The Community Multiscale Air Quality (CMAQ) Model: Updates and Future Development

Jonathan Pleim and the CMAQ development Team, USEPA

An interim version the Community Multiscale Air Quality (CMAQ) Model (version 5.0.2) will be released by December 2013, and will include extended capability for diagnostic output including source apportionment for particulate matter and ozone, the Direct Decoupled Method (DDM) for tracking and output of sensitivities to input perturbations, and the sulfur tracking tool to account for the production of sulfate via the various gas and aqueous phase processes. The CMAQv5.0.2 model will also include community contributed model components for an alternative secondary organic aerosol (SOA) model and new plume-in-grid (PinG) approach. The SOA model is based on the Volatility Basis Set (VBS). The PinG model is known as the Advanced Plume Treatment (APT) which is based on a second-order turbulence plume model with chemistry known as SCICHEM. In both cases the community contributions have been tested and benchmarked against data from the contributor.

Model development is ongoing toward the next major version of CMAQ scheduled for release in 2015. Development research is being conducted in five workgroups that focus on: Chemistry including gas, aqueous, aerosol and heterogeneous; Meteorology including data assimilation and PBL, cloud, and radiation physics; Surface processes including biogenic and geogenic emissions and dry deposition and bi-directional surfaces fluxes; Microphysical processes for both aerosols and clouds; and Model structure which includes computational efficiency, version tracking, and code design. Each workgroup has developed detailed plans for enhancements to CMAQ that will be described in the presentation.

A regional chemical reanalysis prototype

Pius Lee ¹, Greg Carmichael², Tianfeng Chai^1,3, Rick Saylor⁴, Li Pan^1,3, Hyuncheol Kim^1,3, Daniel Tong^1,3, and Ariel Stein¹

1. Air Resource Laboratory (ARL), NOAA, College Park, MD

2. College of Engineering, University of Iowa, Iowa City, IA

3. Cooperative Institute for Climate and Satellites, University of Maryland, College Park, MD

4. Air Resource Laboratory (ARL), NOAA, Oak Ridge, TN

Abstract:

This study aims to launch a prototype chemical reanalysis modeling system. We focus on providing three dimensional reanalysis fields over the conterminous U.S. for gaseous and aerosol species in 12 km horizontal grid spacing and 22 uneven vertical layers between the surface and 100 hPa. Temporally the data are archived hourly and made available to the users at a NOAA portal. The three principal components of the modeling system are the Weather research and Forecasting meteorological model, the data assimilation model based on an optimal interpolation scheme, and the U.S. EPA Community Multi-scale Air Quality Model. We adopt an incremental approach to incorporate observation data sets and to evaluate their individual impact sensitivities. In this initial stage, we started by first assimilating the daily atmospheric column MODIS AOD retrieved by the Terra and Aqua satellites attributable to both fine and coarse mode atmospheric aerosols. Secondly in addition to AOD, we assimilated surface concentrations of PM_2.5 measured by monitors of the AIRNow network at every six hours. Preliminary evaluation of the impacts of these two increments in assimilating observation data was performed and showed that the bulk of contribution in improving the re-analysis fields was attributable to the AOD-ingestion step. Further improvement in either assimilation steps is promising and would be further explored. We also consider additional observation sets such as OMI observed NO₂ column for emission adjustment. We hope to gradually expand the prototype to a comprehensive system deployable for generation of a multiple year reanalysis data base.

A Computationally Efficient Model for Estimating Background Concentrations of NOx, NO2, and O3

Sam Pournazeri* , Si Tan*, Nico Schulte, Qiguo Jing, Akula Venkatram

*These authors contributed equally to the work.

We formulate a Lagrangian model to supplement comprehensive Eulerian grid models such as CMAQ, to estimate concentrations of NO_x, NO₂, and O₃ averaged over a spatial scale of the order of a kilometer over a domain extending over hundreds of kilometers. The model can be used to estimate hourly concentrations of these species over time periods of years. It achieves the required computational efficiency by separating transport and chemistry using the concept of species age. The model computes concentrations by tracing the history of an air parcel reaching a receptor through back trajectories driven by surface winds. Chemical reactions within the parcel are modeled through the Generic Reaction Set (GRS) chemistry model, which approximates the photochemical processes that lead to the production of ozone. The model is evaluated with concentrations measured over two years, 2005 and 2007. Evaluation with data measured at 21 stations distributed over the California South Coast Air Basin (SoCAB) during 2007 indicates that the model provides an adequate description of the spatial and temporal variation of the concentrations of NO₂, and NO_x. Estimates of maximum hourly O₃ concentrations show little bias (less than 10%) compared to observations, and the scatter (s_g² d 2.56 ~95% confidence interval of the ratio of predicted to observed concentrations) is comparable to those associated with more computationally demanding models. The model was also evaluated with data collected at monitors in the San Joaquin Valley Air Basin (SJVAB) in 2005, and it shows similar performance to that at SoCAB. The paper also illustrates the application of the model to 1) screening regions for attainment of statistically based air quality standards, such as that for the daily maximum 8-hour average O₃, and 2) improving methods to interpolate observations.

Adjoint-based climate penalty factor: a perspective on air quality decision-making in a changing world

Amir Hakami, Shunliu Zhao, Amanda Pappin, Morteza Mesbah

Department of Civil and Environmental Engineering

Carleton University, Ottawa, Canada

Of various parameters that will change in a future climate, temperature is believed to affect ozone most significantly. The expected increase in ozone levels due to increasing temperature is referred to as the climate penalty factor (CPF). Values for the CPF have been estimated before from historical ozone and temperature data, or based on forward sensitivity studies.

We use the adjoint of gas-phase CMAQ, to estimate how temperature changes in various locations affect North American nonattainment and mortality metrics. We combine CMAQ adjoint with the decoupled direct method (DDM) sensitivities of chemistry to consider three pathways for estimating influences, i.e. the impact of temperature a) through change in reaction rates, b) change in absolute humidity, and c) increase in biogenic emissions. We find general agreement between our regional adjoint CPF estimations and observed values, but notice significant spatial variability in local CPFs, with great emphasis on urban areas for both exposure and attainment metrics. We further estimate an additional 600 deaths per summer for each degree increase in continental temperatures based on 2007 atmosphere. We will discuss what policy information these values hold, and will explore how future changes in temperature and emissions can have an impact on long-term decision-making for reducing surface ozone levels.

The application and validation of two-way coupled WRF/CMAQ modeling system in China

Wang Jiandong, Wang Shuxiao, Zhao Bin, Wang Long, David Wong, Hao Jiming

With the fast growth of the domestic economy and urbanization, China has become one of the countries with the highest concentration of fine particular matters. To represent and assess the direct radiative effects of aerosol on simulated meteorological parameters and air quality, we applied the two-way coupled WRF/CMAQ modeling system in China, which couples the Weather Research & Forecasting Model (WRF) version 3 and the Community Multiscale Air Quality modeling system (CMAQ) version 5.0. Scenarios with/without feedback were conducted for four months (January, May, August and November) in 2010. The results show that, compared with the scenario without feedback, the radiation on the land surface in the scenario with feedback was reduced owing to the extinction of aerosol. The average reduction was about 23% and it could be 50% in some heavy pollution period in North China Plain. The reduction of radiation leads to a reduction of surface temperature and planetary boundary layer (PBL) height, which could in turn influence the subsequent chemistry transport simulation. The average increase of PM_2.5 concentrations was 15% and it can be more than 40% in heavy pollution period in North China Plain. However, the change of different species of PM_2.5 differs, implying the chemistry reaction rate could be affected during this process. It's supposed to be due to the change of photolysis rate which is affected by the change of radiation. Process analysis is utilized to quantify the contribution of the change of vertical diffusion and the atmospheric photo-chemistry.

A significant source of isoprene aerosol controlled by acidity

Havala Pye, Rob Pinder, Ivan Piletic, Ying Xie, Shannon Capps, Ying-Hsuan Lin, Jason Surratt, Zhenfa Zhang, Avram Gold, Deborah Luecken, Bill Hutzell, Mohammed Jaoui, John Offenberg, Tad Kleindienst, Michael Lewandowski, Ed Edney

Affiliations: US EPA, UNC-Chapel Hill, Alion Science and Technology

Isoprene is a significant contributor to organic aerosol in the southeastern United States where biogenic hydrocarbons mix with anthropogenic emissions. In this work, CMAQ provides explicit predictions of known isoprene-derived aerosol species (2-methyltetrols, 2-methyglyceric acid, and organosulfates) over the United States for the first time. Isoprene aerosol predictions are directly evaluated against ambient data sets quantifying the concentrations of 2-methyltetrols and 2-methylglyceric acid. The new framework includes a mechanistic dependence of isoprene aerosol on acidity and NOx levels. We demonstrate that these pathways more accurately represent the concentration of isoprene-derived aerosol species and respond differently to changes in SOx and NOx emissions than an Odum 2-product model.

Development of a new parameterization for below-cloud scavenging of size-resolved PM by both rain and snow

Xihong Wang¹, Leiming Zhang², and Michael D. Moran^2*

¹Kellys Environmental Services, Toronto, Ontario, Canada
²Air Quality Research Division, Environment Canada, Toronto, Ontario, Canada
*Email Contact: mike.moran@ec.gc.ca

A number of theoretical and empirical size-resolved parameterizations of the scavenging coefficient have been proposed to represent below-cloud particle scavenging by rain. Some but fewer parameterizations exist for below-cloud scavenging by snow. Two recent reviews have suggested that the uncertainty associated with available theoretical parameterizations for rain scavenging spans one to two orders of magnitude while for snow it spans nearly three orders of magnitude. These findings raise concerns over the suitability and representativeness of current individual schemes. A new semi-empirical parameterization for size-resolved as a function of precipitation intensity for both rain and snow has been developed through curve-fitting over a wide range of precipitation conditions using both the sets of existing parameterizations and measurements from field studies. This new scheme can be implemented in any size-distributed particulate-matter model.

Implementation of the Berkeley Dalhousie Soil NOx Parameterization into CMAQ

Benjamin Lash, Dan Cohan, Jesse Bash

Accurate estimation of nitrogen oxide emissions is critical to the simulation of ozone and particulate matter in the atmosphere. Several studies indicate that the Yienger and Levy scheme used to predict soil NO in CMAQ and most other air quality models underestimates emissions by a significant amount. The Berkeley Dalhousie Soil NOx Parameterization (BDSNP) introduced by Hudman et al. 2012 updates soil NO emissions to be more consistent with measurements. The BDSNP is implemented into CMAQ based on a previous implementation in the GEOS-Chem model with minor adjustments. The implementation capitalizes upon the inline biogenics module introduced in the most recent release of CMAQ. BDSNP provides a more physically realistic representation of how soil NO emissions respond to various conditions, including pulsing of emissions following rain events and a more dynamic response of emissions to fertilizer application, nitrogen deposition, and changes in temperature and soil moisture. Preliminary results comparing BDSNP with the Yienger and Levy scheme show soil NO emissions over the Southeast US increase by a factor of 3.5 in a summer 2006 simulation, and in some places soil NO dominates industrial and mobile sources. The updated fertilizer treatment based on Potter et al. 2010 and emission factors from Steinkamp and Lawrence 2011 drive a large fraction of the increase. Emissions are reduced in very wet areas, such as parts of Florida and Louisiana, due to the more complex relationship of NO emissions to soil moisture in BDSNP. Ongoing work will enhance BDSNP to consider year-specific fertilizer inputs, to simulate bidirectional exchange, and to simulate emissions of other nitrogen species.

Implementation and evaluation of new HONO mechanisms in CAMx 6.1

Prakash Karamchandani¹, Chris Emery¹, Barry Lefer², Jochen Stutz³, William Vizuete⁴, Evan Couzo⁴, and Greg Yarwood¹

¹ENVIRON, 773 San Marin Drive, Suite 2115, Novato, CA 94998

²University of Houston, Department of Earth and Atmospheric Sciences

³University of California at Los Angeles, Department of Atmospheric and Oceanic Sciences

⁴University of North Carolina, Department of Environmental Sciences and Engineering

Missing pathways for nitrous acid (HONO) formation are implemented in the latest version (Version 6.1) of CAMx, a photochemical model that is used routinely for regulatory applications in Texas and other areas. This model update is expected to improve the model's ability to simulate ozone concentrations, because HONO is a potential daytime source of the hydroxyl radical, OH, which plays an important role in the ozone formation cycle. Research based on both modeling and field measurements has shown that HONO significantly affects the HO_x budget in urban environments like Houston. Measurements during the Study of Houston Atmospheric Radical Precursor (SHARP) study showed that radical production in the early morning in Houston was dominated by HONO photolysis. Based on current understanding of the important processes governing HONO formation, parameterizations are developed and implemented in CAMx 6.1. The parameterizations make use of a surface model that is added to the core photochemical model. The surface model allows the treatment of the surface as a reservoir of deposited species, and simulates the deposition of chemicals to the surface, sorption and penetration into soils and vegetation, chemical degradation and transformation, and volatilization back into the air (re-emissions). For the HONO parameterizations, the deposited species of interest are NO₂, HNO₃ and HONO, and the surface chemistry includes the photolysis of surface adsorbed HNO₃, the relative humidity dependent conversion of NO₂ to HONO at night, and the photo-enhanced conversion of NO₂ to HONO during the day. Part of the HONO formed by these processes is re-emitted. In addition, the effect of introducing HONO surface emissions as a fraction of NO_x emissions, based on measurements of direct HONO emissions from on-road transport, is investigated. The parameterizations are evaluated and refined using existing modeling databases for the Houston area during the SHARP study period.

On using process-based statistical models of air pollutants to address regulatory and research needs.

Amy Nail, Ph.D.

I will use two process-based statistical models (PBSMs) of ozone-PB SMO VOC and PB SMO-to make the case for development and use of PBSMs of air pollutants to meet the following regulatory and research needs: assessment of the efficacy of past and future emission controls; decomposition of ozone into components attributable to background, creation/destruction from local emissions, and regional transport; exceptional event analyses; space-time prediction backward in time for exposure quantification; and mutual evaluation/validation with other models.

The first model, PB SMO VOC, produces the following four output fields: process-based ozone, process-based VOCs, the error in the process-based ozone field, and the error in the process-based VOC field. The process-based ozone field can be broken down into background, local creation/destruction, and transport fields. It is a hierarchical model of 8-hour ozone as a function of meteorology, observed NOx, and a latent (unobservable) VOC field. The VOC field is a function of meteorology and VOC emissions from different sectors, and it might be interpreted as the reactivity-weighted VOC field where the weight applied to emissions from different sectors are those most relevant to ozone production. The ozone and VOC error fields quantify the uncertainty in the ozone and VOC fields attributable to inaccuracies in the respective process models.

Though PB SMO VOC will produce the daily 8-hour ozone value at latitude and longitude resolution, the spatio-temporal resolution of the inputs was coarser. The anthropogenic emissions data was the annual total for each county, broken down into the following categories: on-road, non-road, storage & transport, and other area sources. The biogenic emissions were available at the monthly resolution for a single county. Despite these limitations, the root mean squared error of predictions of 8-hour ozone in a withheld dataset compares favorably to that for CMAQ predictions, but the performance of CMAQ with respect to getting the highest values of 8-hour ozone exceeds that of PB SMO VOC.

The second model, PB SMO, also produces the decomposable, process-based ozone field, and an ozone error field, but no VOC or VOC error field. This time the VOC emissions were in the form of hourly SMOKE output at the 36 and 12 km resolution, and they were broken down into 10 species categories. Note that the latent VOC field could be included or excluded with these new inputs. Its inclusion would add computation time, but allows learning about the VOC field, the relative contributions of the different species, and comparison with chemical mechanisms in other models. Using the 36 km VOC inputs, the RMSE of PB SMO was about the same as for PB SMO VOC and CMAQ, but using the 12 km inputs, the RMSE of PB SMO was far lower than all of the others.

The performance of 36 km PB SMO with respect to predicting the highest values of ozone was much better than that of PB SMO VOC, and the performance of 12 km PB SMO was even better than that, but the performance of CMAQ was still superior. These performance differences are probably due in part to the fact that the withheld dataset had a value of ozone that was higher than any in the dataset to which the model was fit. Another likely explanation is that both PB SMO VOC and PB SMO were fit to 8-hour ozone rather than to hourly ozone. Fitting to hourly ozone, and then aggregating to 8-hour ozone might result in a PBSM of ozone that does a better job of predicting the highest ozone values.

At the end of the talk I will discuss some responses to this work and my subsequent thoughts about varied perceptions of different model types. I will close by noting the modest amount of time and resources used to produce PB SMO VOC and PB SMO, to further the case for investing in their development and use.

Perspective on urban canopy modeling for weather climate and air quality applications

Jason Ching, Gerald Mills, Johannes Fedemma, Keith Oleson, Linda See, Iain Stewart, Marina Neophytou and Adel Hanna

Environmental issues and impacts to society will be exacerbated with increased population, diminishing resources and climate changes. Current models available for weather, climate and air quality applications are powerful state-of-science modeling systems can be employed to address the impact of these issues. This presentation reviews a selected subset of such systems, considered representative of community-based publically available modeling systems and focus on their utilization for urban applications. Special attention is required given the complex and high degree of spatial inhomogeneity of the underlying surface areas. Such applications optimally require relatively fine grid meshes and scale appropriate science description for the varied and complex land surface atmospheric processes commensurate to the fine scale land surface variability structure. We briefly review means and science parameterizations for urban focused modeling in these major modeling systems. Several issues, limitations as well as innovative opportunities specific to the optimal operations of these urban systems, with focus on fine mesh size and data needs including global coverage of city specific gridded morphology are identified and discussed.

Methodology to Visualize and Validate Modelling Data with Open-Sourced Software

Hamish Hains, M.A.Sc., EIT

Dr. Xin Qiu, Ph.D., ACM, EP

Air contaminant modelling is continuously improving as advancements in the understanding of pollutant transport and computational ability improve model accuracy. In contrast to the advances being made to numerical air quality and meteorological models, the graphic presentations being produced are lacking in many aspects. Tools exist which allow for rapid and accurate visualization of geospatial data which are generally being under-utilized. For example, there are numbers of freeware or open-sourced GIS based software packages available which can aid in data visualization and analysis. The ability to integrate these model results with geospatial data (e.g. measured meteorological data, source locations) leads to a better understanding of the model and provides a fast and accurate way to communicate its results. Novus has developed a method to rapidly display and validate the results of next-generation air quality and meteorological models using entirely open source software, such as Quantum GIS.

Preliminary analyses of flight measurements and CMAQ simulation during Southeast Nexus (SENEX) field experiment

Li Pan ^1,2, Pius Lee ¹, Hyun Cheol Kim ^1,2, Daniel Tong ^1,2 ,Rick Saylor ³ and Tianfeng Chai ^1,2

¹ NOAA/Air Resources Laboratory, College Park, MD

² UMD/Cooperative Institute for Climate and Satellites, College Park, MD

³ NOAA/ARL/Atmospheric Turbulence and Diffusion Division, Oak Ridge, TN

Southeast Nexus (SENEX) is a NOAA field study conducted in the Southeast U.S. in summer 2013. The NOAA WP-3D research plane made on-board measurements of trace gases and atmospheric aerosol downwind from different source regions (urban, point and forest). The field intensive targeted the investigation of the interactions between natural and anthropogenic emissions and their impact on air quality and climate change. To support this field campaign, NOAA/Air Resources Laboratory provides high horizontal grid resolution (4km) CMAQ model simulations covering the Southeast U.S. to conduct a post-campaign analysis.  The 4km CMAQ simulation is driven by time varying lateral boundary conditions extracted from a 12km CMAQ simulation over the continental United States.  Fire emissions derived from HMS (Hazard Mapping system) and Smoke Blue-sky calculation are included in both CMAQ simulations. In this study we mainly focus on two aspects: The synoptic perspective to distinguish general findings and major events such as large wildfires and pressure fronts. Secondly, the process level prospective to understand the differences between CMAQ simulations and campaign measurements to identify the contribution from major model uncertainties such as input uncertainty (emissions) and model algorithm uncertainty (chemical and physical mechanism).  

AirNow Satellite Data Processor: Improving EPAs AirNow Air Quality Index Maps Using NASA/NOAA Satellite Data

Adam N. Pasch¹, Patrick H. Zahn¹, Jennifer L. DeWinter¹, Michael D. Haderman¹, James J. Szykman²_, John E. White³, Phillip Dickerson³, Timothy S. Dye¹, Aaron van Donkelaar⁴, and Randall V. Martin⁴

¹Sonoma Technology, Inc., Petaluma, CA, USA

²NASA Langley Research Center, Hampton, VA, USA

³U.S. Environmental Protection Agency, Research Triangle Park, NC, USA

⁴Dalhousie University, Halifax, Nova Scotia, Canada

The U.S. Environmental Protection Agency's (EPA) AirNow program provides the public with easy access to national air quality information using the Air Quality Index (AQI), which is a standardized metric based on health effects. AQI levels are currently calculated from ground-based measurements and are interpolated to a grid to create maps that cover national, regional, and local spatial scales. The ground monitoring network has significant gaps in coverage in some parts of the United States; mapped AQI levels have high uncertainty in monitor-sparse locations.

To improve the usefulness of AQI maps, we developed the AirNow Satellite Data Processor (ASDP) to integrate ground-based measurements of surface PM_2.5 concentrations with satellite-estimated concentrations. The satellite estimates are derived from NASA/NOAA satellite aerosol optical depth (AOD) retrievals and from the ratios of surface PM_2.5 concentrations to AOD as modeled by GEOS-Chem. The ASDP uses a weighted-average method based on the uncertainty in the satellite estimates and the interpolated surface PM_2.5 observations to generate AQI maps with improved spatial coverage. The ASDP is designed to be flexible to additional datasets, such as future satellite data products, data from other surface observation networks, and air quality model data. The goal of ASDP is to provide more detailed AQI information in monitor-sparse locations and to augment monitor-dense locations with more information.

Our initial evaluation of the ASDP showed that satellite data adds information to current AirNow maps in regions lacking sufficient monitoring (e.g., the Intermountain West and the Great Plains). However, we found that the performance of the ASDP fused PM_2.5 concentration predictions was impacted by several factors, including the calculation method for observation data uncertainty and the test site selection procedure. To address these issues, we (1) improved the uncertainty calculation for ground-based PM_2.5 concentration data to be consistent with the approach used for satellite-estimated data; and (2) based the selection of test sites on data completeness, distance to nearest monitors, and concentration gradients between monitors. We will present a statistical analysis for 2010-2012 of ASDP predictions of PM_2.5 concentrations focusing on performance at test sites, as well as case studies evaluating the performance for multiple regions and seasons. Furthermore, we will illustrate how the ASDP applies to air quality modeling.

Characterization of Gulf of Mexico Background Ozone Concentrations

1. Jim Smith
2. Fernando Mercado
3. Mark Estes

A perplexing problem with model performance in the Houston/Galveston area is the frequent over-prediction of ozone concentrations in air arriving from the Gulf of Mexico. This work attempts to characterize the air masses arriving at near-shore sites on the Texas coast by calculating three-dimensional back-trajectories, associating each with observed ozone concentrations at the trajectory's terminus, and finally using a clustering algorithm to characterize groups of like trajectories. Color-coding the trajectories according to the ozone concentrations at the terminus provides a highly intuitive approach to characterizing the sources of background ozone.

Reconstructing fire records from ground-based routine aerosol

Hongmei Zhao^{1, 3}, Daniel Q. Tong^{2, 3}, Pius Lee³, Hyuncheol Kim^2,3, Hang Lei³

¹ Key Laboratory of Wetland Ecology and Environment, Northeast Institute of Geography and Agroecology, Chinese Academy of Sciences, Changchun, Jilin, 130102, China

² UMD/Cooperative Institute for Climate and Satellites, College Park, MD, 20740, USA

³ NOAA/Air Resources Laboratory, College Park, MD, 20740, USA

We report our recent efforts to reconstruct a long-term record of fire events based on observations from a continuous ground-based aerosol monitoring network in continental United States. Long-term records of fire events are important in analyzing air quality and climate change at both region and global scales. During a fire episode, high-level aerosols and thick smoke are usually recorded by the ground-based monitors and satellite sensors. Based on the physical and chemical characteristics of fire-dominated aerosols reported by literature, we analyzed the surface aerosol observations from the Interagency Monitoring of Protected Visual Environments (IMPROVE) network during satellite-detected fire events to establish a suite of indicators to identify fire events from routine aerosol monitoring data. Five identification criteria are chosen to pinpoint a fire event: (1) high PM_2.5, PM₁₀ (particles smaller than 2.5 and 10 in diameters) concentrations; (2) higher PM_2.5/PM₁₀ ratio; (3) higher organic carbon (OC/PM_2.5) and element carbon (EC/PM_2.5) ratios; (4) high potassium (K/PM_2.5) ratio and (5) low soil/PM_2.5 ratio. Using those criteria, we are able to identify fire events close to 15 IMPROVE monitoring sites from 2001 to 2011. Most of them were located in the western (California and Montana) and central (Missouri, Oklahoma and Texas) United States. In any given year within the study period fire events often occurred between April and September, especially in the two months of April and September. The ground-based fire climatology is also consistent with that observed by satellites retrievals. This suggests that records of fire events can be reconstructed based on a continuous ground-based aerosol monitoring network to some extent. However, caution needs to be exercised that those indicators are based on limited number of fire events, and should be tested and confirmed by further research.

A NEW PARALLEL VERSION OF MOVESMRG FOR SMOKE

George Delic, HiPERiSM Consulting, LLC, P.O. Box 569, Chapel Hill, NC 27514

This presentation reports on the first major thread parallel enhancements for the SMOKE 3.1 model as developed by the U.S. EPA. These modifications replaced the serial version of the MOVESMRG Fortran procedure with a thread parallel version and removed several serial coding constructs while also streamlining I/O operations. This procedure merges emission factors with activity data to create the on-road mobile emissions input required by CMAQ. It is one of the most time-consuming components of SMOKE and therefore it is of interest to reduce wall clock time. In the thread parallel version of MOVESMRG performance limiting procedures were hoisted outside the parallel region and preprocessed in a separate serial loop on the time step (T) followed by a second parallel loop on T. This new parallel version (hereafter SMOKE-HC) was tested with data provided by the U.S. EPA [1] for five cases with increasing number of time steps in each, corresponding to T=25, 43, 73, 97, and 121 simulation hours (all with 146 reference counties). Results with the Portland compiler are reported for Intel (INTEL) and Advanced Micro Devices (AMD) platforms with 8 and 48 cores, respectively. The rate per distance simulation consumes the largest fraction of wall clock time in MOVESMRG. With increasing T step values results showed that both thread scaling and speedup (over the standard EPA version), improved steadily. For example, with T=121 time steps the multi-threaded parallel speedup over the EPA version was better than 2.3 (INTEL, 8 threads), and 3.2 (AMD, 24 threads), respectively. Trends in values for the parallel wall clock times in the five cases (T=25-121) were accurately fitted with linear trend lines as a function of T. These trend lines, when extrapolated to T=433, showed parallel scaling could reach 3.3 (INTEL, 8 threads), or 5.1 (AMD, 24 threads). These extrapolations were also used to simulate workloads of multiple concurrent SMOKE-HC runs. One possible workload metric measures how many weeks of simulation (multiples of 7 x 24 = 168 hours) complete for each wall clock hour. Sample results showed a best metric value of 1.3 on the INTEL node, with 2 x 4-thread runs at T=265. Whereas, an AMD workload exceeds this with a metric value of 1.37 for 3 x 12-thread runs at T=217 each. Thus, although the INTEL platform is the speedier of the two, there is more opportunity for increased SMOKE workload throughput with SMOKE-HC on the AMD node because of the higher core count.

[1] The author gratefully acknowledges SMOKE input data provided by B.H. Baek (UNC Institute for the Environment) and the Emissions Inventory and Analysis Group, OAQPS, U.S. EPA.

Optimizing a coated-sphere module for use in coupled WRF-CMAQ

David Wong and Francis Binkowski

A widely used coated-sphere module, BHCOAT (Bohren and Huffman, 1983), has been adopted for use in coupled WRF-CMAQ applications. This module calculates the efficiencies for extinction, total and back scattering as well as the asymmetry factor for a particle consisting of an absorbing spherical core surrounded by a non-absorbing or weakly absorbing shell.

Our starting base version was from a standard double-precision implementation of the Bohren and Huffman code by Prof. Bruce T. Draine of Princeton University, from whom we obtained the code.

During the model-testing phase, we encountered numerous cases that caused the failure in BHCOAT algorithm. We have examined the algorithm with 32 specific cases and found that in four of those cases, intermediate calculation ends up with NaN (Not-a-Number) value. Only one case exceeded the limits that Bohren and Huffman had suggested. Such a NaN value is caused by a denominator becomes zero due to subtraction of two similar numbers. We applied a classical approach by increasing the real precision to 16 bytes (32 byte complex) . However, the run time increased up to 32 fold.

We then devised a simple approach to deal with this numerical precision issue without increasing the precision beyond standard double-precision. We also optimized the code to improve computational performance for our applications. The final version of the algorithm reduced of the execution time to a range of 43% to 96% with respect to the 32 byte complex version. We will present example calculations to illustrate this new approach and optimization the techniques.

I/O Analysis for the Community Multiscale Air Quality (CMAQ) Model

David Wong¹, Cheng-En Yang², Joshua S. Fu², Kwai Wong², Yang Gao³

¹U.S. Environmental Protection Agency, Research Triangle Park, NC, USA

²University of Tennessee, Knoxville, TN, USA

³Pacific Northwest National Laboratory, Richland, WA, USA

This study attempts to develop a renovated module of true parallel data input and output (I/O) in the Community Multiscale Air Quality (CMAQ), to improve its I/O efficiency. Various I/O methods which include the one being used in the current CMAQ model, parallel netCDF (pnetCDF, and parallel netCDF with domain collapsing approach, were devised and a series of experiments with a "pseudo" I/O module were performed. pnetCDF parameters: stripe counts and size, were tested to determine the optimal setting on various machines. The efficiency for each method is evaluated through three different sizes of domains: small, medium, and large domain. Results show three major findings: (1) I/O performs better with stripe size 1 Megabyte (MB) and 2 MB than any other stripe size; (2) higher I/O performance is found with stripe count 8 and 11; (3) the collapsing method, reducing the maximum I/O time by 62-96% compared to the regular NetCDF method. This new approach has been implemented in the CMAQ model and currently it is undergoing testing.

Comparisons of CMAQ model performance over the Northeast United States as a function of grid resolution (12km vs. 4km) for a 2007 annual model simulation

Pat Dolwick, Kirk Baker, James Kelly, Chris Misenis, Sharon Phillips, Norm Possiel, Heather Simon, and Brian Timin

As computing power increases, it is increasingly feasible to apply regional chemical transport models (CTM) at finer scales over relatively large regions. Finer scale model applications have the potential of yielding better linkages between air quality and the health studies of the populations exposed to those air quality concentrations. However, the fundamental question of whether a given finer-scale CTM simulation represents a more accurate characterization of urban air quality will need to be answered before one should consider its use for more detailed health studies. This analysis presents results from a 2007 annual CMAQ simulation modeled with a horizontal grid resolution of 4km over a 174 x 201 domain covering the northeast U.S. This presentation compares the results of an operational model evaluation for this finer scale model output against a similarly-constructed 12-km grid over the same geographic area. In addition to the statistical summaries of model bias and error as a function of resolution, this analysis also compares the depiction of specific features that are important within the northeast U.S. In particular, the report looks at: 1) how model performance varies between the two simulations during periods in which interactions with sea- or bay-breezes are resulting in poor air quality, and 2) how model performance varies at urban, suburban, and rural sites over the two grids.

Inter-comparison of Photochemical Modeling by EPA and CARB for the Calnex 2010 Study

James Kelly and Kirk Baker

Office of Air Quality Planning & Standards, US EPA, RTP, NC 27711

Chenxia Cai, Jeremy Avise, and Ajith Kaduwela

Planning and Technical Support Division, California Air Resources Board, Sacramento, CA 95812

Despite much progress in recent decades, many regions of California suffer from poor air quality due to ozone and particulate matter (PM) pollution.  To develop optimal emission control strategies, air quality modeling systems are needed that can accurately predict the response of pollution concentrations to emission reductions.  However, simulating air quality in California is challenging due to the complex terrain, diverse emissions, and other complicating factors.  In May-June 2010, the Calnex field study was conducted to answer scientific questions about emissions, chemistry, climate, transport, and meteorology in California.  The study provides a rich dataset of observations from aircrafts, ships, and supersites, which are supplemented by California’s routine monitoring networks.  This dataset is being used by the California Air Resources Board (CARB) and US EPA in diagnostic evaluations of their photochemical modeling systems to improve understanding of their modeling capabilities and to identify areas for model development.  In this study, predictions of the CARB and US EPA modeling platforms for fine-scale (4-km) simulations of the Calnex 2010 study period are inter-compared and evaluated against observations.  An emphasis is placed on isolating the causes of differences in predictions for PM components, ozone, and precursor species.  Future work will involve improving both modeling systems based on the findings of this inter-comparison and evaluation. 

Examining the impacts of emissions from single-sources on PM2.5 and ozone using a photochemical model

James Kelly and Kirk Baker

US EPA, Office of Air Quality Planning & Standards, RTP, NC

Estimates of the impacts of emissions from single sources on criteria pollutant concentrations are useful in a variety of regulatory applications related to New Source Review, Prevention of Significant Deterioration, and National Environmental Policy Act provisions.  Modeling approaches for estimating the single-source impacts of primary pollutant emissions (e.g., direct PM2.5) are relatively well-developed compared with those for estimating the impacts of secondary-pollutant precursor emissions (e.g., NOx and VOCs).  We previously presented results indicating that photochemical model predictions based on 4-km horizontal resolution compared reasonably well with in-plume aircraft observations downwind of the Tennessee Valley Authority Cumberland power plant.  Here, we present photochemical model predictions of the impacts of hypothetical new ground-level sources on PM2.5 and ozone concentrations in highly polluted areas of California.  The simulations were conducted with the Community Multiscale Air Quality (CMAQ) using 4-km horizontal resolution for short (~10 days) winter and summer periods in 2007.  The maximum impacts of emissions on ozone and PM2.5 concentration are presented as a function of emission species (i.e., NH3, SO2, VOC, NOx, and EC) and downwind distance.  The advantages and disadvantages of using photochemical models for estimating single-source secondary impacts are also discussed.  Future work will focus on comparing these results with those of other single-source modeling approaches to inform the development of screening tools for estimating single-source impacts on secondary pollutants.

Meteorological Evaluation to Support a 2011 Air Quality Modeling Platform

Chris Misenis

An ongoing effort is being made to evaluate the performance of the Weather Research and Forecasting (WRF) model for its use in supporting the development of a 2011 air quality modeling platform. A 12-km grid resolution simulation was conducted over the continental United States. WRF version 3.4 with the Advanced Research WRF (ARW) core was implemented, with several notable physics options: Pleim-Xiu land surface model, Asymmetric Convective Model version 2 planetary boundary layer scheme, Kain-Fritsch cumulus parameterization, Morrison double moment microphysics, and the YYTMG longwave and shortwave radiation schemes. A 35-layer vertical spacing was used with a 20-m lowest layer. The WRF model was initialized using the 12 km NAM analysis provided by NCDC.  Additionally, updated data assimilation techniques as described by Gilliam et al. (2012) were implemented. Surface observations of temperature, mixing ratio, wind speed and direction will be compared with model outputs. Monthly spatial comparisons of accumulated rainfall and diurnal measurements of incoming shortwave radiation will also be presented.

Development of a 2007 Fine Scale Platform: Meteorological Evaluation

Chris Misenis, Kirk Baker, Pat Dolwick

As part of an effort to understand the benefits of fine-scale meteorological and photochemical modeling, the Weather Research and Forecasting (WRF) model was applied to several domains centered over the following areas: 1) California, 2) Northeast Corridor, 3) Atlanta, 4) Detroit. By using high-resolution modeling techniques, we are better able to understand area-specific processes that have significant impacts on downstream photochemical simulations. WRF was applied using a 4-km horizontal resolution for the entire year of 2007. Initialization was done by nesting each region with a 12-km CONUS domain that was initialized using 12-km NAM analysis data. The analysis of these simulations is multi-faceted: 1) A comparison to routine meteorological observations (surface temperature, mixing ratio, wind speed and direction), monthly precipitation data and incoming shortwave radiation observations (where available), 2) Differences between 12-km and 4-km resolution fields in each domain, and 3) Examination of the characterization of region-specific meteorological processes (e.g., land/ocean interface) for each domain. These simulations are used for input into CMAQ simulations to better understand the impact of fine-scale meteorological data.

Calculation of sensitivity coefficients for airport emissions in the Continental United States using CMAQ DDM-3D/PM

Scott Boone, Sergey Napelenok and Saravanan Arunachalam

Sensitivity modeling techniques have been at the core of environmental policy development since the first appearance of atmospheric chemical transport models. Methods in current use include the subtractive (brute force) and response surface methods. All such techniques, however, are limited in scope or resolution by available computational power. As a result, impact assessment for several emissions sectors from large to small, including the growing aviation sector, have thus far been implemented only at the domain-wide or regional scales. By using the direct decoupled method in three dimensions for fine particulate matter (DDM-3D/PM), a newly-implemented advanced sensitivity analysis technique for CMAQ, we seek to characterize the sensitivity of atmospheric PM_2.5 concentrations to aircraft emissions during the landing and takeoff (LTO) cycles (occurring within the lowest 3,000 ft of the atmosphere) from each of the major U.S. airports. We use emissions inventories from the recently developed global-scale Aviation Environmental Design Tool (AEDT) for the year 2005. We will present the design of experiments to perform this modeling study that will provide individual airport-level sensitivities, as opposed to previously available U.S.-wide sensitivities to combined aircraft emissions from all U.S. airports. The results from this study should help characterize the air quality impacts and target policy goals achievable by specific precursor-wise emissions reductions on an airport-by-airport basis.

Assessing sources of ozone damages to human health and ecosystems with the CMAQ adjoint

Shannon Capps, Rob Pinder, Ellen Cooter, Matthew Turner, Daven Henze, Peter Percell, Shunliu Zhao, Matthew Russell, Amir Hakami

Tropospheric ozone (O₃) degrades human and ecosystem health. Controlling it is challenging due to its secondary formation and relatively long atmospheric lifetime (approx. 22 days). Here, we present a modeling framework that elucidates the spatially-resolved and species-specific emissions that lead to detrimental ozone effects, which range from urban to rural, acute to chronic. The CMAQ adjoint simultaneously calculates the relative influence of each emissions source in the domain on selected outcomes. We have augmented the model, which has been used in the past for human exposure and mortality assessments, to also account for ecosystem impacts on crops, timber, and federally-protected areas.

In particular, we determine the yield reduction of specific crops and timber species due to chronic O₃ exposure using a sigmoidally-weighted sum of O₃ over the summer months (W126), which is unique to each crop or tree species distribution. Furthermore, vegetative damage in federally-protected wilderness areas or national parks is estimated by a threshold-based cumulative exposure metric modified by the vapor pressure deficit (mAOT30). Finally, we calculate human mortality in a manner consistent with BenMAP. With these new developments, we assess the relative influence of emissions on human health and ecosystem degradation in the continental U.S. during the summer of 2007.

Towards building better linkages between aqueous phase chemistry and microphysics in CMAQ

Fahey, K., Wong, D., Hutzell, W., Yu, S., and J. Pleim

Atmospheric Modeling and Analysis Division, National Exposure Research Lab, U.S. EPA, RTP, NC 27711

 

Currently, CMAQ’s aqueous phase chemistry routine (AQCHEM-base) assumes Henry’s Law equilibrium and employs a forward Euler method to solve a small set of oxidation equations, considering the additional processes of Aitken scavenging and wet deposition in series and employing a bisection method to calculate H⁺ concentrations.  With potentially hundreds of reactions that may be important in cloud water and only seven reactions in the current model, expansion of the existing mechanism is an important area of investigation.  However, with the current mechanism hardwired into the solver code, the module is difficult to expand with additional chemistry.  It also ignores the impacts of mass transfer limitations on cloud chemistry which may be significant.  Here, the Kinetic PreProcessor has been applied to generate a Rosenbrock solver for the CMAQ v5.0.1 aqueous phase chemistry mechanism.  The module has been updated to simultaneously solve kinetic mass transfer between the phases, dissociation/association, chemical kinetics, Aitken scavenging, and wet deposition.  This will allow for easier expansion of the chemical mechanism in the future and a better link between aqueous phase chemistry and droplet microphysics.

Evaluating and improving atmospheric chemical mechanisms used for modeling ozone formation from alkenes

Gookyoung Heo,^a William P. L. Carter,^a Peng Wang,^b Qi Ying,^b Ron Thomas^c

^aCollege of Engineering, Center for Environmental Research and Technology, University of California, Riverside, CA 92521, USA

^bZachry Department of Civil Engineering, Texas A&M University, 3136 TAMU, College Station, Texas 77843-3136, USA

^cTexas Commission on Environmental Quality, 12100 Park 35 Circle, Austin, TX 78753, USA

Abstract: As air quality standards become more stringent to protect human health, more accurate and reliable scientific information is needed to formulate scientifically-sound and cost-effective air quality policies, particularly for secondary pollutants such as ozone. Environmental chamber data generated under well-controlled experimental conditions (e.g., without involving uncertainties in emissions and meteorology while minimizing the impact of chamber artifacts) are useful in evaluating chemical processes that contribute to ozone formation. Mechanism evaluation using chamber experimental data generates evidence of credibility of the chemical mechanism (e.g., SAPRC-07) used in air quality modeling by providing comparison of experimentally-measured and model-simulated concentrations of key species such as ozone. In this study, we generated chamber experimental data that can be used to evaluate mechanisms for 5 of the alkenes classified as the Highly Reactive Volatile Organic Compounds and regulated in southeast Texas (1,3-butadiene, isobutene, cis-2-butene, trans-2-butene, 1-butene) and 5 additional alkenes commonly observed in urban atmospheres in the U.S. (1-pentene, 1-hexene, cis-2-pentene, trans-2-pentene, 2-methyl-2-butene), and improved the mechanisms in SAPRC-07 used for modeling ozone formation from alkenes. Using the improved mechanisms for alkenes in air quality models such as the Community Multiscale Air Quality modeling system (CMAQ) will contribute to (1) increasing credibility of the chemical mechanism used for air quality modeling, (2) reducing uncertainty in the chemical mechanism used for modeling ozone formation from alkenes, and (3) assessing the impact of industrial alkene emissions in southeast Texas on ozone formation. We will present major findings based on newly obtained chamber experimental data and key improvements made to the mechanisms in SAPRC-07 for modeling ozone formation from alkenes, and we will also present preliminary results of CMAQ simulations for the Texas Air Quality Study 2006 (TexAQS) while focusing on alkene chemistry.

Key words: chemical mechanism, condensed chemical mechanism, mechanism evaluation, environmental chamber data, alkene, ozone, air quality modeling

Source Sensitivity of Secondary Inorganic Fine Particles Based on Sectors and Regions Using the GEOS-Chem Adjoint Model

 

Xin Li^{1, 2}, Yang Zhang², Qiang Zhang¹, Kebin He³, Daven K. Henze⁴, Xiaoye Zhang⁵

¹Center of Earth System Sciences, Tsinghua University, Beijing, 100084, China

²Air Quality Forecasting Laboratory, Department of Marine, Earth, and Atmospheric Sciences, North Carolina State University, Raleigh, NC, 27606, USA

³School of Environment, Tsinghua University, Beijing, 100084, China

⁴Department of Mechanical Engineering, University of Colorado, Boulder, Colorado, USA

⁵Key Laboratory of Atmospheric Chemistry, Chinese Academy of Meteorological Sciences, CMA, 46 Zhong Guan Cun S. Ave., Beijing, 100081, China

 

China suffers severe pollution of fine particles (PM_2.5) for many years. Strict control of SO₂ emissions has been enforced since 2005 and resulted in significant reductions of sulfate; however, high PM_2.5 concentrations remain and NO_x and NH₃ are now believed to be important species to further control inorganic fine particles. This work aims at exploring the policy of emission control in China, by calculating the sensitivities of surface concentrations of inorganic fine particles in Beijing to three precursor gases (i.e., SO₂, NO_x, and NH₃) in each sector. The GEOS-Chem Adjoint model is applied to East Asia (11°S~55°N, 70°E~150°E) at a grid resolution of 0.5° (Latitude) × 0.67° (Longitude). Emission inputs and the division of anthropogenic sectors are based on the Multi-resolution Emission Inventory for China. Surface concentrations from the forward simulation are evaluated by several observational datasets. The model underpredicts concentrations of surface sulfate but overpredicts those of surface nitrate and ammonium, with annual normalized mean biases of -31.4%, 41.9%, and 29.8%, respectively.

Previous work demonstrated that inorganic particles are more sensitive to NO_x emissions than to SO₂ emissions in North China, because the dominant component in this area is changing from sulfate to nitrate. According to the source sensitivity results in this work, with the same level of emission control, the reduction of NO_x emissions is more efficient than the reduction of SO₂ emissions, especially in winter, in reducing PM_2.5 concentrations, which is consistent with previous work. The GEOS-Chem adjoint model simulations identify the top three major emission sectors to be agriculture, residential, and transportation in winter and agriculture, industry and power plant in summer. To examine the most influential emission areas, north China is divided into 10 subareas (including a local source area that covers Beijing) mainly based on province boundaries. The top four contributor areas are North Hebei (36.1%), local (31.6%), Neimenggu (6.5%), and Liaoning (5.2%) in winter and North Hebei (31.7%), local (27.4%), Shandong (13.0%), and South Hebei (12.9%) in summer. These results identify the most influential source emissions and areas for PM_2.5 pollution control, thus providing a scientific basis for the determination of the national emission control priority. 

Implementing explicit SOA aqueous chemistry in CMAQ

Neha Sareen, Kathleen Fahey, Bill Hutzell, Ann Marie Carlton

Aqueous multiphase chemistry in the atmosphere has a substantial impact on climate and can lead to air quality changes that adversely impact human health and the environment. The chemistry is complex because of the variety of compounds present in the atmosphere and the phase transitions associated with multiphase reactions. These reactions can lead to the formation of secondary organic aerosols (SOA_AQ) in the atmosphere. When included, current photochemical models typically use a simple parameterization to describe SOA_AQ formation. Here, we discuss the implementation of explicit aqueous SOA chemistry in a box model of the CMAQ 5.0.1 aqueous phase chemistry mechanism using the Kinetic PreProcessor (KPP). The expanded chemistry model includes reactions of glyoxal, methylglyoxal, and glycolaldehyde as precursors to form SOA_AQ and is based on the mechanism from Lim et. al. 2010 (1). The current aqueous phase chemistry module in CMAQ uses a forward euler method to solve the system of oxidation equations, estimating the pH with a bisection method assuming electroneutrality, and multiphase processes are solved sequentially. This is not robust for systems with large dynamic range (e.g., multiphase systems), and inhibits expansion of the aqueous phase chemical mechanism to adequately incorporate the growing body of literature that describes multiphase organic chemistry. The KPP solver allows for all processes to be solved simultaneously and facilitates expansion of the current mechanism. Addition of explicit organic reactions and H₂O₂ photolysis in the KPP box model results in increased mass of organic aerosol and more realistic predictions. For particulate matter focused air quality management strategies to be effective, it is important that models move away from the yield-based approach currently used and expand to include more explicit organic chemistry.

 

(1) Lim, Y. B., Tan, Y., Perri, M. J., Seitzinger, S. P., and Turpin, B. J.: Aqueous chemistry and its role in secondary organic aerosol (SOA) formation, ACP, 10, 10521-10539, 2010.

 

Impacts of CB6 and CB05TU chemical mechanisms on air quality

Feng Liu and Golam Sarwar

In this study, we incorporate the newly developed Carbon Bond chemical mechanism (CB6) into the Community Multiscale Air Quality modeling system (CMAQv5.0.1) and perform air quality model simulations with the CB6 and the existing Carbon Bond 2005 chemical mechanism with updated toluene chemistry (CB05TU) for May, 2011. The modeling domain covers southern part of Canada, United States, and northern portion of Mexico using 36-km horizontal grid-resolution and contains 24 vertical layers. We use meteorological fields predicted by the Weather Research and Forecasting (WRF) (version 3.5) model for the study. We generate model-ready emissions for the two mechanisms by using the Sparse Matrix Operator Kernel Emission (SMOKE, version 3.1). Anthropogenic emissions are derived from the 2008 National Emissions Inventory while biogenic emissions are derived from the Model of Emissions of Gases and Aerosols from Nature (MEGAN, version 2.1). We use the predefined clean air vertical profiles for boundary conditions. We compare ozone, hydroxyl radical, hydrogen peroxide, peroxy acetyl nitrate, organic nitrate, and selected speciated organic and inorganic aerosol species predicted by the two mechanisms. The accompanying paper describes a detailed analysis of the model predictions obtained with the two mechanisms along with a comparison of the predictions with observed data.

Examination of sulfate production by CB05TU, RACM2, and RACM2 with SCI initiated SO2 oxidation in the Northern Hemisphere

Golam Sarwar, James Godowitch, Kathleen Fahey, Rohit Mathur

 

We employ the hemispheric Community Multiscale Air Quality model to examine tropospheric sulfate production in the northern hemisphere using the Carbon Bond 2005 chemical mechanism with updated toluene chemistry (CB05TU) and the Regional Atmospheric Chemistry Mechanism (RACM2) without and with sulfur dioxide (SO₂) oxidation by Stabilized Criegee Intermediate (SCI). The base model accounts for gas-phase SO₂ oxidation by hydroxyl radical and five aqueous-phase sulfur oxidation pathways: (1) hydrogen peroxide (2) ozone (3) peroxyacetic acid (4) methyl hydroperoxide and (5) oxygen catalyzed by manganese and iron. We modify RACM2 to explicitly represent several SCIs from alkene/ozone reactions. These SCIs subsequently react with SO₂ to produce sulfuric acid which partitions to particulate sulfate. The formation of sulfuric acid via this pathway has been traditionally thought to be a small contributor to tropospheric sulfate budgets, though with high uncertainty in the formation rate. Recent laboratory chamber experiments directly measured the rate coefficient for SO₂ oxidation by SCI and reported a high value. SCIs also react with water monomer, water dimer, and other chemical species and these reactions are accounted for in the study. We combine the detailed SCI chemistry with RACM2 and perform three air quality model simulations for summer 2006 using the following chemical mechanisms: (1) CB05TU (2) RACM2 and (3) RACM2 with detailed SCI chemistry. Concentration changes between the first two simulations are attributed to the differences between the CB05 and RACM2 chemical mechanisms; while changes between the second and third simulations are attributed to SCI initiated oxidation. CB05TU predicts the lowest sulfate concentrations. RACM2 enhances sulfate concentrations compared to those with CB05TU due primarily to enhanced gas-phase SO₂ oxidation by hydroxyl radical and aqueous-phase oxidation by ozone. The SO₂ oxidation by SCI further enhances sulfate. The accompanying paper presents a detailed analysis of predicted sulfate concentrations obtained for each simulation along with a comparison of the predictions with available measurements.

Comparison of CMAQ Aerosol Concentrations, Dry Deposition, and Size Distributions using AE6 and the Volatility Basis Set

Matthew Woody, Saravanan Arunachalam, Francis S. Binkowski

Initial tests of CMAQ v5 using the volatility basis set (VBS) to model organic aerosols with the SAPRC-07 chemical mechanism were performed and results compared against CMAQ with ae6. In general, VBS predicted lower concentrations of all species including primary species (75% decrease for primary organics, 42.5-81.5% decrease for biogenic SOA, and 9-17.5% decrease for non-organics), with the exception of anthropogenic SOA (118.7-443.3% increase). The reduction in concentrations of non-organic PM species was attributable to an increase in dry deposition rates as VBS altered the size distribution (geometric mean diameter and standard deviation) of the bulk aerosol compared to ae6. Specifically, VBS decreased the average geometric mean diameter of the accumulation mode from 0.126 µm to 0.096 µm while increasing the standard deviation from 2.01 to 2.26.  Results comparing aerosol concentrations, dry deposition, and bulk aerosol size distributions for ae6 and VBS for a case study over the Eastern U.S. in 2002 will be presented.

An Updated Framework to Model Organic Aerosol Formation from Aircraft Emissions in CMAQ

Matthew Woody, Saravanan Arunachalam, J. Jason West, Hsi-Wu Wong

As an integral part of daily activities on a global scale, aviation is also a source of several air pollutants, including PM_2.5. Modeling efforts using CMAQ to quantify one component of PM_2.5 formed from aviation activities, secondary organic aerosols (SOA), have generally underpredicted contributions when compared against smog chamber studies of aircraft exhaust emissions. Additionally, modeled SOA predictions from aircraft emissions in CMAQ can vary significantly with grid size, with instances of aircraft emissions reducing SOA concentrations at coarser grid resolutions (12k/36k and despite the presence of SOA precursors in aircraft emissions) while increasing SOA at a finer grid resolution (4k). This work attempts to improve the modeling framework to predict organic aerosols formed from aircraft emissions in CMAQ v5, through the integration of three components, for a case study in the Hartsfield-Jackson international airport in Atlanta. First, we use the Aerosol Dynamics Simulation Code (ADSC), a 1-D plume scale model that estimates engine specific PM and vapor phase semi-volatile PM components at ambient conditions and in the immediate vicinity of an aircraft engine. Second, ADSC outputs will be interfaced with the plume-in-grid model CMAQ-APT to provide for the sub-grid scale treatment of aircraft emissions. The plume-in-grid treatment will help to reduce spatial uncertainties, providing an intermediary between the near-field treatment of ADSC and the underlying CMAQ grid, allowing for aircraft plumes to evolve at sub-grid scales prior to being diluted into the grid. The third component includes an aircraft-specific SOA parameterization using the volatility basis set to include the representation of SOA formed from intermediate and semi-volatile organic carbon, with ADSC providing individual aircraft engine-specific SOA precursor emission estimates.

A Reduced-Form Model to Estimate Near-Road Air Quality for Communities: The Screening Tool for Roadway Emissions and Exposure to Toxics (STREET)

A reduced-form model was developed to estimate air pollutant concentrations within 500 meters of busy roadways. The Screening Tool for Roadway Emissions and Exposure to Toxics (STREET) uses parameterized lookup-tables to estimate vehicular emissions; an integrated Gaussian plume formulation to estimate dispersion; and accesses nationally-available datasets for traffic and meteorological conditions. The model is considered “reduced-form” or “screening” because it foregoes the extensive array of inputs required by related, full-form emissions and dispersion models, and presents results for representative hours as opposed to a continuous time series. It uses parameterization schemes and draws inputs from national datasets to make it applicable to community-scale areas (order 100 square kilometers) across the country. STREET includes a GIS-based system to visualize results, and a user-friendly graphical user interface (GUI) to facilitate model applications. Because of its many components, STREET is considered more of a modeling system than a single model. STREET outputs include pollutant concentrations based on conditions related to traffic types, counts, and speed, and meteorological conditions, and so is well-suited to conduct what-if scenarios based on different input parameters (e.g., current conditions versus future projections). STREET estimates compare well with measured concentrations and results from other models that require more extensive inputs. It is designed to support and inform community groups and near-road researchers exploring potential health and environmental impacts related to traffic. It does not fulfill regulatory requirements related to permits or national ambient air quality standards.

Characterizing variability in exposures to traffic-related air pollutants for NEXUS: Comparison of exposure estimation approaches

Janet Burke¹, Kathie Dionisio¹, Vlad Isakov¹, Michelle Snyder¹, Michael Breen¹, Sarah Bereznicki¹, Gary Norris¹, Alan Vette¹, Stuart Batterman² and CAAA³

¹ National Exposure Research Laboratory, Office of Research and Development, U.S. EPA, Research Triangle Park, NC

²Environmental Health Sciences, School of Public Health, University of Michigan, Ann Arbor, MI

³Community Action Against Asthma

The Near-road EXposures and effects of Urban air pollutants Study (NEXUS) was designed to examine the relationship between exposures to traffic-related air pollutants and respiratory outcomes in a cohort of asthmatic children living near major roadways in Detroit, MI. A tiered exposure assessment approach was applied in NEXUS to compare increasingly complex methods for estimating traffic-related air pollutant exposures and their associations with observed health effects. The exposure assessment approaches included: (1) ambient pollutant concentrations from existing monitoring sites and two seasonal monitoring intensives; (2) GIS-based exposure indicators that incorporated data on proximity to major roadways and traffic volumes; (3) spatially and temporally refined air quality model estimates of pollutant concentrations for each residence and school location; and (4) exposure model estimates that account for residential factors that influence pollutant infiltration and for time spent in different locations (home, school, near roadways).

This tiered approach produced exposure estimates of varying spatial and temporal detail that were compared to investigate whether the increased modeling complexity (and the resultant input data requirements) of the higher tier approaches provided substantial improvement in both characterizing exposure variability, and in the epidemiology analyses with health outcome data. These results should be useful for improving exposure assessments in future air pollution epidemiology studies, by assessing the level of complexity and spatiotemporal resolution required in the exposure estimates for similar study designs.

Techniques Using HDDM Sensitivity to Estimate Future Ozone Frequency Distributions from Reduced Emissions

Chris Emery, Uarporn Nopmongcol, Tan Sakulyanontvittaya, Jaegun Jung, Justin Zagunis, Greg Yarwood

ENVIRON International Corporation

The US EPA is conducting a Risk and Exposure Assessment (REA) for the current review of the ozone National Ambient Air Quality Standard (NAAQS). Photochemical model simulations employing the High Order Decoupled Direct Method (HDDM) are being used to estimate ozone frequency distributions resulting from US anthropogenic emissions reductions that just attain alternative proposed standards in several US cities (Simon et al., 2013; Yarwood et al., 2013). The projected ozone frequency distributions will be input to risk models to estimate health and mortality effects for alternative standards. We have developed a method for estimating ozone frequency distributions in years beyond the simulation year (2006 in this case) that combines HDDM results with ambient data using regressed relationships, similar to Simon et al. (2013). Our method differs because of details in the HDDM approach. The techniques, results, and limitations stemming from our implementation of the regression approach will be presented. Calculations from the regression method will be compared to brute force results and to the original "model only" HDDM projections described by Yarwood et al. (2013). An a priori assumption is that HDDM sensitivities derived for one year (2006) can be combined with ambient data from other years. Evaluation reveals that our regression approach reduces hourly variability in ozone time series resulting in narrower ozone frequency distributions than the "model only" distributions.

Air Quality Modeling in Support of the Near-road EXposures and effects of Urban air pollutants Study (NEXUS)

Vlad Isakov, Michelle Snyder, Janet Burke, Kathie Dionisio, David Heist, Steven Perry, Sarav Arunachalam, Stuart Batterman, and Community Action Against Asthma

Population-based epidemiological studies of air pollution have traditionally relied upon imperfect surrogates of personal exposures, such as area-wide ambient air pollution levels based on readily available concentrations from central monitoring sites. U.S. EPA in collaboration with University of Michigan is developing and evaluating several types or tiers of exposure metrics for traffic-related and regional pollutants that differ in their modeling approaches for addressing the spatial and temporal heterogeneity of pollutant concentrations. We hypothesize that using more refined exposure estimates will provide greater power to detect associations with health outcomes, particularly for traffic-related pollutants that can vary considerably over short distances and time scales. The Near-road Exposures to Urban air pollutant Study (NEXUS) design includes determining if children in Detroit, MI with asthma living in close proximity to major roadways have greater health impacts associated with air pollutants than those living farther away, particularly for children living near roadways with high diesel traffic. One tier for estimating exposures to traffic-generated pollutants uses local-scale dispersion modeling. Temporally and spatially-resolved pollutant concentrations, associated with local variations of emissions and meteorology, were estimated using a combination of AERMOD and RLINE dispersion models, local emission source information from the National Emissions Inventory, detailed road network locations and traffic activity, and meteorological data from the Detroit City Airport. Hourly pollutant concentrations for CO, NO_x, PM_2.5 and its components (EC and OC) were predicted at each study participant location. The exposure metrics were evaluated in their ability to characterize the spatial and temporal variations of multiple ambient air pollutants across the study area. This research will be used for improving exposure assessments in future air pollution epidemiology studies, and for informing future multipollutant exposure analyses

Spatiotemporal Comparison of Novel Hybrid Source Apportionment and Receptor Modeling Results

CESUNICA IVEY¹, Heather Holmes¹, Yongtao Hu¹, Sivaraman Balachandran¹, Xinxin Zhai¹, Jeremiah Redman¹, Kyle Digby¹, James Mulholland¹, Armistead Russell¹

¹ School of Civil and Environmental Engineering, Georgia Institute of Technology, Atlanta, GA

 A hybrid source apportionment (SA) model was developed that takes into account source-oriented and receptor model results to generate improved spatially resolved source impact estimates [1, 2].  The model also considers uncertainties in emissions inputs, chemical transport model outputs, and observed data.  This work extends the hybrid model both spatially and temporally to generate daily source impacts over the continental US for 33 unique source categories over the month of January 2004.   Spatially and temporally dense source impact estimates can be used in studies that require spatial air quality and source impact fields (i.e. epidemiology studies that correlate daily health outcomes and air quality metrics [3-5]).  Data used to produce results include 36-km resolution CMAQ-DDM sensitivities and species concentrations from the Chemical Speciation Network (CSN, U.S. EPA). CMAQ-DDM apportions pollutant impacts to sources and accounts for the formation secondary species (e.g. secondary organic aerosols from biogenic emissions and ammonium from livestock emissions), which is typically not resolved using traditional receptor modeling approaches.  Results from the spatiotemporal extension of the hybrid SA model were evaluated by comparing results from independent SA studies that utilized traditional approaches [6, 7].  Hybrid results were aggregated from 33 to 13 source categories to compare with receptor molding SA estimates.  This work will present the model evaluation results for the spatial-temporal extensions.  Original CMAQ results are also included to highlight the improvements made by the application of the spatiotemporal extensions of the hybrid model.  

1. Hu, Y.T., et al., Fine Particulate Matter Source Apportionment using a Hybrid Chemical Transport and Receptor Model Approach, 2013.

2. Ivey, C., et al., Development of Spatial Source Impact Fields Using a Hybrid Source Apportionment Air Quality Model, 2013.

3. Ito, K., et al., Fine Particulate Matter Constituents Associated with Cardiovascular Hospitalizations and Mortality in New York City. Environmental Health Perspectives, 2011. 119(4): p. 467-473.

4. Jerrett, M., et al., Spatial analysis of air pollution and mortality in Los Angeles. Epidemiology, 2005. 16(6): p. 727-736.

5. Darrow, L.A., et al., Ambient Air Pollution and Birth Weight in Full-Term Infants in Atlanta, 1994-2004. Environmental Health Perspectives, 2011. 119(5): p. 731-737.

6. Balachandran, S., et al., Ensemble-trained source apportionment of fine particulate matter and method uncertainty analysis. Atmospheric Environment, 2012. 61: p. 387-394.

7. Maier, M.L., et al., Application of an Ensemble-Trained Source Apportionment Approach at a Site Impacted by Multiple Point Sources. Environmental Science & Technology, 2013. 47(8): p. 3743-3751.

Fine-scale comparison of GOME-2, OMI and CMAQ NO2 columns over Southern California in 2008

Hyun Cheol Kim^1,2, Sang-Mi Lee ³, Fong Ngan^1,2, and Pius Lee¹

¹ NOAA/Air Resources Laboratory, College Park, MD

² UMD/Cooperative Institute for Climate and Satellites, College Park, MD

³ South Coast Air Quality Management District, Diamond Bar, CA

In this study, NO₂ columns from the Global Ozone Monitoring Experiment-2 (GOME-2), the Ozone Monitoring Instrument (OMI) and the Community Multiscale Air Quality (CMAQ) model are compared. We developed spatial regridding algorithms for the purpose and applied them over a case in southern California in 2008. We noticed that traditional spatial regridding methods applied to the space-based observation tend to cause serious underestimation for retrieved data with coarse resolution. We designed two spatial regridding algorithms: a so-called "Conservative remapping"; and "Downscaling" algorithm to reconstruct coarse resolution satellite data into fine scale using spatial weighting kernels from fine resolution CMAQ simulations. The new method successfully produces fine scale satellite-derived NO₂ columns, showing excellent agreement with CMAQ NO₂ columns (R=0.96 for GOME-2 and R=0.93 for OMI in August 2008) and with surface NO₂ concentrations from U.S. Environmental Protection Agency (EPA) Air Quality System (AQS) monitoring sites (R=0.91 for GOME-2 and R=0.80 for OMI in August 2008). Nonetheless, traditional methods show significant negative biases in highly urbanized regions or near active traffic activities. This study has several important implications for the usability of satellite-based NO₂ column measurements: Without proper spatial regridding method, the bias can be more than 100% in highly urbanized regions; Reconstructed satellite NO₂ columns can be used to detect incomplete emission inventory, and to monitor yearly changes of emission inventory; It can generate fine scale surface NO₂ concentration fields for epidemiology and urbanization studies.

Diurnal Influences of the Urban Heat Island Effect on Ozone Mortality: An Adjoint Case Study

Amanda Pappin, Shunliu Zhao, Morteza Mesbah, Amir Hakami

Changes in land use as a result of urbanization have long been associated with elevated urban temperatures when compared to surrounding rural environments; a phenomenon termed the urban heat island (UHI) effect. More recent studies in urban areas have quantified hour-by-hour diurnal changes in UHIs, noting that the majority of urban heating occurs at night and into the early hours of the morning. Observational and model-based studies have found that temperature, of all meteorological variables, most affects ground-level ozone production. UHIs thus play an important role in photochemical production of ozone and associated adverse health effects in populations.

Our past adjoint sensitivity analysis study attributed ozone-related mortality to rising temperatures in major cities in the U.S. (e.g., 3 excess deaths during the ozone season attributed to a 1 K rise in temperature in New York, NY). We extend our study by including the time-of-day temperature influences of UHIs on short-term ozone mortality in North America. Our adjoint sensitivities thus indicate the human health consequences of city-specific, hour-by-hour UHI effects. Furthermore, given the diurnal profile of UHIs and the tendency of urban warming to be greatest during periods of low sunlight, ozone mortality quantified over longer averaging periods is likely most influenced by UHIs. We thus examine the role of choice of averaging period in estimating responses of ozone mortality metrics to diurnal changes in UHIs.

Generating spatially-resolved mobile source impacts using an observation-CMAQ data fusion technique and an emission-based indicator source apportionment approach

XINXIN ZHAI¹, Sheila A Sororian¹, Mariel Friberg¹, Heather Holmes¹, Yongtao Hu¹, James Mulholland¹, Armistead Russell¹

¹ School of Civil and Environmental Engineering, Georgia Institute of Technology, Atlanta, GA

The relationship between geo-coded patient information and ambient air quality is being investigated over the state of Georgia in a spatially resolved health study. In Atlanta, epidemiologic results using central monitor data suggest associations of acute health effects and mobile source emissions. Here, we develop a procedure to provide spatially resolved daily mobile source PM_2.5 impact estimates using data fusion that combines observations and chemical transport model (CMAQ, 4 km resolution) predictions to yield daily concentration fields of single pollutants for 2010. Daily PM_2.5 elemental carbon (24-hr average), CO (1-hr maximum), and NO_x (1-hr maximum) concentration fields were generated for Georgia using ambient monitor data and CMAQ predictions; data withholding was used to evaluate the performance of the data fusion method. The fields were then used as input to an emission-based integrated mobile source indicator method (IMSI) developed by Pachon et al. (2012) ^[1] to estimate mobile source impact fields at a 4 km resolution for Georgia. The IMSI uses EC, CO, and NO_x concentrations and the ratio of mobile emissions to total emissions for each species to estimate mobile source impacts.  The IMSI is applied to the Georgia domain by using emissions modeling results based on data from the National Emissions Inventory (NEI). The results are scaled using mobile source impact estimated by receptor model CMB applied at nine chemical speciation network sites in Georgia. The spatially-resolved daily source impact estimates across Georgia allow for the spatial analysis of health data to assess pollutant exposure risks of susceptible and vulnerable populations. Quantifying differences in source impacts estimated using different methods provides enhanced estimates of uncertainty since source impacts cannot be measured directly.

1. Pachon, et al., Development of outcome-based, multipollutant mobile source indicators, 2012.

 

Evaluation of Cumulus Cloud - Radiation Interaction Effects on Air Quality-Relevant Meteorological Variables from WRF from a Regional Climate Perspective

Jerold Herwehe, Kiran Alapaty, Tanya Otte, Chris Nolte, and Russ Bullock

U.S. EPA/ORD/NERL/Atmospheric Modeling and Analysis Division

Research Triangle Park, NC

Aware only of the resolved, grid-scale clouds, the Weather Research & Forecasting model (WRF) does not consider the interactions between subgrid-scale convective clouds and radiation. One consequence of this omission may be WRF's overestimation of surface precipitation during summer. To address this problem, our regional climate modeling group at the U.S. EPA modified WRF to provide feedbacks from the Kain-Fritsch (KF) convection parameterization to the Rapid Radiative Transfer Model - Global (YYTMG) radiation schemes in order to allow the subgrid cumulus clouds, along with the resolved clouds, to affect both shortwave and longwave radiative processes (Alapaty et al., Geophys. Res. Lett., 2012). We implemented this cumulus cloud - radiation connection in a recent release of WRF (Version 3.5) and used it to simulate a multiyear period over the contiguous U.S. The purpose of this study is to perform an initial evaluation of the effects on air quality-relevant meteorological variables when including the cumulus cloud - radiation interactions in regional climate simulations. The focus of this evaluation will be on temporally- and regionally-averaged analyses of parameters such as 2-meter temperature (for biogenic emissions, reaction rates), 10-meter wind speed and planetary boundary layer height (to gauge ventilation), specific humidity (important in particulate matter processes), cloud fraction (which affects photolysis and aqueous chemistry), precipitation (a sink for PM), and, possibly other quantities, such as the frequency of frontal passages, for example. For users of the Community Multiscale Air Quality modeling system (CMAQ), this study should provide some indication of the changes and potential improvements expected in WRF-driven CMAQ air quality predictions when including subgrid-scale cloud effects on radiation.

Sensitivity of ozone and its prediction to air mass classification

Fong (Fantine) Ngan^1,2, HyunCheol Kim^1,2, Zhaoyang Chen³ and Pius Lee¹

¹National Oceanic and Atmospheric Administration/Air Resources Laboratory, College Park, MD

²Cooperative Institute for Climate and Satellites, University of Maryland, College Park, MD

³The Department of Statistics, George Washington University, DC

Meteorological conditions determine the transport, chemical reaction and removal processes of pollutants. For the numerical modeling of regional air quality, one of the primary concerns has been that model biases of predicting air pollutants concentrations were caused by inaccuracies of meteorological inputs. By utilizing the Spatial Synoptic Classification (SSC), a commonly used weather typing method, we investigate the sensitivity of surface ozone and its prediction in different weather patterns. The analysis will be conducted during the ozone season (May to September) in years 2009 to 2011 for seven regions in contiguous US, including California (CA), Michigan Lake (MI), Ohio River Valley (OV), Northeastern (NE), Southeastern (SE), Texas (TX) and Central Mountain (CM). The ozone predictions are from NOAA National Air Quality Forecast Capability while the ozone observations are from US EPA air quality system. Through the analysis, we observed that Moist Tropical and Dry Tropical are the most important air mass types conducive to high surface ozone episodes in occurrence frequency and production efficiency. However, coincidentally the Moist Tropical also experienced the highest biases in modeled surface ozone among all other weather types underscoring the large uncertainty in ozone predictions.

Annual WRF Simulations for the Utah Bureau of Land Managements Air Resource Management Strategy (ARMS) Air Quality Modeling

Kenneth Craig¹, Zion Wang², Stephen Reid¹, Courtney Taylor², Jason Reed²

¹Sonoma Technology, Inc., Petaluma, CA
²AECOM, Fort Collins, CO

An annual Weather Research and Forecasting (WRF) model simulation was performed for use as part of an air quality management tool to assess the potential air impacts from future activities on land administered by the Bureau of Land Management in Utah's Uinta Basin. Cold pool stagnation events are common in the Uinta Basin during the winter months, but can be challenging for the WRF model to simulate correctly. The combination of snow cover, light winds, strong and shallow inversions, and weak mixing commonly observed during these cold pool stagnation events creates conditions that can lead to elevated particulate matter and ozone concentrations.

Because the air quality management tool will be used by stakeholders to assess air quality impacts during all seasons, adequate meteorological model performance must be demonstrated not just during cold pool stagnation events, but throughout the year under a wide variety of meteorological conditions. Multiple WRF configurations were tested to determine the preferred configuration for the annual simulation; different WRF configurations were found to perform better depending on the season evaluated. The results from these tests suggested that two different WRF configurations should be used during the annual simulation: one for winter months and another for non-winter months.

Unusually poor model performance for temperature within the Uinta Basin and throughout the innermountain west led to the discovery of an unintended consequence of applying buddy check quality assessment tests to observational data in complex terrain. The buddy check algorithm within the WRF pre-processor, OBSGRID, compares observations to other neighboring observations and rejects those observations that differ substantially from neighboring ones. In complex terrain, temperature readings from observations separated by just a few kilometers in horizontal distance can differ dramatically due to large differences in elevation. Although the observed temperature differences may be valid, the buddy check algorithm rejects the observations. For the Uinta Basin, a large portion of valid temperature observations were omitted from the WRF data assimilation process. Therefore, we deactivated the OBSGRID buddy check test but retained the other OBSGRID quality assurance tests. This adjustment increased the impact of the temperature data assimilation and improved model performance in the annual WRF simulation without reducing the quality of the assimilated observational data. This underscores the importance of examining not only the quality of assimilated observation data, but also any automated quality control processes, as part of the overall model quality assurance procedure.

Retrospective Analysis of Meteorological Effects on Photochemical Modeling in North Carolina

Elliot Tardif, Nicholas Witcraft and Bradley McLamb

North Carolina Division of Air Quality

Average concentrations of ozone have decreased steadily and significantly across much of the nation in recent decades, thanks to rules and regulations that have markedly curtailed emissions that cause ozone.  However, meteorological impacts can either diminish or exaggerate these effects when analyzing annual 4^th-highest ozone values and other statistical benchmarks.  This project attempted to analyze these meteorological effects over the course of eleven ozone seasons, leaving emissions constant during this time to isolate the effects of meteorology on ozone concentrations.  Meteorology data was produced using WRF v.3.4.1, initialized using NAYY data, and converted for use in CMAQ using MCIP v.3.6.  Emissions data was taken from the USEPA's year 2005 MATS modeling and kept constant across all studied years.  The output of the CMAQ runs were then analyzed against observed ozone readings within North Carolina and nearby large cities during this time.  Our hypothesis was to expect generally higher-than-observed concentrations of ozone in the most recent years due to modeled emissions that were likely higher than what were otherwise observed, especially during periods of excessive heat that appeared to be more frequent in the past few years.

GOES data assimilation in the WRF model: Development and Evaluation

Aijun Xiu, Limei Ran, Adel Hanna

Institute for the Environment, University of North Carolina at Chapel Hill

Arastoo Biazar, Richard McNider, Kevin Doty

National Space Science Technology Center, University of Alabama at Huntsville

Recent advances in tropospheric remote sensing provide an ample opportunity to enhance and validate meteorology and air quality models at various temporal and spatial scales. One of the satellite systems is the Geostationary Operational Environmental Satellites (GOES) operated by NOAA that have been used for monitoring weather and assisting weather prediction, as well as for research studies. Right now the CMAQ air quality model has an option to use the GOES cloud optical depth data to adjust the photolysis rates that are critical for good performance in CMAQ for ozone and secondary PM as these rates drive photochemical production of many oxidants in cloudy air. In order to improve the meteorological simulations by the WRF model in support of air quality modeling applications, we have developed the regridding capability in the Spatial Allocator tool and the GOES data assimilation in the WRF model.

We will present the development of the GOES regridding tool and data assimilation scheme in the WRF model. The WRF simulations will be evaluated and analyzed comparing to observations.

Sensitivity to changes in HONO emissions from mobile sources simulated for Houston area

Beata Czader, Yunsoo Choi, Lijun Diao

University of Houston, Department of Earth and Atmospheric Sciences

Nitrous acid (HONO) is an important source of hydroxyl radical (OH), which plays a crucial role in oxidation of volatile organic compounds (VOCs) leading to the formation of ozone. Accurate estimation of HONO in air quality modeling is important as it affects predictions of HO_x as well as ozone concentrations.

Current mobile-emission model, MOVES, estimates HONO emissions based on the HONO/NO_x ratio derived from the tunnel studies done in 2001. However, recent measurement studies in Houston suggest that this ratio is higher. We propose to use the latest HONO/NO_x ratio in estimating emissions of HONO from mobile sources and to perform air quality simulations with the CMAQ model to evaluate the effect of changing mobile emissions on HONO predictions as well as on O₃ and HO_x concentrations.

Assessment of the two-way Coupled WRF-CMAQ Model with Observations from the CARES

Chuen-Meei Gan, Francis Binkowski, Jia Xing, Robert Gilliam, David Wong, Jonathan Pleim, Rohit Mathur, Kirk Baker and James Kelly

The main goal of this assessment is to evaluate the improved aerosol component of two-way coupled WRF-CMAQ model particularly in representing aerosol physical and optical properties by utilizing observations from the Carbonaceous Aerosol and Radiative Effects Study (CARES) in May 2010 which was held in central California. The objective of the CARES was to investigate the evolution of carbonaceous aerosols of different types and their optical and hygroscopic properties. Since various instruments (e.g. AMS, PSAP and FIMS) were deployed aboard two aircrafts (DOE G-1 and NASA B-200) during the field campaign, the available in situ measurements provide an opportunity to explore the aerosol radiative effects in a detailed manner. For example, measurements of PM size, species and optical properties are made not only at the surface but also in the vertical - these can be utilized to assess the simulation output with a single column model. In addition, satellite data (e.g. SeaWiFS) will be used for regional assessment. The new two-way coupled model has been updated with several modifications such as densities and refractive indices for different particulate matter species based on OPAC dataset and Mie and Core-Shell scattering approaches. Two months (May and June 2010) simulations (feedback and no feedback) at 4-km horizontal resolution are conducted. Detailed comparisons of various meteorological (e.g., PBL heights, radiation fields), chemical (PM composition and relevant gas-phase precursors), and optical (extinction, AOD) characteristics simulated by the model with corresponding measurements from the field campaign will be discussed for simulations involving both aerosol feedback and no-feedback.

ERTAC Electric Generating Unit Emissions Projection Tool - Methodology and Results

Julie McDill PE, MARAMA

The Eastern Regional Technical Advisory Committee (ERTAC) has developed a tool to estimate future year NOX and SO2 emissions from Electric Generating Units (EGU) based on measured hourly Continuous Emission Monitoring (CEM) data for use in regional chemical transport modeling to assess air quality impacts on both an annual and episodic peak basis. Growth considers fuel-specific generation trends and constraints. Unit operations are not grown past installed capacity limits. Regional operational reserve requirements are evaluated. The tool combines information from a number of sources including: (1) Energy Information Administration (EIA) Annual Energy Output (AEO) fuel specific average growth projections, (2) North American Electric Reliability Corporation (NERC) peak growth projections; (3) base year Clean Air Markets Division (CAMD) hourly emission files and (4) state and source provided information on planned expansions, curtailments and controls. The tool consists of a preprocessor, processor and post-processors. The preprocessor sets up the base year data structure and identifies data outliers. The processor calculates future generation using an iterative approach to adjust for individual unit overload or system capacity deficit. The post-processors summarize the large hourly output files to aid in data review. The methodology is coded using Python and SQLite. Details of the tool operation and results from a projection from 2007 to 2017 will be presented.

Particle phase liquid water measurements during the Southern Oxidant and Aerosol Study

Thien Khoi V. Nguyen¹, Markus D. Petters², Annmarie G. Carlton¹, Sarah R. Suda²

¹Department of Environmental Sciences, Rutgers University, New Brunswick NJ 08901, USA.

²Department of Marine Earth and Atmospheric Sciences, North Carolina State University, Raleigh, NC 27695, USA.

Particle-phase liquid water (H₂O_ptcl) contributes to total aerosol mass concentrations. Previous studies established links between inorganic species, particle hygroscopicity, ambient relative humidity, and condensed phase liquid water. These relationships are also included in thermodynamic modules of atmospheric chemistry models. Conversely, relationships between H₂O_ptcl and organic species are poorly understood, and there are few field measurements linking the two. Furthermore, the potential of H₂O_ptcl to facilitate secondary organic aerosol (SOA) formation, which contributes to the atmospheric fine particle burden, is well established in laboratory studies of aqueous phase organic chemistry, but the extent to which this occurs in the atmosphere is unknown. In this study, we conducted in situ measurements of H₂O_ptcl using a newly developed technique - the semi-volatile differential mobility analysis (SVDMA). Measurements were conducted June 1 - July 15, 2013, during the Southern Oxidant and Aerosol Study (SOAS) in the southeast U.S., a biogenically-dominated environment impacted by anthropogenic pollution and known to contain high concentrations of organic aerosol mass. The SVDMA measures volume distributions of ambient atmospheric aerosols in three states: unperturbed, dried, and dried then re-humidified. Unperturbed measurements characterize the aerosol distribution at ambient conditions. For dry spectra, the sample is routed through a cold trap (T = -30K) upstream of the DMA inlet. The total volume of water and SVOCs lost during drying are quantified by differencing dry and unperturbed volumes from the integrated size spectra. SVOC volumes are quantified by re-humidifying the sample and referencing to the unperturbed state. Results indicate that liquid water is an important contributor to ambient aerosol volume in the SE US during early morning times when relative humidities are highest. Measured H₂O_ptcl volumes show that the aerosol can be characterized by ~ 0.2 to 0.3, which is consistent with a mix of hygroscopic organic and inorganic compounds. The data also demonstrate that the presence of water can mediate the partitioning of semi-volatile organic compounds to the condensed phase, suggesting that particle drying may lead to underestimates of organic aerosol mass concentrations. A better understanding of the connections between biogenic emissions and SOA formed in the presence of anthropogenic perturbations will improve the chemical mechanisms in atmospheric photochemical models, yielding more informed and more accurate model predictions for climate and air quality.

Inter-comparison of Emission Projection Methods for NOx and SO2 Emissions From Electricity Generating Units

Byeong-Uk Kim¹ and Doris McLeod²

¹Georgia Department of Natural Resources (GA DNR)

²Virginia Department of Environmental Quality (VA DEQ)

Estimating NOx and SO2 emissions from electricity generating units (EGUs) in future years is a key component of technical analyses in US air quality management. Three approaches were reviewed: (1) The Southeastern Modeling, Analysis, and Planning (SEMAP) approach, (2) The Eastern Regional Technical Advisory Committee (ERTAC) approach, and (3) the Integrated Planning Model (IPM) approach. The SEMAP approach is to apply growth/control factors to the annual emission inventory. This is a simple and fast approach but does not account for the impact of new units or shutting down old units explicitly. The ERTAC approach is to consider electricity generation demand among EGUs in the same fuel type explicitly. This is done by taking supply and demand of hourly generation into account when hourly heat input values are computed. NOx and SO2 emissions are, in turn, estimated with hourly heat input data and emission rates reflecting control information specific to each EGU. The IPM approach is to utilize simulation outcome of complex interactions among all energy sectors. This paper compares methodological differences as well as SO2 and NOx emission estimated by each approach.

The January-February 2013 DISCOVER-AQ Field Campaign in theSan Joaquin Valley of California: Observations and Initial CMAQ Simulations

Kenneth Pickering, Christopher Loughner, James Crawford, and the DISCOVER-AQ Observation Team

The second deployment of the NASA Earth Venture - 1 DISCOVER-AQ (Deriving Information on Surface Conditions from Column and Vertically Resolved Observations Relevant to Air Quality) field program took place in January and February 2013 in the San Joaquin Valley of California. This region frequently experiences large particulate matter concentrations in the wintertime. Over 170 lower tropospheric profiles of aerosol types and optical and microphysical properties, as well as trace gases (e.g., O3, NO, NO2, HCHO, CO, HNO3, and selected VOCs), were observed by the P-3B aircraft over a set of existing air quality monitoring stations, which were upgraded to include Aeronet sun photometers and Pandora UV/Vis spectrometers. In addition, over 150 missed approaches were conducted by the P-3B at a set of airports in the region to obtain observations as close to the surface as possible. A King Air aircraft conducted remote sensing with the High Spectral Resolution Lidar (HSRL) for aerosols and the Airborne Compact Atmospheric Mapper (ACAM) instrument for trace gases. Ground-based lidar, tethered balloon, and ozonesonde observations were performed in the valley. The deployment period was simulated with the CMAQ v5.0 model and evaluated using data collected by the NASA P-3B aircraft. The evaluation was conducted using output from a 4-km resolution domain and focused on the ability of the model to simulate regional aerosol concentrations.

Key Non-EGU Projections Issues for the 2011 Emissions Modeling Platform

Rich Mason, Alison Eyth, Alexis Zubrow

The U.S. Environmental Protection Agency (EPA)'s Office of Air Quality Planning and Standards (OAQPS) has developed a 2011 emissions modeling platform, including projections for a range of future years. The emissions for this platform are based on Version 1 of the 2011 National Emissions Inventory (NEI) and future year emission projections for EGUs, onroad mobile and most nonroad mobile sources are provided by the IPM, SMOKE-MOVES system and NONROAD models, respectively. There is very little difference between base and future year emissions for the remaining, primarily non-EGU, sources; this is primarily a result of the limited collection of known controls and/or growth manipulations needed for future year estimates. We demonstrate that these non-EGU sources comprise a larger fraction of total anthropogenic PM_2.5 and ozone precursor emissions that previous versions of the NEI and going forward, will be an even larger component in the future thanks to existing EGU and mobile source measures. We describe some of the more significant non-EGU sources, such as Residential Wood Combustion and Oil and Gas Production, with respect to the limited existing set of projection methods as well as efforts to improve these projection methods in the near future.

Evaluating fine-scale photochemical modeling for California during May-June 2010

James Kelly, Kirk Baker, and Chris Misenis

EPA, Office of Air Quality Planning and Standards, RTP, NC

Rob Gilliam, Jon Pleim, Golam Sarwar, and Donna Schwede

EPA, Office of Research and Development, RTP, NC

Multiple areas in California are designated as nonattainment of the ozone and PM_2.5 National Ambient Air Quality Standards (NAAQS). Air quality modeling for California is an important aspect of national scale modeling for EPA rulemaking. Fine-scale air quality modeling for key population centers in California is also useful to inform health and exposure studies. However, air quality modeling is challenging in California due to complex emissions, terrain, meteorology, and chemistry. In May-June 2010, the CalNex field study was conducted to answer scientific questions about emissions, chemistry, climate, transport, and meteorology in California. The CalNex observational dataset provides an opportunity to identify model performance issues in fine-scale simulations of California as well as potential causes and remedies. In this study, we conduct fine-scale (4-km horizontal resolution) photochemical air quality model simulations for California for the May-June 2010 period using the Community Multiscale Air Quality model. An operational model performance evaluation is conducted to identify topics for further exploration with sensitivity simulations and field study measurements. Model predictions from base and sensitivity simulations for PM2.5 components, ozone, and their precursors are then evaluated with field study data. The impacts of model updates are discussed and areas of future work are identified.

Influences of Drought on Biogenic VOC Emissions

Erin Chavez-Figueroa and Daniel Cohan

Biogenic volatile organic compound (BVOC) emissions depend upon plant species, air temperature, soil moisture, insolation, and canopy properties. At the same time, the canopy itself will change with varying climate conditions. The effect of drought on both the canopy and the biogenic emissions is of particular interest due to the predicted increase of both frequency and severity of drought in much of the United States under future climate scenarios. Influence of drought on canopy condition was assessed by comparing interannual changes in satellite-derived LAI to Palmer Drought Severity Index (PDSI), temperature, and precipitation. While the heavily forested areas that produce the most isoprene are not particularly sensitive to drought, grasslands show a strong correlation between LAI and drought.

Sensitivity of biogenic emissions to drought was assessed using two techniques. In the first, isoprene measurements from Photochemical Assessment Monitoring Stations (PAMS) was regressed with temperature, precipitation, PDSI, and LAI. Temperature was found to be the most important factor for most locations, with effects from precipitation, PDSI, and LAI showing little statistical or physical significance. In the second assessment, the biogenic emissions model MEGAN was used to find the sensitivity of isoprene and monoterpene emissions to LAI, sunlight, temperature, and soil moisture. An ensemble of model runs were performed for the Continental United States using multiple data sources of differing variability for each variable. Two years were used, 2005 and 2007. In 2005, much of the US had higher than normal rain, particularly in the heavy isoprene producing region of the Southeast US, while 2007 brought extreme drought to the same region. Preliminary results indicate that interannual variability in LAI has marginal effect on biogenic emissions, but that the average LAI value chosen has a large impact on the emissions prediction. The results of the ensemble modeling will also be compared to the PAMS measurements, allowing exploration of how the modeled relationships compare to observations in order to resolve some of the expected discrepancy between models and observations caused by exclusion of drought impacts.

HCHO and NO2 column comparisons between OMI, GOME-2 and CMAQ during 2013 SENEX campaign

Hyun Cheol Kim ^1,2, Li Pan ^1,2, Pius Lee ¹, Rick Saylor ³, and Daniel Tong ^1,2

¹ NOAA/Air Resources Laboratory, College Park, MD

² UMD/Cooperative Institute for Climate and Satellites, College Park, MD

³ NOAA/ARL/Atmospheric Turbulence and Diffusion Division, Oak Ridge, TN

Formaldehyde (HCHO) and nitrogen dioxide (NO₂) columns observed from space are compared with fine resolution Community Multiscale Air Quality (CMAQ) simulation and aircraft measurements during the Southeast Nexus campaign, which is designed to conduct an airborne study to investigate the roles by anthropogenic and natural emission in the formation of ozone and aerosol in Southeastern U.S. during summer of 2013. Since HCHO is a common intermediate product from the degradation of volatile organic compounds (VOCs), it can be used as a proxy to monitor emissions of VOCs from biogenic, biomass burning, and anthropogenic sources. Monitoring the spatial and temporal variability of VOC emissions in the southeast US region is crucial to understanding control processes in this region's air quality, especially for ozone and secondary organic aerosols productions. On the other hand, monitoring NO₂ column is important to investigate locations of natural and anthropogenic NO_x emission sources, and chemical regimes according to Ozone-NO_x-VOC chemistry. We utilized two HCHO and NO₂ column retrievals from two space-borne instruments: the Ozone Monitoring Instrument (OMI) and the Global Ozone Monitoring Experiment-2 (GOME-2), available from the Tropospheric Emission Monitoring Internet Service (TEMIS), the Satellite Application Facility on Ozone Monitoring (O3M SAF); and National Aeronautics and Space Administration (NASA). OMI, onboard the EOS Aura satellite has a 13:30 local overpass time with 13x24 km² resolution, and GOME-2, onboard the EUMETSAT MetOp-A satellite has a 09:30 overpass with 40x80 km² resolution. Herein we developed a downscaling spatial regridding method. It is applied to resolve biases resulted from different footprint pixel resolutions, and regridded fields are compared with predicted fields from a fine-scale (4-km) CMAQ simulation and in-situ measurements.

Estimation of Biomass Burning Emissions over Turkey using SEVIRI Fire Characterization data: the Antalya fire, August 2008

G. Baldassarre¹, L. Pozzoli¹, J.W. Kaiser², C.C. Schimdt³, A. Unal¹, T. Kindap¹

¹Eurasia Institute of Earth Sciences, Istanbul Technical University, Turkey.

²KCL, London, UK; ECMWF, Reading, UK; MPI for Chemistry, Mainz, Germany

³University of Wisconsin Cooperative Institute for Meteorological Satellite Studies- Consultant, Madison, WI, United States

From an air quality management perspective, the availability of information on forest fires and their gaseous and aerosol emissions becomes critical for specific regions and seasons. Recent improvements of air quality models, such as CMAQ, permit to simulate the chemical composition of the atmosphere at finer resolutions. Therefore emission inventories must also be provided with higher level of detail in terms of both spatial and temporal resolution. In particular, forest fire emissions, due to their episodic nature, are characterized by high spatial and temporal variations. In order to better simulate the impact of fire emissions on air quality it is fundamental to better describe the entire evolution of the fires.

SEVIRI-based Fire Radiative Power (FRP) can be directly linked to the biomass combustion and emissions. And it has sufficient time resolution (15 minutes) to observe the complete fire life cycle and thus capture fires when they reach their peak intensity. The Wildfire Automated Biomass Burning Algorithm (WF_ABBA) and the EUMETSAT Land SAF Fire Radiative Power provide operational fire radiative power products based on SEVIRI observations using different algorithms. In this presentation, we present a case study of a large forest fire occurred during August 2008 in the province of Antalya, South of Turkey, which burned an area of 4000-4500 hectares of forest land. The emission estimates of the principal pollutants from the two fire dataset based on SEVIRI can describe the entire evolution of the fire episode improving the temporal resolution. The new estimates are comparable with other available fire emission inventories, which are based on MODIS satellite observations. For example 3.2 Gg of PM2.5 and 27.1 Gg of CO are estimated for the entire Antalya episode from WF_ABBA, while the GFASv1.1 based on MODIS estimates are 2.2 Gg and 27.3 Gg for PM25 and CO, respectively. SEVIRI-based fire emission inventories, derived over a larger region of the Eastern Mediterranean Basin surrounding Turkey during Antalya fire life time, has been used as input in the Community Multiscale Air Quality (CMAQ) model, providing a more accurate description of fire contribution (hourly emissions at 10 x 10 km horizontal resolution) in determining the concentration of major contaminants in the study area and during the selected period. The variability of aerosol and gaseous species due to forest fire emissions will be quantified by comparing a set of simulations using 2 hourly emission inventories based on SEVIRI dataset and a daily emission inventory based on MODIS (GFAS).

Ammonia Measurements by the NASA Tropospheric Emission Spectrometer (TES)

K. Cady-Pereira, M. Shephard, D. K. Henze, L. Zhu, R. W. Pinder, J. O. Bash, J. T. Walker, M. Lou

The high spectral resolution and good SNR provided by the TES instrument allow for the detection and retrieval of numerous trace species. Advanced optimal estimation algorithms have been developed to retrieve three of these, ammonia, methanol and formic acid, from TES radiances. Ammonia is currently a standard TES operational product, while methanol and formic acid will be standard products in the next TES software update (V006). Ammonia is highly reactive, with concurrent high spatial and temporal variability; it can play a key role in determining air quality through its part in the formation of PM_2.5 particles, yet in situ measurements are sparse, especially over areas beyond North America and Europe. The air quality community has a pressing need for global information on the diurnal and seasonal cycles as well as the distribution and strength of the ammonia sources, thus, there is great interest in using these new satellite derived products, but there is often no clear idea on the information they provide.

We will first provide a short summary of the characteristics of TES retrieved ammonia, discuss the distinct characteristics of point and satellite measurements and illustrate how information from the latter is related to the former. We will then present results from comparisons with in situ measurements. Specifically, we will compare TES NH₃ with surface measurements in North Carolina and China and surface and aircraft measurements in the San Joaquin Valley in California. We will also compare global TES NH₃ with outcomes from the GEOS-CHEM model. We will present results from the application of inverse methods using TES ammonia to constrain model emissions, an area of research that has showcased the value provided by satellite data. Finally, we will demonstrate the potential of a sensor with TES characteristics on a geostationary platform to provide data with quality sufficient to evaluate models of the ammonia bi-directional exchange at the surface and we will show some preliminary ammonia retrievals from the Cross-track Infrared Sounder (CrIS) currently flying on the NASA NPP Suomi satellite.

Development of a Uinta Basin Oil and Gas Emissions Inventory Suitable for a Model Performance Evaluation

Courtney Taylor¹, Caitlin Shaw¹, Tiffany Samuelson¹, Erin Pollard², Stephen Reid², Leonard Herr³

¹AECOM Inc.

²Sonoma Technology, Inc.

³Bureau of Land Management, Utah State Office

The Bureau of Land Management (BLM), Utah State Office, has initiated several studies focused on air quality in the Uinta Basin; one of these studies is the Air Resource Management Strategy (ARMS) Modeling Study. The Uinta Basin is an area in northeastern Utah that is projected to have continued development of oil and gas reserves in the foreseeable future. The goal of the ARMS Modeling Study is to develop a reusable modeling platform to streamline future air quality impact analyses required under the National Environmental Policy Act. Since the modeling platform will be used to assess air quality impacts for future conditions, adequate model performance must be demonstrated. Year 2010 was selected for the model performance evaluation based on data availability, including ambient ozone measurements, emissions inventories, and meteorological measurements.

Oil and gas emissions inventories have received increased attention in multiple areas of the United States. As part of the ARMS Modeling Study, existing oil and gas emissions inventories for the Uinta Basin were revised to reflect current operations and fill known data gaps. The 2006 Western Regional Air Partnership (WRAP) Phase III Uinta Basin emissions inventory was used in conjunction with 2010 actual production information as the basis for the 2010 emissions inventory. The 2006 WRAP Phase III Uinta Basin emissions inventory does not contain emissions emitted from oil and gas mobile sources nor produced water ponds. Emissions from these activities were estimated and incorporated into the 2010 emissions inventory. In addition, oil and gas producers in the Uinta Basin participated in a survey. The results were used to assess if previous assumptions are still accurate and update assumptions when necessary, as well as apportion drilling and completion emissions to a specific time and place.

The resulting 2010 oil and gas emissions inventory was combined with other regional emissions for a comprehensive emissions inventory suitable for evaluating the performance of both CMAQ and CAMx models. As a final step, the 2010 oil and gas emissions inventory is used to estimate future oil and gas emissions in 2021 for four future year scenarios.

Developing a High-Spatial-Resolution Aerosol Optical Depth Product Using MODIS Data for Evaluating Aerosol During Large Wildfire Events

Jennifer L. DeWinter¹, Sean M. Raffuse¹, Michael C. McCarthy¹, Kenneth J. Craig¹,
Loayeh K. Jumbam², Scott Fruin³, Fred W. Lurmann¹

¹ Sonoma Technology, Inc., 1455 N. McDowell Blvd., Suite D, Petaluma, CA 94954

² Esri, 380 New York St., Redlands, CA 92373

³ Keck School of Medicine, University of Southern California, Los Angeles, CA 90089

Satellite-derived aerosol optical depth (AOD) has been used to understand spatial variations in aerosol that are not well-represented by sparse ground-based monitoring networks. AOD products have also been used for data assimilation and model evaluation by the air quality modeling community. Remotely sensed AOD can be particularly useful during wildfire events, when aerosol levels can vary widely over small spatial scales. However, the standard AOD product at 10-km spatial resolution, available from the NASA Moderate Resolution Imaging Spectroradiometer (MODIS) instrument, is too spatially coarse to adequately capture intra-urban variability or other fine-scale variations near smoke plumes during wildfires. In addition, the aerosol optical properties used for the standard NASA AOD product are not representative of aerosol from biomass burning, the cloud-masking algorithm can incorrectly label heavy smoke as clouds, and there are often missing AOD data due to failed retrievals over the bright land surfaces of the western United States.

To improve the usefulness of the AOD product during wildfires, we developed a localized AOD product covering Northern California at 2.5-km spatial resolution using raw MODIS data for the summer of 2008, a period with multiple large and lengthy wildfires in the region. The algorithm uses local biomass burning aerosol optical properties, local surface reflectance data, and a "relaxed" cloud filter. The high-resolution AOD was regressed against hourly surface-based PM2.5 concentrations observed at monitors throughout the domain; results show that the AOD explains more than 50% of the variance in hourly surface PM2.5 concentrations observed during the wildfires. The AOD-PM2.5 relationship was then used to estimate ground-level PM2.5 concentrations. We will present our methods for developing high-resolution estimates of AOD over California during the 2008 Northern California fires and will discuss implications for the air quality modeling community.

Advances in Support of the CMAQ Bidirectional Science Option for the estimation of ammonia flux from agricultural cropland.

Ellen Cooter¹, Limei Ran², Verel Benson³, Jesse Bash¹

1/ corresponding author, cooter.ellen@epa.gov, 919-541-1334, USEPA/ORD/NERL, RTP, NC

2/UNC Institute for the Environment, Univ. of North Carolina at Chapel Hill, Chapel Hill, NC

3/Benson Consulting, Columbia, MO

Quantifying regional background ozone for Houston, Texas
Shaena R. Berlin1,2, Andrew O. Langford1, Mark Estes3, Melody Dong1,4 and David D. Parrish1

1NOAA ESRL Chemical Sciences Division, 325 Broadway, Boulder, CO, USA
2Massachusetts Institute of Technology, Cambridge, MA, USA
3Texas Commission on Environmental Quality, Austin, Texas, USA
4presently at Department of Bioengineering, University of California, San Diego, USA

Regional background ozone in Houston was estimated by two independent methods for ozone seasons between 1998 and 2012. One method considered the lowest daily peak eight-hour ozone observed at the periphery of the Houston metropolitan area; the other method obtained background estimates using principal components analysis of ozone monitoring data. From these estimates of regional background ozone, trends in regional background, peak ozone, and estimated local contributions were obtained. The trends for the two methods were compared to each other and to recently-published estimates of background ozone trends for the US. In addition, the relationships between regional background ozone, wind flow patterns, and Spatial Synoptic Classifications were quantified. The ozone trends for different meteorological conditions differ substantially. Finally, the relationship between regional background ozone and peak ozone was examined to find days likely to have high local ozone formation rates.

Assessment of black carbon in the Arctic: new emission inventory of Russia, model evaluation and implications

Kan Huang¹, Joshua S. Fu^1*, Xinyi Dong¹

¹Department of Civil and Environmental Engineering, The University of Tennessee, Knoxville, TN

*Correspondence to: jsfu@utk.edu

Anthropogenic emissions of black carbon (BC) contribute to global warming, and that effect is most pronounced in the Arctic where both atmospheric heat retention and deposition to the ice promote warmer temperatures and snow melt by reducing the surface albedo. To simulate the effect of black carbon on the regional climate over the Arctic, a reliable emission inventory is required as the input for any chemical transport models. Currently, all of the Arctic region countries have relatively reliable emission inventories with the exception of Russia, due to the difficulties of quantifying the local emission factors and locating emission sources. Various trajectory models concluded that northern and central Russia was one of the major source regions contributing to the Arctic haze. However, CTM using current BC emission inventory have shown contrasting results. Models generally underpredicted BC concentrations in the Arctic [Koch and Hansen, 2005; Shindell et al., 2008], and the largest model diversity occurred in northern Eurasia and the remote Arctic from the AeroCom model intercomparison [Koch et al., 2009].

In this study, we build an up-to-date BC emission inventory of Russia by using both the "bottom-up" and "top-down" methods. Several main emission sectors are considered, including gas flaring, power generation and heating, residential, industry and transportation. Russia's federal statistics of fuel usage, emission data, and emission factors from literatures are used for a best estimate of the BC emission. Different proxy datasets are used for the spatial allocation of sectoral BC emission into a fine grid resolution. Model simulation using both Hemispheric CMAQ and GEOS-Chem with the new emission inventory is evaluated by comparison to various observational datasets, including surface BC/aerosol absorption measurements, AERONET column absorption AOD, and satellite retrievals. It is concluded that the improved emission inventory of Russia has obvious improvement for reproducing the black carbon over the Arctic region. It is assessed that the role of the black carbon emission from Russia on the Arctic has more significant impacts than previous modeling studies. This study serves as an urgent need for the communities of both regional and global models to narrow down the gap between the measured and simulated black carbon levels in the Arctic Circle.

References:

Koch, D., and J. Hansen (2005), Distant origins of Arctic black carbon: A Goddard Institute for Space Studies ModelE experiment, J. Geophys. Res., 110(D04204, doi:10.1029/2004JD005296), Doi 10.1029/2004jd005296.

Koch, D., et al. (2009), Evaluation of black carbon estimations in global aerosol models, Atmos. Chem. Phys., 9(22), 9001-9026, doi:9010.5194/acp-9009-9001-2009.

Shindell, D. T., et al. (2008), A multi-model assessment of pollution transport to the Arctic, Atmos. Chem. Phys., 8(17), 5353-5372, doi:5310.5194/acp-5358-5353-2008.

Intercomparison of secondary organic aerosol models based on SOA/Ox ratio

Yu Morino, Kiyoshi Tanabe, Kei Sato, and Toshimasa Ohara
National Institute for Environmental Studies, Japan

Improvement of secondary organic aerosol (SOA) models is critical in order to accurately understand behaviors and sources of atmospheric aerosols. Over the last decade, a number of SOA production pathways were newly found, and several new SOA models have been developed. However, comparative studies on performance of multiple SOA models are limited to date. In this study, results of five SOA models, including yield model (SAPRC99-AERO4 and SAPRC99-AERO5) , mechanistic model (CACM-MADRID2), near-explicit model (MCM), and volatility basis set model (SAPRC99-VBS), were compared. In addition, the model performance of the SOA models were evaluated by comparing with the observed ratio of SOA and odd oxygen ([Ox] = [O3] + [NO2]). It has been shown that, in Tokyo, SOA correlated well with Ox, and thus, Tokyo is an appropriate research field for this model intercomparison. All the five models showed similar results for gaseous species, including ozone, reactive nitrogen, hydroxy radical, and volatile organic species. By contrast, simulated SOA concentration largely varied among five models. VBS model well reproduced the observed SOA/Ox ratio, while other four models largely underestimated this ratio. Source contributions of SOA, which were estimated by sensitivity simulation (brute-force method), showed large discrepancies among several SOA models. Choice of SOA model is critical in the source apportionment of SOA.

Ozone Reactivity Analyses of Air Pollutant Emission Inventories

Maximum Incremental Reactivity (MIR) scales quantify the relative ground-level ozone impacts of different volatile organic compounds (VOCs). Based on box-model calculations of ozone formation under a variety of atmospheric conditions, MIR scales estimate the mass of additional ozone formed per mass of VOC emissions.In this presentation we demonstrate a technique for using the Speciation Tool to calculate total MIRs for each VOC speciation profile. We apply these profile MIRs to example emission inventories to produce ozone reactivity-based analyses of individual air pollutant emissions sources.In combination with SMOKE, the profile reactivity estimates can be used to compare and/or rank inventory sources based on ozone formation potentials.Along with comparisons of individual sources in the inventory, the reactivity assessments can be coupled with spatial and temporal information to highlight locations and time periods of emissions sources with high reactivity.

Example analyses and how the results can be used to evaluate inventories for both quality assurance and ozone impacts will be presented. Because the MIR scale is designed to represent radical-limited photochemical regimes that have high sensitivity to VOC, we also evaluate and compare the results for MIR and Maximum Ozone Incremental Reactivity (MOIR) conditions that have lower sensitivity to VOC.

Development and Status of EPAs 2011 Emissions Modeling Platform

Alison Eyth, Rich Mason, Alexis Zubrow

EPA has developed an emissions modeling platform for 2011 based on Version 1 of the 2011 National Emissions Inventory (NEI). The 2011 NEI and corresponding modeling platform have been developed at an unprecedented pace, allowing the most currently available data to be used for regulatory efforts and other modeling studies. Key enhancements to the inventory, supporting data, and methods for the modeling platform will be discussed. These include updates to oil and gas estimation techniques, modeling methods for onroad mobile sources, and improved spatial surrogates. The impact of collaboration with the states on platform development will be discussed, as will areas identified as needing continued refinement, such as the modeling and inventory of commercial marine vessel emissions.

Mesoscale meteorological modeling at kilometer scale grid meshes for air quality simulations.

Jason Ching, Rich Rotunno, Peggy Lemone, Branko Kosovich, Jimy Dudhia, Pedro Jimenez , and Alberto Martilli

We address a basic issue associated with mesoscale WX models used for AQ models with fine [O(1km)] grid meshes. Such meshes are small enough to capture features of the larger turbulent eddies that may exist in the planetary boundary layer (PBL), but too large for the simulation of the turbulent cascade that regulates their amplitude and structure. Higher-resolution mesoscale modeling with current PBL parameterizations developed for the larger-grid-mesh simulations can simulate the often observed quasi two-dimensional convectively induced secondary circulation (CISCs) features occurring in the heated PBL but performed without sound theoretical foundation, and thus are inaccurate and misleading. We explain and provide initial guidance using WRF LES as a guide towards more soundly based simulation of these ubiquitous CISCs and as a means for improving future mesoscale modeling parameterizations. We also explore alternative options as a means to filter such modeled CISCs for more operational applications.

Car Time: Updating onroad temporal profiles

Alexis Zubrow, Chris Allen, and James Beidler

EPA's Office of Air Quality Planning and Standards (OAQPS) has revised the diurnal temporal profiles used in the integrated system for mobile emissions, SMOKE-MOVES. Temporal profiles are used within SMOKE to distribute typically annual emissions to finer temporal resolutions (month, day, and ultimately hour). Temporal profiles help identify the "when" of emissions, which is often as important as the total amount of emissions. For the onroad sector using SMOKE-MOVES, the temporal profiles distribute vehicle miles traveled (VMT) from monthly values to daily and then to hourly VMT. Historically the hourly temporal profiles have varied by road type, but have neither varied geographically nor by vehicle type. The 2011 NEI was the first time that the Emissions Inventory System (EIS) accepted MOVES inputs. These inputs provided an opportunity to reevaluate the temporal profiles used in SMOKE-MOVES. This presentation describes the process by which new temporal profiles were developed based on the MOVES inputs and the resulting profiles that vary by geography, road type, and vehicle. These new temporal profiles have been used in the 2011 platform and in the 2011 NEI v1.

Fine-scale Meteorological Simulation of Cold Pools in Salt Lake City

Chris Misenis, Kirk Baker, Pat Dolwick

Areas in northern Utah sometimes experience long-term periods of elevated levels of fine particulate matter (PM2.5) during the winter months. Several factors (atmospheric stability, topography, lake breeze) can combine to create the formation of cold pools that can cause increases in PM2.5 levels. We present here the results of fine-scale Weather Research and Forecasting (WRF) model simulations centered over the Intermountain West and more specifically, Salt Lake City. The purpose is to better understand the capabilities of the model to replicate the conditions necessary for these long-term high pollution events. WRF was applied at grid resolutions of 12-, 4- and 1-km for December 2010 through March 2011. Several sensitivities to physical parameterizations, nudging options and initialization datasets were performed. This modeling period coincides with elevated PM2.5 in Salt Lake City and an intensive meteorological field study known as the Persistent Cold-Air Pool Study (PCAPS). The results of the WRF simulation werecompared with routine meteorological data (surface temperature, mixing ratio, wind speed and direction) as well as additional observations made during the PCAPS campaign.

Summary of the Emission Inventories compiled for the AQMEII Phase 2 Simulations

George Pouliot¹, Christian Hogrefe¹, Ryan Cleary², Junhua Zhang³, Paul Makar³, Shawn Roselle¹, Rohit Mathur¹

¹Atmospheric Modeling and Analysis Division, National Exposure Research Laboratory, Office of Research and Development, U.S. Environmental Protection Agency

²Computer Sciences Corporation, Research Triangle Park, NC

³Air Quality Research Division, Environment Canada, Toronto, Ontario, Canada

We present a summary of the emission inventories from the US, Canada, and Mexico developed for the second phase of the Air Quality Model Evaluation International Initiative (AQMEII). Activities in this second phase are focused on the application and evaluation of coupled meteorology-chemistry models over both North America and Europe using common emissions and boundary conditions for all modeling groups for the years of 2006 and 2010. We will compare the emission inventories developed for these two years focusing on the SO2 and NOx reductions over these years and compare with socio-economic data. In addition we will highlight the differences in the inventories for the US and Canada compared with the inventories used in the phase 1 of this project.

Sensitivity of Simulated Cloud Properties to Meteorological Model Configurations

Junhua Zhang and Wanmin Gong

Air Quality Research Division, Environment Canada,

4905 Dufferin Street, Toronto, ON M3H 5T4, Canada

The relationship between clouds and air quality is complex. On one hand, aerosol particles, a key index of air quality, play an important role in the formation and characteristics of clouds by acting as cloud condensation nuclei and ice nuclei. On the other hand, clouds provide a favored environment in which heterogeneous chemical reactions can take place, which in turn alters the concentrations of gases and particles in the atmosphere. Clouds also affect photochemical processes in the atmosphere by modulating radiation intensity, and they can remove gases and particles through precipitation processes. A number of modeling studies have shown the importance of clouds on air quality: most of them focused on microphysical schemes for cloud formation (e.g., Gong, et al., 2013, Impact of Cloud Microphysics Parameterization on Model Simulation of Chemistry-Aerosol-Cloud Interaction: a Case Study) while some of them have studied the sensitivity of cloud properties to meteorological model setups. During the 2004 International Consortium for Atmospheric Research on Transport and Transformation (ICARTT) field study, photochemical and heterogeneous chemical processes over North America, the North Atlantic Ocean, and Western Europe were intensively studied. One component of the ICARTT campaign over North America was the Cloud-Aerosol Study, during which two aircraft were deployed to measure cloud microphysical and chemical properties. In this study, cloud properties simulated by the Canadian Global Environmental Multiscale (GEM) meteorological model for part of the ICARTT period will be discussed. Instead of focusing on examining the sensitivity of cloud formation to different cloud microphysical parameterizations, the role of other model configuration aspects, such as model domain setup, initial conditions, forecast duration, and land surface processes, on cloud formation and life cycle will be investigated and will be shown to have significant impacts. The potential impact of modeled cloud properties on air quality modeling will also be discussed.

Top-down estimate of surface flux in the Los Angeles Basin using a mesoscale inverse modeling technique: assessing anthropogenic emissions of CO, NOx and CO2 and their impacts

J. Brioude, W. M. Angevine, R. Ahmadov, S.-W. Kim, S. Evan, S. A. McKeen, E.-Y. Hsie, G. J. Frost, J. A.Neuman, I. B. Pollack, J. Peischl, T. B. Ryerson, J. Holloway, S. S. Brown, J. B.Nowak, J. M. Roberts, S. C. Wofsy, G. W. Santoni, T. Oda, and M. Trainer

We present top-down estimates of anthropogenic CO, NO_x and CO₂ surface fluxes at mesoscale using a Lagrangian model in combination with three different WRF model configurations, driven by data from aircraft flights during the CALNEX campaign in southern California in May-June 2010. The US EPA National Emission Inventory 2005 (NEI 2005) was the prior in the CO and NO_x inversion calculations. The flux ratio inversion method, based on linear relationships between chemical species, was used to calculate the CO₂ inventory without prior knowledge of CO₂ surface fluxes. The inversion was applied to each flight to estimate the variability of single-flight-based flux estimates. In Los Angeles (LA) County, the uncertainties on CO and NO_x fluxes were 10% and 15%, respectively. Compared with NEI 2005, the CO posterior emissions were lower by 43% in LA County and by 37% in the South Coast Air Basin (SoCAB). NO_x posterior emissions were lower by 32% in LA County and by 27% in the SoCAB. NO_x posterior emissions were 40% lower on weekends relative to weekdays. The CO₂ posterior estimates were 183 Tg yr¹ in SoCAB. A flight during ITCT (Intercontinental Transport and Chemical Transformation) in 2002 was used to estimate emissions in the LA Basin in 2002. From 2002 to 2010, the CO and NO_x posterior emissions decreased by 41% and 37%, respectively, in agreement with previous studies. Over the same time period, CO₂ emissions increased by 10% in LA County but decreased by 4% in the SoCAB, a statistically insignificant change. Overall, the posterior estimates were in good agreement with the California Air Resources Board (CARB) inventory, with differences of 15% or less. However, the posterior spatial distribution in the basin was significantly different from CARB for NO_x emissions. WRF-Chem mesoscale chemical-transport model simulations allowed an evaluation of differences in chemistry using different inventory assumptions, including NEI 2005, a gridded CARB inventory and the posterior inventories derived in this study. The biases in WRF-Chem ozone were reduced and correlations were increased using the posterior from this study compared with simulations with the two bottom-up inventories, suggesting that improving the spatial distribution of ozone precursor surface emissions is also important in mesoscale chemistry simulations.

Application of a synoptic typing scheme to assess multi-year ozone model performance to variable meteorological patterns

Pat Dolwick, Christian Hogrefe, Mark Evangelista, Chris Misenis, Sharon Phillips, Norm Possiel, Shawn Roselle, Brian Timin, and Ben Wells

Several previous studies have shown the value of binning meteorological patterns into specific synoptic types and assessing how well air quality models capture the sensitivity of air pollution to varying meteorology (Eder et al., 2006; Hogrefe et al., 2012). To date, the majority of these studies have been constrained to relatively short modeling periods, typically one year or less. This analysis assesses how CMAQ ozone model performance for a 5-year simulation period (2006-2010) varies as a function of several discrete patterns of mean sea-level pressure patterns over the eastern U.S. The meteorological typing scheme was developed using the European Centre for Medium-Range Weather Forecasts (ECMWF) ERA-interim reanalysis fields. The CMAQ modeling was completed over a 12-km resolution domain using year-specific emissions estimates and meteorological inputs from the Weather Research Forecast (WRF) model. The first stage of the analysis determined the magnitude of the ozone anomalies associated with each of the 15 identified synoptic types. In some locations, mean daily peak 8-hour ozone concentrations can vary by as much as +/- 10 ppb as a function of synoptic type. Additional analysis will show how model performance patterns of ozone bias and error vary by synoptic type. Additionally, the analysis will determine the relationships between performance and meteorological type vary by year over the 5-year period.

Evaluation of NOx emission inventories in California using multi-satellite data sets, AMAX-DOAS and in-situ airborne measurements, and regional model simulations during the CalNex field campaign

S.-W. Kim^1,2, S. Baidar^1,3, K. F. Boersma^4,5, J. Brioude^1,2, E. Bucsela⁶, J. P. Burrows^7,8, E. A. Celarier^9,10, R. C. Cohen^11,12, G. J. Frost^1,2, N. A. Krotkov⁹, L. N. Lamsal^9,10, R. V. Martin¹³, S. A. McKeen^1,2, H. Oetjen¹, I. Pollack^1,2, A. Richter⁷, A. R. Russell¹¹, T. Ryerson², M. Trainer², L. C. Valin¹¹, and R. Volkamer^1,3

[1]{Cooperative Institute for Research in Environmental Sciences, University of Colorado, Boulder, CO 80309, USA}

[2]{NOAA Earth System Research Laboratory, Boulder, CO 80305, USA}

[3]{Chemistry and Biochemistry, University of Colorado, Boulder, CO 80309, USA}

[4]{Royal Netherlands Meteorological Institute (KNMI), De Bilt, the Netherlands}

[5]{Eindhoven University of Technology, Eindhoven, the Netherlands}

[6]{SRI International, Menlo Park, CA 94025, USA}

[7]{Institute of Environmental Physics, University of Bremen, Germany}

[8]{Center for Ecology and Hydrology, Maclean Building, Benson Lane, Crowmarsh Gifford, 16 Wallingford, Oxfordshire, OX10 8BB, United Kingdom}

[9]{NASA Goddard Space Flight Center, Laboratory for Atmosphere, Greenbelt, MD 20771, USA}

[10]{Universities Space Research Association (USRA), MD, USA}

[11]{Department of Chemistry, University of California Berkeley, Berkeley, CA, USA}

[12]{Department of Earth and Planetary Sciences, University of California Berkeley, Berkeley, CA, USA}

[13]{Department of Physics and Atmospheric Science, Dalhousie University, Halifax, Canada}

Satellite NO₂ column measurements indicate large NO_x emissions from urban and agricultural sources in California. In this presentation, we highlight the NO_x sources identified in California using the satellite measurements. Comparison of regional model-simulated NO₂ columns with satellite retrievals has proven useful in evaluating emission inventories for various sectors. We compare the NO₂ columns from the WRF-Chem model with the multi-satellite data sets from different instruments and retrieval groups for a variety of California sources. Use of multiple satellite data sets help to define the uncertainties in the satellite retrievals. In addition, the CalNex 2010 intensive field campaign provides a unique opportunity to independently assess California's emission inventories. CU-AMAX-DOAS and in-situ airborne observations from CalNex 2010 and fine-resolution model simulations are used to estimate the accuracy of the satellite NO₂ column retrievals.

*Note: please assign this presentation to either "Emission Inventories, Models and Processes session" or "Model evaluation and analysis".

Recent Advances in High-Resolution Lagrangian Transport Modeling

Jennifer Hegarty, Roland Draxler, Ariel Stein, Jerome Brioude, Thomas Nehrkorn, Janusz Eluszkiewicz, John Henderson, Mark Leidner, Marikate Mountain, Kathryn McKain, Fong Ngan, Steven Wofsy, Phillip DeCola, and Arlyn Andrews

The accuracy of air quality simulations reflects the fidelity of the atmospheric transport model employed that in turn is highly dependent on the accuracy of the meteorological input data. In this work we present some recent advances in Lagrangian transport modeling obtained through high-resolution Lagrangian particle dispersion model (LPDM) simulations and field measurements. First we applied controlled tracer release experiments to evaluate the Hybrid Single Particle Lagrangian Integrated Trajectory Model (HYSPLIT), Stochastic Time-Inverted Lagrangian Transport (STILT) and FLEXPART LPDMs driven by identical meteorological fields from the Weather Research and Forecasting (WRF) model. This rigorous evaluation demonstrated that the WRF configuration was the dominant factor governing the accuracy of the transport simulations, rather than inter-model differences between the three LPDMs. Next, we describe the application of STILT driven by WRF inputs (WRF-STILT) to top-down quantification of greenhouse gas (GHG) fluxes in urban areas including Salt Lake City, Boston and Krakow, Poland. These studies demonstrate the improvements achieved by increasing the spatial resolution from 4 to 1.33 km horizontal grid spacing and employing advanced parameterizations such as an urban canopy model in WRF, and the feasibility of incorporating advanced instrumentation such as Mini Micro-Pulse LiDARs (Mini MPL) to evaluate and potentially improve the fidelity of the WRF simulations.

Developing Spatial Surrogates for High-Resolution Modeling Domains

Carol McClellan, J. Wayne Boulton, Martin Gauthier, Saba Hajaghassi, Jeff Lundgren, Greg Conley

RWDI, Guelph, Ontario, Canada

Michael Moran, Junhua Zhang, Qiong Zheng

Environment Canada, Toronto, Ontario, Canada

Louise Aubin, Kim McAdam

Region of Peel - Public Health, Mississauga, Ontario, Canada

As high-end computing resources become less costly and increasingly available, air quality modellers are continually migrating towards higher-resolution modelling grids. The temptation behind using higher-resolution grids is the assumption that this will inherently lead to improved accuracy of model results. However, with an increase in spatial resolution, modellers need to be equally mindful of the inputs to their models.

Spatial surrogates are used to allocate geographically aggregated (e.g., county or state-wide) emissions data to model grid cells. The higher the grid resolution, the more important it is to assess the appropriateness of both the spatial representation (the "where") and weighting factors (the "what") used to create spatial surrogates. In this paper we highlight the importance of these issues with examples from a large (100 km by 108 km), 1.0 km grid resolution SMOKE and CMAQ modelling exercise being performed for the Region of Peel Department of Public Health in Southwestern Ontario, Canada.

Assessment of deposition of reactive nitrogen to various national parks using CMAQ 5.0.1- bidirectional ammonia exchange

Aika Y. Davis, Armistead G. Russell

Deposited reactive nitrogen, which originates from NO_x and NH₃ emissions, can pose a threat to ecosystems when exceeding critical load. CMAQ v5.0.1 with bidirectional exchange of ammonia was used to look at the sensitivity of nitrogen deposition to emissions from domestic anthropogenic emissions and natural sources. Brute force method was applied to study the sensitivity of reactive nitrogen deposition to NO_x emissions from mobile, power plant, and biogenic as well as to NH₃ emissions from livestock, fertilizer, and biogenic. Simulations were done for the year of 2010 over continental US using 36km resolution. Furthermore, we focused on multiple national parks with 4km grid resolution.  Sensitivity analysis on nitrogen emissions and grid resolutions are presented.

EPA's Particulate Matter (PM) Augmentation Tool: Automating Quality Assurance and PM Speciation to Generate a Model Ready Inventory

Dr. Jonathan Dorn, David Cooley, Roy Huntley

The Consolidated Emissions Reporting Rule requires that state, local, and tribal agencies (SLT) submit particulate matter (PM) emissions for point and nonpoint sources to EPA's National Emissions Inventory (NEI). Since PM is both a National Ambient Air Quality Standard (NAAQS) pollutant and a major contributor to visibility impairment, the PM_2.5 NAAQS and the Regional Haze Rule each emphasize emissions inventory development for the PM species required in regional air quality modeling. While submission of PM emissions to the NEI by SLT agencies should include filterable and primary PM (PM10-PRI, PM10-FIL, PM25-PRI, and PM25-FIL) along with condensible PM (PM-CON), SLT agencies often submit incomplete speciation. Augmentation of the PM species in the NEI point and nonpoint source inventories is necessary to ensure completeness of the PM inventories for air quality modeling and to ensure that SLT inventories do not contain erroneous pollutant reporting. This presentation explains the basis of EPA's PM Augmentation Tool and the recently revised procedures developed to correct reporting inconsistencies and to populate missing PM species in the NEI.

Studying distribution and long-range transport of polycyclic aromatic hydrocarbons in Europe Application of the SMOKE-EU/CMAQ modeling system

Christos Efstathiou, Johannes Bieser, Jana Matejovicova, Gerhard Lammel

Polycyclic aromatic hydrocarbons (PAHs) are persistent organic pollutants (POPs) that are unintentional by-products of incomplete combustion processes. Several PAHs are of public concern because of their high toxicity (e.g. benzo[a]pyrene, dibenzo[a,h]anthracene, indeno[1,2,3,-cd]pyrene). PAHs can travel through the atmosphere across long distances and are observed in remote regions far from sources. PAHs are regulated in many countries and internationally by the United Nations Economic Commission for Europe’s (UNECE’s) Convention on Long-RangeTransboundary Air Pollution (CLRTAP). However, there are still questions about the pathways by which they reach remote regions, specifically with respect to gas-particle partitioning and oxidation of both the gas- and particulate-phase species. In this study, we employ a modified version of the CMAQ model along with emission inventories for selected PAHs obtained using SMOKE-EU. The influence of uncertain PAH properties on atmospheric transport and source-receptor relationships on the regional scale is investigated and the results are evaluated using observational data.  

Addressing Science and Policy Needs with Community Emissions Efforts

Gregory Frost (1), Claire Granier (1,2), Paulette Middleton (3), & Leonor Tarrasen (4)

(1) University of Colorado/CIRES and NOAA/ESRL, Boulder, Colorado, USA

(2) CNRS/INSU, LATMOS-IPSL, UPMC, Paris, France

(3) Panorama Pathways, Boulder, CO, USA

(4) Norwegian Institute for Air Research, Kjeller, Norway

Accurate, timely, and accessible emissions information is critical for understanding and making predictions about the atmosphere. We present an overview of the Global Emissions InitiAtive (GEIA, http://www.geiacenter.org/), a community-driven, joint activity of IGAC/iLEAPS/AIMES within the International Geosphere-Biosphere Programme. Since 1990, GEIA has served as a forum for the exchange of expertise and information on anthropogenic and natural emissions of trace gases and aerosols. GEIA supports a worldwide network of about 1300 emissions data developers and users, providing a solid scientific foundation for atmospheric chemistry research. Moving forward, GEIA seeks to build bridges between the environmental science, regulatory, policy, and operational communities. GEIA's core activities include 1) facilitating analysis that improves the scientific basis for emissions data, 2) enhancing access to emissions information, and 3) strengthening linkages within the international emissions community. GEIA pursues these activities in collaboration with the ECCAD project (http://pole-ether.fr/eccad) and as a member of the GEO Air Quality Community of Practice (http://wiki.esipfed.org/index.php/GEO_AQ_CoP). GEIA welcomes new partnerships that advance emissions knowledge for the future.

Modeling the Impact of Emissions of HCL During Rocket Launch Events in the Region of Alcantara Launch Center

Erick Giovani Sperandio Nascimento^a, Davidson Martins Moreira^a, Gilberto Fisch^b, Taciana Toledo de Almeida Albuquerque^a

^a Federal University of Esperito Santo

Av. Fernando Ferrari, s/n
29060-970 - Vitoria/ES - Brazil

Tel. (Fax): (27) 4009-2677

erick@lcad.inf.ufes.br, davidson@pq.cnpq.br, taciana@model.iag.usp.br

^b Institute of Aeronautics and Space

Praja Marechal Eduardo Gomes, 50

12228-904 - Sao Jose dos Campos/SP - Brazil

Tel. (Fax): (12) 3941-5635

gilbertofischgf@iae.cta.br

During vehicle launching events, the burning of rocket engines causes the formation of a huge and hot cloud composed by buoyant exhaust products during a few minutes. These products are mainly composed by alumina, carbon monoxide and hydrogen chloride (HCl). In this work, we are interested in assessing the environmental impact of such huge emissions in the Alcantara Launch Center (ALC), Brazil, specially paying attention to the hydrogen chloride, due to the important role it plays in ambient air when released from an anthropogenic source. Due to the lack of experimental data from rocket exhaust clouds in ALC, we conducted the experiments based on hypothetical rocket launch events found in the literature using the CMAQ modeling system to simulate the dispersion, chemical interactions, transport, photolysis, dry and wet deposition of hydrogen chloride and its derivatives. This work is part of a major project whose intention is to provide a more meaningful model that consider the Brazilian site characteristics, representing an effort in the construction of a computational tool for normal and/or accidental events during rocket launchings, making possible to predict or simulate the concentration in accordance with emergency plans and pre and post-launchings for environmental management.

Alternative Way to Reduce Vehicle Emissions in Summer with the help of Car Window Filming and Car Window Filming's Economic Benefits over WA, NY, NC, U.S.A. and Istanbul, TURKEY

Nazenin GURE
Environmental Engineer
M.Sc. Student, Mechanical Engineering Department, Marmara University, Istanbul, TURKEY

Mustafa YILMAZ
Assist. Prof. Dr., Mechanical Engineering Department, Marmara University, Istanbul, TURKEY

Validation of Soumi-NPP VIIRS marine isoprene emission products with in-situ measurements

Hang Lei, Daniel Tong, Li Pan, Pius Lee, Menghua Wang, Brett Gantt

Marine emissions of non-methane hydrocarbons influence atmospheric chemistry and cloud formation over the oceans and coastal areas. However, such sources are not included in CMAQ simulations. Using a revised emission algorithm originally proposed by Gantt et al. (2009), we have developed a marine isoprene emission product based on the ocean color data retrieved from the Soumi-NPP Visible Infrared Imaging Radiometer Suite (VIIRS) and global meteorology simulated by NOAA Global Forecasting System (GFS). This work presents the first comprehensive validation of the VIIRS marine isoprene emission product based on a collection of in-situ measurements conducted at various open-ocean and near-shore sites around the globe. The results show that the satellite dataset captures the total isoprene budget over global and regional scales. To further test this product, we compare the emission flux with estimated isoprene emission fluxes with measurement. We also estimated additional marine isoprene emission fluxes through consideration of air-sea exchange pathways. Measurements of isoprene concentration in the ocean mixing layer and corresponding marine boundary layer isoprene concentration are used. The results show that the measurement-based marine isoprene fluxes in coastal regions have large uncertainty. Satellite derived isoprene emission fluxes are within the range of estimated fluxes, but lower than the mean flux estimated by previous studies. A comparison of additional isoprene data and VIIRS marine isoprene emission fluxes show that the mean fluxes of air-sea exchanged derived flux data are close to that observed by VIIRS. VIIRS data captures the seasonal variability and the spatial distribution of marine isoprene emission. The study also indicates that the estimated fluxes based on surface measurements are sensitive to converting methods.

Emissions Inventory for the Metropolitan Area of Vitoria ES, Brazil Using the SMOKE Modeling System

Ayres G. Loriato¹, Taciana T. de A. Albuquerque¹, Renato S. Marinho², Alexandre M. Santiago¹, Nadir Salvador¹, Erick G. S. Nascimento¹, Neyval C. Reis. Jr.¹.

¹Federal Universisity of Espírito Santo - Environmental Engineering Department

¹Federal Universisity do Espírito Santo - Geography Departament

contact: ayresloriato@yahoo.com.br

Large industrial facilities were installed in the northeastern region of Metropolitan Area of Vitória, Espírito Santo, Brazil. These facilities were built to enable production and port exportation of goods from the ore, steel and pelletizing industries, but the main wind direction in the area carry the gases and particles emitted to the most populated sector of the city. This problem has led to legal disputes between residents and companies, generating additional costs for the enterprises in the form of mitigating actions and upsetting the populace, which has to deal with the impact of the industrial emissions, including breathing the polluted air. Emission inventories are a fundamental input to atmospheric chemical transport models (CTMs). This study presents an adaptation of the official emission inventory for anthropogenic sources covering the Metropolitan Area of Vitoria (MAV) for the reference year 2009 using the SMOKE (Sparce Matrix Operator Kernel Sistem). The SMOKE model is used to compile a high spatially and temporally resolved emission inventory. The bulk of the pollution is due to motorization, facilities, and residential activities. The inventory provided by the local environmental agency has been adapted and tested with the air quality model (AQM), Chemistry Model Air Quality(CMAQ). 

High-resolution CMAQ simulation of air pollution over the Metropolitan Area of Vitria, Brazil.

Alexandre M. Santiago¹, Ayres G. Loriato¹, Nadir Salvador¹, Erick G. S. Nascimento¹, Renato S. Marinho¹, Neyval C. Reis Jr¹, Jane M. Santos¹, Taciana T. de A. Albuquerque¹

1. Environmental Engineering Department - Federal University of Espirito Santo - UFES - BRAZIL
2. Geography Department - Federal University of Espirito Santo - UFES - BRAZIL

magalhaes.es@gmail.com

The main objective of this study is to evaluate the air pollution concentrations over the Metropolitan Area of Vitoria (MAV), Esperito Santo, Brazil using The Models-3 Community Multiscale Air Quality Modeling System (CMAQ) version 4.6. An experimental campaign was performed during the winter of 2012, from July 22 to 31, to quantify the aerosols formation and transportation using a LIDAR and SODAR instruments. Meteorological numerical data was obtained using the WRFv3.2.1 (Weather Research and Forecasting) model. Four nested grid was used, 27km (70 x 70 cells), 9km (100 x 100 cells), 3km (100 x 100 cells) and 1km (64 x 82 cells), only the 1-km domain was aligned with the CMAQ domain, which covers the most polluted cities on the State. CMAQ modeling simulations were conducted over 306 hours from 20 to 31 of July, 2012. The SMOKEv3.7 emissions model was applied to build a spatially and temporally resolved the official emission inventory for MAV. The CMAQ and SMOKE domains consist of 64 x 82 grid cells with 1 km horizontal spacing and 20 vertical layers. The air quality simulations use measured concentrations as initial and boundary conditions. Aerosol processes and aqueous chemistry in CMAQ (AERO4) were used, as well as the Carbon Bond V gas phase mechanism. The highest PM10 average concentrations measured at the Local Air Quality Monitoring Stations was observed in Cariacica (54μg/m³), Serra (Laranjeiras) (43μg/m³), Vila Velha Downtown (31μg/m³). The numerical results showed a varied considerably among the local stations, but Vitoria Downtown station showed much less variation, where the average PM10 concentration observed was 23μg/m³ and the simulated was 31μg/m³.

FEST-C 1.0 for CMAQ Bi-directional NH3 Modeling and Spatial Allocator 4.1

Accurate estimation of ammonia emissions in space and time has been a challenge in meso-scale air quality modeling. For instance, fertilizer applications vary in the date of application and amount by crop types and geographical area. With the support of the U.S EPA, we have developed an agricultural Fertilizer Emission Scenario Tool for CMAQ (FEST-C) to be used with the new CMAQ bi-directional option for estimating ammonia flux which was released last year. The FEST-C system contains three components: FEST-C interface, Environmental Policy Integrated Climate (EPIC) modeling system, and Spatial Allocator (SA) Raster Tools. The FEST-C interface is a Java-based interface which guides users through the simulation of daily fertilizer application information using the EPIC model. This information is a required input for CMAQ bi-directional NH3 modeling. The FEST-C Interface integrates WRF/CMAQ with EPIC through the current release of SA Raster Tools system. In addition, under the EPA's support we have completed the land use processing tool which generates land use data to be consistently used across the WRF/CMAQ modeling system for bi-directional NH3 flux, biogenic emission, land surface fluxes, and dry deposition modeling. The goal of this presentation is to describe the FEST-C 1.0 system and SA 4.1 which we have just released through the CMAS. We will briefly demonstrate the use of the FEST-C system for 2001 and 2006 fertilizer simulations on the CMAQ CONUS 12km domain (see presentation by Cooter et al. for more details in different domain applications). We will also demonstrate the use and applications of the SA 4.1 land use tool for WRF/CMAQ modeling.

2011 NEI v1

Madeleine Strum, Venkatesh Rao, Sally Dombrowski, Laurel Driver, Jonathan Miller, Roy Huntley, Jennifer Snyder, Marc Houyoux, Rhonda Thompson, Lee Tooly

The poster will highlight version 1 of the  2011 National Emission Inventory.  The poster will review the methodology,  sources of data and will provide sector summaries of criteria pollutants and select hazardous air pollutants.

Impact of dramatic land use change on gaseous pollutants emission from biomass burning

Hongmei Zhao^1,3, Daniel Q. Tong^2,3, Xianguo Lu¹, Guoping Wang¹, Pius Lee³

¹ Key Laboratory of Wetland Ecology and Environment, Northeast Institute of Geography and Agroecology, Chinese Academy of Sciences, Changchun, Jilin,130102, China

² UMD/Cooperative Institute for Climate and Satellites, College Park, MD, 20740, USA

³ NOAA/Air Resources Laboratory, College Park, MD, 20740, USA

Biomass burning is a global phenomenon that releases large qualities of gases and aerosol particles that affect the atmospheric chemistry and climate. Northeast China is one of the three areas that experienced the largest temperature increase in the past century, during which this region has also experimented the most dramatic land use change. From 1986 to 2005, the wetland area was decline from 139×10⁴ ha to 84×10⁴ ha, while the rice paddy area was increased from 57.7×10⁴ ha to 150.9×10⁴ ha, and the dry land area from 394.8×10⁴ ha to 449.3×10⁴ ha in this region (Huang, 2009). During this rapid cropland reclamation, prescribed burning was widely used as an effective measure to enforce land use changes. At present in China, the agricultural residues are increasingly being burned in the field, usually right after harvesting. This study combines measured emission factor of dominant cropland (rice and soybean) and wetland plants (Calamagrostis angu-stifolia, Carex lasiocarpa, Carex pseudo-curaica), and remote sensing data of the land use to estimate the effect of dramatic land use change on gaseous pollutants emission from biomass burning. Wetland plant and crop residues were collected in the field after harvesting (in October), and grounded and passed through 60 mesh sieve. About 5.0 g samples were burned in a burning device. The emission factors of CO₂, CO, SO₂, NO and C_XH_Y were determined by an integrative smoke analyzer (KM-9106, England). For each vegetation type, three burning tests were conducted successively. Gaseous pollutant concentrations were recorded every 10 s. The release curves were fitted by the Table Curve 2D V5.0 Trial. Results show that the EFs of CO₂, CO, C_XH_Y, SO₂, and NO are significantly different in rice, soybean and wetland vegetation. Next, we estimated emissions of gaseous pollutants from wetland, rice paddy and dry land in this region, according to EFs and mass of burned fuel. The results show that from 1986 to 2005, the total emissions of CO₂, CO, C_XH_Y, SO₂ and NO have increased by 18.6%, 35.7%, 26.8%, 66.2% and 33.2%, respectively. In addition, the emission allocation of wetland, rice paddy and dry land were calculated. The contribution of wetland was declined from 1986 to 2005. And the main source of gaseous pollutants was changed from wetland to agricultural crop residue, as a result of more wetlands converted into cropland in this period.

Can we bridge the gap between climate change projections and users to stop misuse of climate change data

Jared H. Bowden

Tanya Otte

Adrienne Wooten

Ryan Boyes

North Carolina State, University of North Carolina at Chapel Hill, and the U.S. Environmental Protection Agency in partnership with the Department of Interior Southeast Climate Science Center hosted a workshop in May 2013 titled Regional Climate Variations and Change for Terrestrial Ecosystems. This multi-disciplinary effort was unique because the intent was to bridge the knowledge gap between climate and ecosystem sciences. The workshop provided a summary of recommendations that likely applies to a broader audience and could provide to be a useful resource for other disciplines, including those in the air quality community interested in climate change projections. This experience also provides useful insight into future workshops interested in bridging the gap between climate change projections and other disciplines.

Probe into interactions between mineral dust and biomass burning aerosols over East Asia using WRF/CMAQ

Xinyi Dong, Joshua S. Fu, Kan Huang

       Impact of mineral dust and biomass burning aerosols on air quality and regional climate has been well documented in the last few decades, but the knowledge about the interactions between these two different aerosols is very limited, mainly because they occur in different climate regimes and rarely encounter each other. While East Asia is greatly affected by dust storms in spring from Taklamakan and Gobi deserts, it also suffers from significant biomass burning emission from Southeast Asia during the same season. Observations from both surface monitoring and satellite data indicated that mineral dust and biomass burning aerosols might approach to coastal area of East Asia simultaneous and thus have a very unique impact on the local atmospheric environment and regional climate (Huang et al., 2010). Interactions between mineral dust and biomass burning aerosols may also lead to chemical composition changes and alter the size distribution of aerosols.  So in this study, WRF/CMAQ modeling system will be applied over East Asia to probe into the interactions between these two aerosols and assess their impacts on local atmospheric environment and regional climate forcing budget.  

The WRF/CMAQ modeling domain has horizontal resolution of 27km x 27km, and 34 vertical sigma-pressure coordinates with top height at 50 hPa. Meteorology filed is provided by WRF. The Intercontinental Chemical Transport Experiment-Phase B (INTEX-B) emission (Zhang et al., 2009) will be used to provide anthropogenic emission, the Global Fire Emission Database (GFED) will be used to provide the biomass burning emission, and Model of Emissions of Gases and Aerosols from Nature (MEGAN) will be used to provide biogenic emissions. Dust module in CMAQv5.0 will be applied to provide the mineral dust emissions over East Asia, and global model GEOS-Chem will be used to provide initial and boundary conditions for regional modeling with CMAQ.

North American CMAQ Modeling of Cyclic Siloxanes

Nathan J. Janechek¹, Jaemeen Baek¹, Rachel A. Yucuis², Keri C. Hornbuckle², and Charles O. Stanier¹

¹Department of Chemical and Biochemical Engineering, University of Iowa, Iowa City, Iowa, USA

²Department of Civil and Environmental Engineering, University of Iowa, Iowa City, Iowa, USA

 

 

 

 

            Cyclic volatile siloxanes octamethylcyclotetrasiloxane (D4), decamethylcyclopentasiloxane (D5) and dodecamethylcyclohexasiloxane (D6) are widely used chemicals present in personal care products such as shampoos and deodorants.  Due to their physical properties they readily volatilize and partition preferentially to the atmosphere; typical concentrations are 18-210 ng m^-3 (D5) and 10-54 ng m^-3 (D4).  Despite their ubiquitous use and widespread environmental presence, the number of regional models that include these compounds is very small; and North American measurements needed for model evaluation are very sparse.  Atmospheric cyclic siloxanes primarily degrade via reaction with hydroxide (OH) radicals leading to half-lives of approximately 10-30 days with the potential for long range transport.  We are currently exploring the fate of OH reacted cyclic siloxanes and the possible relationship to explain ultrafine silicon aerosols.  Cyclic siloxanes, especially D4 and D5 have been targeted as potentially harmful ecological contaminants due to concerns of bioaccumulation and chemical persistence.  Modeling of transport and reaction can be used to better understand the regional distribution and test assumptions regarding transport, chemistry, and emissions.  The Community Multiscale Air Quality (CMAQ) model contains preconfigured chemical mechanism data that includes photochemical chemistry such as OH radicals but does not include cyclic siloxanes.  These species will be added using the CHEMMECH program and CMAQ will be used to map D4, D5, and D6 concentrations over North America at a resolution of 36 km.  The regional distribution will be evaluated and the model accuracy will be assessed by comparing model results to cyclic siloxane measurements in West Branch, IA; Cedar Rapids, IA; and Chicago, IL.  In addition to model-observation comparison of cyclic siloxane concentrations, concentration ratios will also be evaluated since seasonal and spatial variation in siloxane compound ratios contains information about relative emission rates and atmospheric lifetimes.  Preliminary modeling suggested D5/D4 ratios of 1.5 and 1.3 in urban and rural continental locations, while measurements indicated values of 3.9 and 1.9, respectively.  This discrepancy will be explored by varying siloxane-related parameters in CMAQ.

An Extended Approach to Calculating Relative Response Factors for use in the Attainment Demonstration of the 1-hour Ozone NAAQS

Sarika Kulkarni¹, Daniel Chau¹, Jeremy Avise¹, John DaMassa¹, and Ajith P. Kaduwela^1,2

¹ Regional Air Quality and Science Division, Air Resources Board, California Environmental Protection Agency, Sacramento, CA 95814

² Department of Land, Air and Water Resources and the Air Quality Research Center, University of California, Davis, CA 95616

With the promulgation of the National Ambient Air Quality Standards (NAAQS) for 8-hour ozone (O₃), the United States Environmental Protection Agency (U.S. EPA) advocated that the results of the photochemical air quality models be used in a relative sense. In doing so, the U.S. EPA provided guidance on how to calculate Relative Response Factors (YYFs) that can project current Design Values (DV) into the future for the purpose of determining the attainment status with respect to NAAQS. The average YYFs recommended by the U.S. EPA represent the response of the photochemical model over a broad range of O₃ concentrations above a specified cutoff threshold. However, it is known that the O₃ response to emissions reductions generally increases at higher O₃ concentrations compared to lower concentrations. We present a segmented YYF method, termed band-YYFs, which take into account the difference in model response at different O₃ concentrations. The new band-YYF method is demonstrated in the San Joaquin Valley of California for the now revoked 1-hour O₃ NAAQS. The applicability of band-YYFs to the attainment demonstrations for both 8-hour O₃ and particulate matter (PM_2.5) NAAQS are also discussed.

Seasonal Modeling of PM2.5 in Californias San Joaquin Valley

Jianjun Chen¹, Jin Lu¹, Jeremy Avise¹, John DaMassa¹, Michael J. Kleeman², and Ajith Kaduwela^1,3

¹ Regional Air Quality and Science Division, California Air Resources Board, 1001 I Street, Sacramento, CA 95814

² Department of Civil and Environmental Engineering, University of California, Davis, CA 95616

³ Air Quality Research Center and the Department of Land, Air and Water Resources, University of California, Davis, CA 95616

California's San Joaquin Valley (SJV) is in non-attainment for the 2006 revised 24-hour PM_2.5 National Ambient Air Quality Standard (NAAQS) established by the United States Environmental Protection Agency. As a part of the emissions control strategy development to bring the SJV into attainment of the standard, the Community Multi-scale Air Quality (CMAQ) model was used to simulate PM_2.5 formation and its response to precursor emission reductions. Since 24-hour PM_2.5 violations typically occur during winter months in the SJV, simulations were only conducted for the first (January-March) and fourth (October-December) quarters of 2007. Sensitivities of the 24-hour average PM_2.5 concentrations to precursor emissions were investigated based on the forecasted baseline 2019 emissions inventory. Individual 50% reductions in anthropogenic primary PM_2.5 and oxides of nitrogen (NO_x) emissions in the SJV lowered the PM_2.5 concentration in Bakersfield by 27% and 24%, respectively. In contrast, 50% reductions in anthropogenic ammonia and sulfur oxides emissions only lowered the PM_2.5 concentration in Bakersfield by 4% and 1%, respectively. A 50% reduction in anthropogenic volatile organic compounds emissions did not provide any PM_2.5 benefits in Bakersfield. The PM_2.5 sensitivities to different precursors were fairly consistent among different urban locations in the SJV. This study demonstrated that the current control strategy of controlling primary PM_2.5 and NO_x emissions in the SJV will continue to be effective for further reducing PM_2.5 in the SJV beyond 2019. We also present carrying capacity diagrams for key sites in the SJV and for the first time, compare and contrast precursor sensitivities for 2000 and 2019.

Characteristics of long-range transport of sulfate, nitrate, and ammonium aerosol in Northeast Asia

Yoo Jung Kim¹, Gregory Carmichael¹, Jung-Hun Woo², Young Sunwoo²

¹University of Iowa, ²Konkuk University

The long-range transport of air pollutant is a relevant issue in Northeast Asia due to large emission of SO2 and NOx. Air pollutants emitted from industrialized regions along the East China coast can be transported over downwind region by the prevailing westerlies. Recently, China and Korea are changing air quality standard for particle pollutant from PM10 to PM2.5 in 2012 and 2015, respectively, because they recognize again the importance of health risk of aerosol. The long-rang transported fine particle certainly attributes to Korean air quality, but there are many unknowns on the quantity, transport pattern, and secondary aerosol production mechanism despite the fact with many studies have been performed.

Dominant contributors of PM2.5 are sulfate, nitrate and ammonium in Korea. Specially high relative contributions of secondary aerosol appear for westerly wind cases. The secondary aerosols are produced by converting from SO2 and NOx during the long-range transport but the contribution varies dramatically depending on season and wind pattern. Sulfate is consistently the primary contributor of PM2.5 still now but we should more concern nitrate because that NOx emissions of China is increasing steeply since 2000 by leading powerplant, industry, and transport, despite downward trend of SO2.

Impact of local and non-local historical air pollution emissions reductions on surface air quality during record breaking heat

Christopher P. Loughner (Earth System Science Interdisciplinary Center - University of Maryland), Bryan Duncan (NASA Goddard Space Flight Center), Jennifer Hains (Maryland Department of the Environment), Melanie Follette-Cook (Morgan State University), Yasuko Yoshida (SSAI), Kenneth E. Pickering (NASA Goddard Space Flight Center), Maria Tzortziou (Earth System Science Interdisciplinary Center - University of Maryland)

Maryland recorded 14 code orange days (maximum 8 hr average ozone greater than 75 ppbv) and 3 code red days (maximum 8 hr average ozone greater than 95 ppbv) during July 2011, which was the hottest month for the Baltimore-Washington metropolitan area in recorded history. However, the air quality during this month could have been worse if it were not for air pollution regulations curtailing emissions that had been put in place since 2002. There have been significant NOx emissions reductions since 2002 in the eastern and central US through a combination of the Environmental Protection Agency (EPA) NOx State Implementation Plan (SIP) call (which required 22 states and the District of Columbia to regulate NOx emissions to mitigate ozone transport), the NOx Budget Trading Program, subsequent EPA rules, court-orders, and state regulations. As reported by the EPA's National Emissions Inventory (NEI), NOx emissions nationwide have been reduced 37% between 2002 and 2011. The benefit of these emissions reductions within and outside of the Baltimore non-attainment zone will be presented by comparing CMAQ simulations for July 2011 from a 12 km domain over an eastern US domain and a 4 km domain over the Mid-Atlantic with 1) 2011 emissions everywhere, 2) 2002 emissions everywhere, 3) 2002 emissions within the Baltimore non-attainment zone and 2011 emissions everywhere else, and 4) 2011 emissions within the Baltimore non-attainment zone and 2002 emissions everywhere else. The model results indicate the historical emissions reductions from 2002 to 2011 prevented 9 to 13 ozone exceedance days throughout much of the Ohio River Valley and 3 to 9 ozone exceedance days throughout the Baltimore-Washington metropolitan area for the month of July 2011. The impact of local emissions reductions vs. regional emissions reductions on surface air quality and the importance of regional transport will be shown. The base case simulation with emissions appropriate for 2011 everywhere was evaluated with ground-, ship-, aircraft-, and satellite-based observations, which include measurements made during the DISCOVER-AQ (Deriving Information on Surface Conditions from Column and Vertically Resolved Observations Relevant to Air Quality) and GeoCAPE-CBODAQ (Geostationary Coastal and Air Pollution Events-Chesapeake Bay Oceanographic Campaign with DISCOVER-AQ) field campaigns.

The effect of alternate representations of lake temperatures and ice on regional climate simulations

Megan S. Mallard, Chris Nolte, Russ Bullock, Tanya Otte, Jerry Herwehe, Kiran Alapaty, Jonathan Gula

Lakes can play a significant role in regional climate, modulating inland extremes in temperature and enhancing precipitation.  Representing these effects becomes more important as regional climate modeling (RCM) efforts focus on simulating smaller scales.    When using the Weather Research and Forecasting (WRF) model to downscale future global climate model (GCM) projections into RCM simulations, model users typically must rely on the GCM to represent temperatures at all water points.  However, GCMs have insufficient resolution to adequately represent even large inland lakes, such as the Great Lakes.  Some interpolation methods, such as setting lake surface temperatures (LSTs) equal to the nearest water point, can result in inland lake temperatures being set from sea surface temperatures (SSTs) that are hundreds of km away.  In other cases, a single point is tasked with representing multiple large, heterogeneous lakes.  Similar consequences can result from interpolating ice from GCMs to inland lake points, resulting in lakes as large as Lake Superior freezing completely in the space of a single timestep.  The use of a computationally-efficient inland lake model can improve RCM simulations where the input data is too coarse to adequately represent inland lake temperatures and ice (Gula and Peltier 2012). 

This study examines three scenarios under which ice and LSTs can be set within the WRF model when applied as an RCM to produce 2-year simulations at 12 km grid spacing.  In order to assess the model’s performance, the 1.875p NCEP–DOE Atmospheric Model Intercomparison Project Reanalysis-2 (R2) data is used as a proxy for a typically-coarse GCM.  This first control run (CTL-R2) represents the usual performance when the GCM is tasked with representing ice and LSTs.  The second control run (CTL-Ob) is driven with high-resolution observations of ice from the National Ice Center and lake surface temperatures from the Advanced Very High Resolution Radiometer (AVHYY) dataset.  This run is a “best case scenario”, where available products that are appropriate for use with a 12-km grid are utilized.  However, such an option is not actually available when producing future simulations.  Therefore, CTL-Ob is a benchmark for the performance of the WRF model when LSTs and ice are well-represented, but does not provide guidance on choosing a preferred RCM setup for future simulations.  The final run utilizes a version of WRF that is dynamically coupled with the Freshwater Lake (FLake) model, providing simulated LSTs and ice concentrations.  FLake is a 1D column model, consisting of a two-layer parametric representation of a time-varying temperature profile that includes a mixed layer and a thermocline extending down to a layer of thermally-active sediment.  Evaluation of these three runs will focus on 2-m temperatures and rainfall, assessing what impact the choice of lake representation has on WRF’s performance in an RCM setu

Assessing the Impacts on Smoke, Fire and Air Quality Due to Changes in Climate, Fuel Loads, and Wildfire Activity Over the Southeastern U.S.

An issue of great concern on federal lands is wildland fires, which have increased in frequency and strength over the past few decades as a possible consequence of climate change. Modeling wildfires under an evolving climate is challenging. There are disparate spatial and temporal scales involved in characterizing wild fire emissions and their effects on ambient air quality and visibility downwind, and in forecasting changes in vegetation and fuel loads in response to the changing climate and resulting changes in fire regimes. Many models altogether ignore these changes in future climate regimes, giving rise to large uncertainties in predicting future climate impacts on fires, air quality and compliance with the National Ambient Air Quality Standards (NAAQS). To address some of the issues underlying the reliable projection of fire emissions and air quality in an evolving climate, the University of North Carolina-Institute for the Environment, in collaboration with research scientists in the US Forest Service, is developing a modeling and analysis study focused on the Southeastern U.S., an area in which management of fire and air quality is already challenging today. Our modeling approach includes dynamic representation of the vegetation and associated changes in fuels, changes in the spatial distribution of daily fire activity during the fire season, and the resulting impacts on smoke emissions and air quality under contemporary conditions and future climate scenarios in high and low fire years.

Global premature mortality due to ozone and PM2.5 outdoor air pollution and the contribution of climate change

Raquel A. Silva, J. Jason West, Yuqiang Zhang, Susan C. Anenberg, Jean-Francois Lamarque, Drew T. Shindell, William J. Collins, Stig Dalsoren, Greg Faluvegi, Gerd Folberth, Larry W. Horowitz, Tatsuya Nagashima, Vaishali Naik, Steven Rumbold, Ragnhild Skeie, Kengo Sudo, Toshihiko Takemura, Daniel Bergmann, Philip Cameron-Smith, Irene Cionni, Ruth M. Doherty, Veronika Eyring, Beatrice Josse, I. A. MacKenzie, David Plummer, Mattia Righi, David S. Stevenson, Sarah Strode, Sophie Szopa, Guang Zeng

Outdoor air pollution is linked to premature human mortality. Current and future concentrations of ground-level ozone and fine particulate matter (PM_2.5) are affected by natural and anthropogenic emissions and climate change. Using modeled concentrations from an ensemble of chemistry-climate models, we estimate the present-day and future global burden of outdoor air pollution on premature human mortality, and the contribution of climate change to that burden. We use preindustrial (1850s), present-day (2000) and future (2030, 2050 and 2100) global ozone and PM_2.5 concentrations from the Atmospheric Chemistry and Climate Model Intercomparison Project (ACCMIP) simulations, which include future global greenhouse gas and air pollutant emissions as projected in the IPCC AR5 Representative Concentration Pathways (RCPs). The effect of climate change on air quality is isolated through simulations where 2000 emissions were used together with 1850 or future year climate. All model outputs are regridded to 0.5x0.5 horizontal resolution. For each model, the present-day burden is evaluated using the difference between air quality in 2000 and 1850, together with present-day population and baseline mortality rates and a health impact function. We estimate that 470,000 (95% confidence interval, 140,000 to 900,000) premature respiratory deaths are associated globally and annually with anthropogenic ozone, and 2.1 (1.3 to 3.0) million deaths with anthropogenic PM_2.5-related cardiopulmonary diseases (93%) and lung cancer (7%). The effects of past climate change have an estimated small contribution of 1,500 (-20,000 to 27,000) deaths yr^-1 due to ozone and 2,200 (-350,000 to 140,000) due to PM_2.5. The large uncertainties reflect the uncertainty in the concentration-response function and, particularly in the case of past climate change, the large variability among models. Estimates of future premature mortality will be evaluated for each future scenario and year relative to 2000, along with projections of future population and baseline mortality rates, and we will likewise estimate the potential contribution of future climate change.

Toward a chemical climatology of ozone contributions from long range transport in the Pacific Northwest -- Incorporation of ozone tracers in the AIRPACT-4 air quality forecast system

Joseph K. Vaughan*,

Serena H. Chung*,

Farren Herron-Thorpe*,

Rui Zhang*,

Brian K. Lamb*,

George H. Mount*.

* Laboratory for Atmospheric Research, Department of Civil and Environmental Engineering, Washington State University, Pullman, WA, 99164-2910

Air-quality modeling is an important tool for evaluating strategies for complying with the NAAQS. A perennial issue is the significance of long-range transport (LRT) effects on air quality. Under the EPA Exceptional Events Policy, for example, a nominal exceedance can be excluded from design value calculation if it can be credibly ascribed to long-range transport. Air-quality modeling is an appropriate tool for attempting demonstration of LRT in making a case for Exceptional Event status for an exceedance. Also, ample evidence exists that local air pollution should sometimes be viewed in the context of a baseline pollution levels, and that these baseline levels are influenced by LRT (Widger et al., 2013). AIRPACT4, a WRF-SMOKE-CMAQ air quality modeling system, uses chemical boundary conditions from global MOZART4 model runs that assimilate MOPPIT/TEYYA satellite CO (Herron-Thorpe et al., 2012). To develop a chemical climatology describing LRT contribution to the ozone background of the Pacific Northwest, we use a non-reactive tracer species version of CMAQv4.7.1 to trace the contribution of ozone originating on the western boundary of the AIRPACT4 domain, generally representing ozone input to the domain from trans-Pacific transport originating in Asia. Discrete tracers are assigned to the boundary condition ozone from each of the 21 model layers. The modeling results are analyzed to determine the ratio of tracer ozone to standard AIRPACT4 ozone in the lowest, ground level, model layer, resulting in monthly statistics. Preliminary results will be presented for ozone-season months, along with discussion of further refinement and application of this approach.

References:

Wigder, N. L., D. A. Jaffe, F. L. Herron-Thorpe, and J. K. Vaughan (2013), Influence of daily variations in baseline ozone on urban air quality in the United States Pacific Northwest, J. Geophys. Res. Atmos., 118, 3343-3354, doi:10.1029/2012JD018738.

Herron-Thorpe, Farren L., George H. Mount, Louisa K. Emmons, Brian K. Lamb, Serena H. Chung, and Joseph K. Vaughan, Regional Air-Quality Forecasting for the Pacific Northwest Using MOPITT/TEYYA Assimilated Carbon Monoxide MOZART-4 Forecasts as a Near Real-Time Boundary Condition, Atmospheric Chemistry and Physics, 12, 5603-5615, 2012. doi:10.5194/acp-12-5603-2012.

Investigation of multi-decadal trends in aerosol direct radiative effect from anthropogenic emission changes over North America by using a multiscale two-way coupled WRF-CMAQ model

Chao Wei, Jia Xing, David Wong, Jonathan Pleim, Rohit Mathur, Chuen-Meei Gan, ST Rao and Francis S. Binkowski

The anthropogenic aerosols play a dominant role in the "global dimming or brightening". However, aerosol radiative effects are still recognized as some of the largest sources of uncertainty among the forcers of climate change. This study will systematically investigate changes in anthropogenic emissions of short-lived aerosol-precursors over the past two decades (1990-2010) in the United States, their impacts on aerosol loading, and subsequent impacts on regional radiation budgets. The hypothesis that changes in surface solar radiation over time are caused by the changing patterns of anthropogenic emissions of aerosols and aerosol precursors will be tested in this study. A new two-way coupled meteorology and atmospheric chemistry model composed of the Weather Research and Forecast (WRF) model and the Community Multiscale Air Quality (CMAQ) model has been developed by U.S. Environmental Protection Agency. This two-way model is being run for 20 years (1990-2010) on both 12-km and 36-km resolution grids that cover most of North. A newly developed 20-years U. S. emission inventory is used in order to accurately reflect the emission trends resulting from progressively more stringent air quality regulations as well as population trends, economic conditions, and technology changes in motor vehicles and electric power generation. The direct effects of aerosols on SW radiation are considered in this WRF/CMAQ model. New algorithms on the calculation of aerosol optical properties and radiation have been developed for considering of both computational efficiency and more realistic aerosol states. Aerosol mixing state is a key factor for the calculations of aerosol optical properties. A more realistic core-shell model is included in this study. Weak nudging in atmosphere and strong nudging on soil temperature are used in our simulations in order to get a balance between strong signal of aerosol effects and good model performance. Preliminary model simulations for 1990-2010 are being evaluated both for their performance in comparison to observed concentrations and simulation of observed trends in concentrations and surface radiation.

Assessment of transboundary ozone contribution over East Asia using multiple source-receptor modeling techniques

 

Jung-Hun Woo¹, Ki-Chul Choi¹ , Cheol-Hee Kim², Soontae Kim³

 

 

¹Department of Advanced Technology Fusion, Konkuk University, Seoul, South Korea

²Department of Atmospheric Sciences, Pusan National University, Busan, South Korea

³Division of Environmental, Civil and Transportation Engineering, Ajou University, Suwon, South Korea

 

East Asia is one of the largest emission source regions in the world because of the large population and fast economic growth for several decades. A number of regional scale transport studies using comprehensive 3D modeling system such as CMAQ have been conducted to understand transboundary air pollution. The contribution assessment using such a comprehensive modeling system, however, was not extensively investigated in this region. Ozone concentrations in East Asia were simulated using the CMAQ model and its source contributions were estimated by multiple source-receptor modeling techniques. Three approaches (Brute-force, HDDM and OPTM) to study relationships between ozone concentrations and precursor emission sources were applied to four months (January, April, July and October 2009) to represent seasonal characteristics and compare the results. The approaches generally show that most of the receptor regions are strongly affected by central China emissions, which is the largest anthropogenic emissions source region over East Asia. A comparison study of multiple approaches to estimate monthly averaged ozone contributed by anthropogenic sources shows similar assessment results, however, OPTM generally provided higher estimates of the biogenic source contribution compared to HDDM. When comparing the BF method and HDDM, the sensitivity results show reasonably good agreement during the same period. The location and time dependent maximum 8-h ozone isopleths over South Korea as a receptor region created by HDDM suggest that each source region generally shows a NO_X limited regime.

Calibrating satellite-based fire emissions using prescribed-fire permit data and assessing/predicting fire impacts on air quality using CMAQ, FLEXPART, and statistical models

Chao Luo¹, Yuhang Wang¹, Dan Chan², Tao Zeng^1,3, Di Tian^1,3, Xiaoyang Zhang⁴

¹School of Earth and Atmospheric Science, Georgia Institute of Technology, Atlanta,

Georgia, 30332

²Georgia Forestry Commission, Macon, Georgia 31202

³Georgia Department of Natural Resources, Environmental Protection Division, Atlanta,

Georgia, 30354

⁴Earth Resources Technology, Inc., NOAA/NESDIS/Center for Satellite Applications

and Research, 5200 Auth Road, Camp Springs, MD 20746, USA

Prescribed burning is a large aerosol source in the southeastern United States in spring. We analyze fire emissions in March 2012 using Georgia fire permit data and the satellite observation based NOAA GOES Biomass Burning Emissions Product (BBEP) inventory. March was selected in the investigation because it is the month with most active prescribed fires. Comparison between GA fire permit and BBEP data shows reasonably good agreement. We then used the fire permit data to calibrate the BBEP fire emissions in the Southeast. The CMAQ model was used to study the impact of prescribed fires on aerosol concentrations. The inclusion of fire emissions significantly improved model performance. Model results showed that prescribed fire emissions led up to ~30% enhancements of monthly mean OC and PM₂₅ concentrations, and daily increase of PM2.5 up to ~30 ug m^-3 in some states of the Southeast. The FLEXPART-WRF model was used to study transport and dispersion of fire emitted PM₂₅ using the GA permitted fire data. The results are comparable to the CMAQ simulations. Multi-linear regression analysis was also applied to predict PM₂₅ aerosol enhancements using fire emissions and meteorological parameters, such as wind and boundary layer height. We compared the predicted PM₂₅ concentrations calculated by multi-linear regression to PM₂₅ enhancement simulated by CMAQ and the result is encouraging. This study shows that calibration of satellite-based fire emissions using the state fire permit data improves the quality of the fire emission estimates and provides a sound basis for air quality impact assessment using CMAQ simulations. The more computationally efficient simulations using the FLEXPART and statistical models are in reasonable agreement with CMAQ simulations. These methods provide the means to predict the potential impacts of prescribed fire burning requests and may be used operationally to facilitate the issuing of fire permits by state agencies.

Modeling Climate Change Effects on Spatiotemporal Distributions of Allergenic Ragweed (Ambrosia) and Mugwort (Artemisia) Pollen

Yong Zhang^1,2, Leonard Bielory³, Yang Gao⁴, Zhongyuan Mi¹, L. Ruby Leung⁵, Joshua Fu⁴, Panos G. Georgopoulos¹

¹Environmental and Occupational Health Sciences Institute (EOHSI), Rutgers University, 170 Frelinghuysen Road,Piscataway, NJ 08854, USA

²Department of Chemical and Biochemical Engineering, Rutgers University, 98 Brett Road, Piscataway, NJ 08854, USA

³Center for Environmental Prediction, Rutgers University, New Brunswick, NJ 08901, USA

⁴Department of Civil and Environmental Engineering, The University of Tennessee, Knoxville, TN 37996, USA

⁵Atmospheric Sciences and Global Change Division, Pacific Northwest National Laboratory, Richland, WA 99352, USA

Climate change is expected to alter the spatiotemporal dynamics of allergenic pollen, which can act synergistically with common air pollutants such as ozone to cause Allergic Airway Diseases (AAD). A comprehensive prognostic modeling system, combining climate models and anthropogenic and biogenic emission models with an expanded version of the Community Multiscale Air Quality (CMAQ) Model has been developed to support integrated studies of the impact of climate change on AAD. The present work has employed the components of this system to model the emission and spatiotemporal distributions of allergenic ragweed and mugwort pollen.

The pollen emission module was by incorporating major physical processes such as direct emission and re-suspension of pollen particles and considering effects of meteorological parameters such as ground surface temperature, friction velocity and humidity, etc. Season start and length of ragweed and mugwort pollen were simulated using Cooling Degree Hours (CDH). Daily and hourly flowering fractions of ragweed and mugwort plants were parameterized using the observed airborne pollen counts, phonology and relevant meteorology factors such as humidity, temperature and sunrise time. Vegetation coverage of ragweed and mugwort were derived using observed airborne pollen counts and Land Use Land Cover data from Biogenic Emissions Land use Database, version 3.1 (BELD3.1). Using the emission module, spatiotemporal distributions of pollen due to transport was simulated via the combined application of the Weather Research and Forecasting (WRF) model and an adapted version of the CMAQ model.

The optimum threshold CDH values for start date were 2,353 and 40,213 degree hours for ragweed and mugwort, respectively. The optimum initial date and base temperature for the CDH model of start date were found to be August 1st and 30°C for ragweed; May 1st and 37°C for mugwort. Ragweed and mugwort plants are distributed mainly in the western and central US. Simulation results indicate that responses of ragweed and mugwort pollen are expected to be heterogeneous across the continental US. The ragweed and mugwort pollen seasons tend to start later in the northern and central US while they start earlier in the southeastern US. The season length appears to be shorter in the north and longer in the south.

Keywords: Climate change, Airborne allergenic pollen, Ragweed, Mugwort

The co-benefits of GHG mitigation for air quality in the US

Yuqiang Zhang1, Jared Bowden2, Zac Adelman1,2, J. Jason West1

¹The University of North Carolina at Chapel Hill, Chapel Hill, North Carolina 27599

2Institute for the Environment, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA

Previous studies have shown that GHG mitigation strategies will benefit both global and regional air quality.  Here we propose to investigate the air quality benefits due to domestic GHGs mitigation in the US alone, compared with foreign contributions, at fine resolution.  We build off of our recent modeling study on global co-benefits, in which a global GHG mitigation scenario, the Representative Concentration Pathway Scenario 4.5 (RCP4.5), is compared with a reference case scenario (REF) to 2100, accounting for both changes in emissions and climate change induced by the GHG mitigation. Using these global MOZART-4 simulations, we aim to conduct meteorological and chemical downscaling to evaluate co-benefits at fine resolution within the US, and to isolate the co-benefits of domestic GHG reductions under current and future climate.

To achieve our goal, we will use the latest Weather Research and Forecasting (WRF) model as a Regional Climate Model (RCM) to dynamically downscale the GFDL AM3 Global Climate Model (GCM) over the US at 36 km resolution.  By doing this we will resolve the lack of resolution in the global GCM, capturing regional-scale meteorological changes like temperature and precipitation. WRF meteorology will then be used as inputs for the newest Community Multiscale Air Quality (CMAQ) modeling system. The MOZART-4 global simulation will also be used to prepare the initial conditions (IC) and dynamic boundary conditions (BC) for CMAQ. Anthropogenic emissions will be processed through SMOKE from REF and RCP4.5 scenarios to prepare temporally- and spatially-resolved emission files. Biogenic emissions will vary in accordance with the changing climate in the CMAQ model. 

Recent Updates to Visualization Environment for Rich Data Interpretation (VERDI)

Darin Del Vecchio1, Donna Schwede2, William Benjey2,Todd Plessel4, Liz Adams1

1 Center for Environmental Modeling for Policy Development, Institute for the Environment, University of North Carolina at Chapel Hill, Chapel Hill, NC

2 Atmospheric Modeling and Analysis Division, U.S. Environmental Protection Agency, Research Triangle Park, NC

3 Air Quality Analysis Division, U.S.Environmental Protection Agency, Research Triangle Park, NC
4 Lockheed Martin Information Technology, Research Triangle Park, NC

Visualization Environment for Rich Data Interpretation (VERDI), an open-source Java tool for visualizing the results from the Community Multiscale Air Quality Model (CMAQ) and associated programs, version 1.4.1 was released in May 2013.  Development of VERDI has continued and an overview of the new features is presented. VERDI is supported via the CMAS center and has a website (http://www.verdi-tool.org/) which provides information about the tool as well as downloads and documentation. A SourceForge repository (http://sourceforge.net/projects/verdi/) provides source code version control and distributions and a link to the developer documentation.  Contributions to the development of VERDI by the user community are encouraged.

Visibility Modeling in the Lower Fraser Valley, B.C

Bruce Ainslie, Robert Nissen, Paul Makar and Roxanne Vingarzan

Visibility has been identified as an important resource to manage in the Lower Fraser Valley (LFV),
British Columbia. The present management strategy includes measurements through a dedicated
visibility monitoring network with the aim to improve visibility over time. Policy work
looks to identify key emission sources and activities that both significantly
contribute to visibility degradation and which are feasible to control.
This policy work is being informed from the use of Environment Canada's photochemical model, AURAMS. We test the fitness of AURAMS for this policy work by exercising it over two summertime events. The first, in August 1993, occured during the Pacific '93 and REVEAL field campaigns, and has a good dataset of measured extinction data; and the second, during the summer of 2012, makes use of observations from our new visibility monitoring network. Over the 1993-2012 timeframe, local precursor emissions have dropped by about 40% - reductions that are roughly the same magnitude as those being considered for the policy work. Thus, evaluation of AURAMS over these two events provides a test for the model's ability to capture the sensitivity of visibility in the airshed to changing emission rates. In the talk, we will present modeling results from the 2 episodes and relate the model-predicted changes in visibility to findings from a local visibility perception study.

Improvements in modeled CMAQ wet deposition and ambient aerosol concentration due to mechanistic improvements in meteorology and emissions

J.O. Bash¹, K.M. Foley¹, R.L. Dennis¹, R.W. Pinder¹

[1] {Environmental Protection Agency, Research Triangle Park, NC, USA}

Recent process level improvements in meteorological dependent emissions and model estimated precipitation have been made available in CMAQ, WRF and the U.S. EPA National Emissions Inventory (NEI). These include the addition of lightning NO production and bidirectional NH₃ exchange, processes previously absent from CMAQ and other regional and global scale models, a mechanistically based diurnal emissions profile for animal NH₃ emissions, and the WRF Kain-Fritsch convective precipitation options. An incremental study was conducted to quantify the impact that these options have on modeled total nitrogen deposition and ambient aerosol concentrations. Model results are evaluated against monitoring network aerosol and wet deposition observations. Preliminary results indicate that; 1) the inclusion of NO produced from lightning reduces the annual modeled NO_x wet deposition bias, 2) the inclusion of bidirectional NH₃ exchange reduces the spring and fall PM_2.5 bias and the annual ammonium wet deposition bias, 3) the new animal NH₃ diurnal emissions profile reduces the modeled PM_2.5 bias and error in the fall, and 4) using the updated WRF Kain-Fritsch convective precipitation option reduces the bias in the summertime convective precipitation, thus reducing the summertime wet deposition bias.

We will first briefly introduce the processes and then quantify and evaluate the resulting changes in wet deposition and/or ambient PM_2.5 concentrations. Nonlinear interactions between these model options and the impact of including all of these processes on seasonal PM_2.5 concentrations and the total (wet and dry) nitrogen deposition will be discussed.

The evaluation of air quality forecasting system based on WRF-CMAQ and WRF-Chem over Houston during the DISCOVER-AQ Houston: surface O3, PM2.5 and tropospheric NO2

Lijun Diao¹, Yunsoo Choi¹, Beata Czader¹, Sunyeon Choi², Joanna Joiner², Hyuncheol Kim³

Two 4-km air quality forecasting systems based on WRF-CMAQ and WRF-Chem are established to forecast air quality over Houston during the 2013 DISCOVER-AQ Houston campaign. The simulated concentrations of O₃, NO_x, CO, HCHO, and PM2.5 between WRF-CMAQ and WRF-Chem are compared at four different altitude levels (surface, 800 hPa, 500 hPa, and 300 hPa), which give characteristics of two different forecasting systems. The WRF-Chem forecasting system uses chemical boundary condition using MOZART and CMAQ forecasting system utilizes chemical boundary condition using 12km CMAQ forecasting results. Both systems perform two-day forecasting every day during the project time period. The concentrations of surface O₃ and PM2.5 from two forecasting systems are compared with corresponding in-situ measurements at EPA AIRNow stations. Further, tropospheric NO₂ column from GOME-2 and/or OMI satellite, and lower tropospheric NO₂ from OMI satellite, derived using cloud-slicing method, will be compared with corresponding forecasting results.

Continuous, Near Real-Time CMAQ Model Simulations: A New Approach for Rapid and Robust Evaluation of the Modeling System

Brian Eder, Robert Gilliam, George Pouliot, Jesse Bash, Daiwen Kang, Shawn Roselle and Rohit Mathur

Traditionally, the Community Multi-scale Air Quality model has been evaluated using retrospective, often annual length simulations that represented conditions occurring many years prior.  While informative, such an approach often masks finer scale temporal (i.e. diurnal to weekly) and spatial (meso to synoptic) variability that often controls atmosphere processes and hence air quality.

It has become clear that newer approaches are necessary; approaches that will allow for more rapidly testing and hence more efficient evolution of the modeling system’s science.   Accordingly, the we have been running CMAQ continuously and in near real-time, allowing for immediate and ongoing analysis at finer spatial and temporal scales, thereby facilitating model evaluation and hence improve model performance.  Results from the simulations have been immediately examined and discussed by scientists in weekly meetings while antecedent meteorological and air quality conditions remain familiar. 

The advantages of such an approach, which will be presented,  have been numerous:

-           it has identified numerous model deficiencies including periodic inaccurate lateral boundary conditions, emissions inaccuracies related to wind-blown dust and an overprediction of PM_2.5 in the upper mid-west that was attributed to ammonia emissions that has since been corrected;

-           it will lead to improvements in characterizing episodic emission events such as wildfires;

-           the approach has also allowed scientist to provide guidance during an air  quality  field experiment (the Southern Oxidant and Aerosol Study).

Evaluate WRF PBL height using data derived from CALIPSO observations

Michael Ku, Winston Hao, Yonghua Wu, Barry Gross

The planetary boundary layer (PBL) height is an influential parameter on air quality modeling for ozone and PM_2.5. The PBL height affects the dispersion and transport of the pollutant that makes the verification of PBL height prediction from a meteorological model essential. However, it is difficult to assess the accuracy of the predicted PBL height due to sparse temporal and spatial PBL observations. One such area lacking observations is coastal areas where a sharp gradient of PBL height usually occurs in the transition from land to water. An improper prediction of PBL height in coastal areas would cause the air quality model to improperly simulate the dispersion and transport of the vast emissions from populated coastal areas.With recent advances in remote sensing capability, PBL height determination is possible by using data from Spaceborne Lidar, which covers vast and remote areas on a regular basis. In this study, the PBL heights derived from the Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) Level-2 5-km aerosol layer product will be used to evaluate the WRF predicted PBL height of 2007 for the eastern United States. The spatial distribution and variation of PBL heights from WRF and CALIPSO will be compared and discussed.

An Analytical Approach for Quantifying Emission-induced Uncertainties in Photochemical Air Quality Modelling

This study presents an analytical approach for quantifying uncertainties in photochemical air quality modelling outputs via error propagation. Uncertainties (i.e., standard deviations) in independent input variables and linear sensitivities of output variables to perturbations in input values are needed in the analytical approach. We demonstrate this method through a case study involving uncertainty analyses of modelling summertime ozone concentrations in the Mid-Atlantic region of the United States (U.S.). The U.S. Environmental Protection Agency’s (EPA) Community Multi-scale Air Quality Model (CMAQ) version 4.7.1 with Decoupled Direct Method-3D (DDM-3D) was used to simulate ambient ozone concentrations and their sensitivities to emission changes for the summer of 2007. The uncertain inputs had been divided into two categories: 1) volatile organic compounds (VOCs) emissions and 2) nitrogen oxides (NOx) emissions. We quantified uncertainties in CMAQ outputs attributed to errors in emissions from four modelling regions: Central Regional Air Planning Association (CENRAP), Lake Michigan Air Directors Consortium (LADCO), Ozone Transport Region (OTR), and Southeastern Modelling, Analysis, and Planning (SEMAP) regions. The results of the case study show that anthropogenic NOx emissions from non-electric generating unit (EGU) sources in the OTR region were the most important contributor to uncertainties in modeled peak ozone concentrations in the four ozone air quality non-attainment areas in the Mid-Atlantic U.S. 

Uinta Basin Winter Ozone Model Performance for the Utah Bureau of Land Managements Air Resource Management Strategy (ARMS) Modeling Study

Marco Rodriguez¹, Chao-Jung Chien¹, Kenneth Craig², Zion Wang¹, Courtney Taylor¹, Stephen Reid², Leonard Herr³

¹AECOM Inc.

²Sonoma Technology, Inc.

³Bureau of Land Management, Utah State Office

The Bureau of Land Management (BLM), Utah State Office, has initiated several studies focused on air quality in the Uinta Basin; one of these studies is the Air Resource Management Strategy (ARMS) Modeling Study. The Uinta Basin is an area in northeastern Utah that is projected to have continued development of oil and gas reserves in the foreseeable future. One of the main air quality concerns in the Uinta Basin is the elevated ozone levels measured during winter. Several winter episodes of elevated 8-hour ozone concentrations have been measured in the Uinta Basin since monitoring began in 2009. The United States Environmental Protection Agency's (USEPA) National Ambient Air Quality Standards (NAAQS) for the 8-hour average ozone concentration is 75 parts per billion (ppb). At a non-regulatory monitor in the Uinta Basin, the 8-hour average concentrations exceeded 120 ppb three times in late winter and early spring 2010. Episodes of elevated ozone concentrations in the Uinta Basin have been monitored in the late winter and early spring during multiple years since 2010.

Year 2010 was selected for the model performance evaluation based on data availability, including ambient ozone measurements, emissions inventories, and meteorological measurements. To produce a suitable modeling platform for the ARMS Modeling Study, the Uinta Basin oil and gas emissions inventory was refined and the Weather Research and Forecasting (WRF) meteorological model was optimized for the study area. These data were used in conjunction with other inputs required for the Community Multi-Scale Air Quality (CMAQ) Model and the Comprehensive Air Quality Model with Extensions (CAMx) Model. The CMAQ and CAMx results are compared and analyzed with available ambient measurements collected during January through March 2010.

Modeled versus Measured Ammonia: Source of discrepancy in the diurnal profiles at Rocky Mountain National Park

We investigate a discrepancy between summertime diurnal patterns of modeled ground level ammonia concentrations and ammonia measurements taken in Rocky Mountain National Park as part of the Rocky Mountain Atmospheric Nitrogen and Sulfur (RoMANS) II study. We run a year-long CAMx episode with 2009 meteorology and 2009 emissions inventories at 36, 12, and 4 km resolution, and find that modeled concentrations of ammonia at the core site in RMNP peak in late evening (on average around 11 P.M.) as evaluated using a 4 km nested grid domain. This nighttime peak of ammonia concentrations has been reported previously in studies when co-located measurement data was not available. In those cases, mixing height variability was often used to explain the result. However, in the case of RoMANS, hourly measurements are available and these measurements report a mid-day peak. Modeled daily average concentrations are within 6% of the measured values on average at the Rocky Mountain core site during the summer months (and within 15% annually), therefore showing good agreement for that metric. In contrast, chemical transport models utilizing different modeling episode inputs consistently predict nighttime maximum ammonia concentrations at Rocky Mountain National Park and in surrounding areas despite measurements that as consistently provide evidence of daytime maximum values. Source apportionment air quality modeling tools are used to identify the contributing ammonia sources to this site on an hourly basis. Process analysis is used to identify the contributions of all physical and chemical processes to ammonia concentrations also on an hourly basis. Using these data and information on the relative uncertainty of contributing factors, we design sensitivity runs to test the influence of select sources and processes on the modeled diurnal profile of ammonia with the goal of identifying areas that need further study in order to improve the model representation of ammonia.

Models reproducibility of ozone concentrations over Japan in warm season

Kazuyo YamajiKobe UniversityJapan Agency for Marine-Earth Science and Technology , Masayuki Takigawa (Japan Agency for Marine-Earth Science and Technology), Kohei Ikeda (Japan Agency for Marine-Earth Science and Technology), Yugo Kanaya (Japan Agency for Marine-Earth Science and Technology), Xiaole Pan (Japan Agency for Marine-Earth Science and Technology), Hiroshi Tanimoto (National Institute for Environmental Studies)

 

Air quality models such as CMAQ, recently, are expected to use for the national air quality measures in Japan. On the other hand, these models tend to overestimate tropospheric ozone by using an East Asian scale air quality model over Japan, especially in warm season. This study is comparing between observed and simulated ozone at remote observational sites over Japan and investigates the reasons for overestimations. This model tends to overestimate several percent of ozone inflow from outside of model domain. This study will discuss the other factors for the overestimation. 

Examining Projected Changes in Weather Extremes Between 2000 and 2030 using Dynamical Downscaling

Tanya L. Otte and Christopher G. Nolte, U.S. Environmental Protection Agency, Research Triangle Park, North Carolina

Climate change can alter regional weather extremes and affect environmental concerns such as the effects of changing energy demands and air quality on human health and preparing the American public for changes in the water supply, agriculture, and pest migration. Dynamical downscaling simulations were conducted with the Weather Research and Forecasting (WRF) Model for the periods 1995-2005 and 2025-2035 over North America. The WRF simulations were driven by the NASA-GISS ModelE2 simulations of contemporary climate and Representative Concentration Pathway (RCP) 6.0, respectively. The presentation will focus on the regional analysis of extreme daily temperature and precipitation events over each 11-year period and the projected changes in those extremes over that 30-year period.

RLINE: A Line Source Dispersion Model for Near-Surface Releases

Saravanan Arunachalam; Michelle G Snyder; Akula Venkatram; David Heist; Steven Perry; William Petersen; Vlad Isakov

Growing concern about human exposure and related adverse health effects near roadways initiated an effort by the U. S. Environmental Protection Agency to reexamine the dispersion of mobile source related pollutants. These adverse effects, in combination with the fact that a significant portion of the population spending time at home, work or school within a few hundred meters of major roadways, support a need for dispersion modeling to capture the temporal and spatial variability of mobile source pollutants in the near-road environment. The RLINE (Research LINE source) model is a research grade dispersion model that is currently under development by EPA ORD for near-roadway assessments. RLINE is a part of EPA ORD's ongoing evaluation of air quality impacts in the near-road environment. The research model is based upon a steady-state Gaussian formulation and is designed to simulate line type source emissions (e.g. mobile sources along roadways) by numerically integrating point source emissions. The current version of RLINE - RLINE 1.0 Beta - is currently formulated for near-surface releases, contains new (field study and wind tunnel based) formulations for the vertical and lateral dispersion rates, simulates low wind meander conditions, includes Monin-Obukhov similarity profiling of winds near the surface and selects plume-weighted winds for transport and dispersion calculations. The current beta version of the model is designed for flat roadways (no surrounding complexities) with future research efforts expected to accommodate new algorithms for simulating the near-source effects of complex roadway configurations (noise and vegetative barriers, depressed roadways, etc.). EPA released a beta version of RLINE during the summer of 2013 for testing purposes, and an updated version is expected to be released via CMAS in Fall. This talk will present an overview of RLINE, its features, I/O requirements, model evaluation performed to date, along with a list of ongoing applications.

Grid-scale indirect radiative forcing of climate due to aerosols over the northern hemisphere simulated by the integrated WRF-CMAQ model

Shaocai Yu, Kiran Alapaty, Jonathan Pleim, Rohit Mathur, David Wong, and Jia Xing

Atmospheric Modeling and Analysis Division,
National Exposure Research Lab, U.S. EPA, RTP, NC 27711

In this study, indirect aerosol effects on grid-scale clouds were implemented in the integrated WRF-CMAQ modeling system by including parameterizations for both cloud droplet and ice number concentrations calculated from the CMAQ-predicted aerosol particles. The resulting cloud droplet and ice number concentrations are provided to the Morrison double moment cloud microphysics scheme(mass and number concentrations for cloud water and ice) to estimate aerosol effects on cloud and ice optical depth and microphysical process rates for indirect aerosol radiative forcing (including first, second and glaciations indirect aerosol forcing). The cloud drop effective radius and cloud ice effective radius from the output of the Morrison cloud microphysics scheme are used in the YYTMG and CAM radiation schemes affecting the radiation fields. Evaluations of model performance on shortwave cloud forcing (SWCF), longwave cloud forcing (LWCF), cloud optical depth, cloud fraction, cloud effective radius, and PM2.5 are carried out over the northern hemisphere by comparing to satellite observation data such as CERES and MODIS and surface monitoring networks (AQS, IMPROVE, CASTNet, STN, and PRISM) over the continental U.S.

Modeling dispersion of traffic-related pollutants in the NEXUS health study

Michelle Snyder¹, Sarav Arunachalam², Vlad Isakov¹, Kevin Talgo², Alejandro Valencia², Brian Naess², David Heist¹, Stuart Batterman³, and CAAA⁴

¹U.S. EPA, Research Triangle Park, NC

² Institute for the Environment, The University of North Carolina, Chapel Hill, NC

³Environmental Health Sciences, School of Public Health, University of Michigan, Ann Arbor, MI

⁴Community Action Against Asthma

Dispersion modeling tools have traditionally provided critical information for air quality management decisions, but have been used recently to provide exposure estimates to support health studies. However, these models can be challenging to implement, particularly in near-road studies, due to the detailed input requirements necessary to estimate traffic-related emissions. Local-scale dispersion modeling is being applied in the Near-road Exposures to Urban air pollutant Study (NEXUS) as one method to estimate the exposure of asthmatic children living in Detroit, MI to traffic-generated pollutants. The study design includes determining if children with asthma living near major roadways with high traffic have greater health impacts associated with air pollutants than those living farther away, particularly for those living near roadways with high diesel traffic.

A major feature of the NEXUS study is an evaluation of different exposure metrics of traffic-related pollutants of varying complexity to determine their utility in examining associations with observed health effects. One of these metrics, modeled air quality concentrations, is examined in detail in this paper. Modeling the dispersion of traffic-generated pollutants requires detailed information including the geometry of the road network, traffic volume, temporal allocation factors, and fleet mixes. In addition, emission factors for various pollutants of interest are required as a function of ambient temperature, traffic speed, time of year, and vehicle class. These various inputs, assembled from a variety of sources, are used in combination with meteorological inputs to generate link-based emissions for use in dispersion modeling to estimate pollutant concentration levels due to traffic. In this paper, we discuss the modeling set-up, the combination of dispersion and emissions modeling to obtain primary pollutant concentrations, and the resulting traffic-related exposure metrics for study participants.

Assessment of aerosol effects on surface radiation in the north hemisphere using two-way WRF-CMAQ model

Jia Xing¹, Jonathan Pleim¹, Rohit Mathur¹, David Wong¹, George Pouliot¹, Christian Hogrefe¹, Chuen-Meei Gan¹, Chao Wei¹

¹ The U.S. Environmental Protection Agency, Research Triangle Park, NC 27711, USA

The anthropogenic aerosols play a dominant role in the "global dimming or brightening" (the decrease or increase in surface solar radiation). However, the simulations of Global Climate Models (GCMs) generally underestimate the decadal changes in surface solar radiation, compared to the observed "diming" and "brightening" trends during the 20th century. Therefore it is important to further reduce the uncertainties and to improve the model's ability of reproducing the decadal changes in surface radiation. A new two-way coupled meteorology and atmospheric chemistry model, i.e., Weather Research and Forecast (WRF) model coupled with the Community Multiscale Air Quality (CMAQ) model has been developed by U.S. Environmental Protection Agency. This model system can be applied as an integrated regional climate and chemistry model (RCCM) which is an important tool for downscaling future projections of global climate to higher resolution, and assessing the interactions between atmospheric chemistry and climate forcing and the effects of air pollutants on atmospheric radiation and secondary effects on meteorology and air concentrations. In this study, we extend the applicability of the two-way WRF-CMAQ model to hemispheric scales. Results of 20 years simulations in the north hemisphere from 1990-2010 where significant of change of emission and radiation was expected (e.g. dimming and brightening) will be presented and discussed. The meteorological inputs needed for WRF simulations were obtained from the NCEP/NCAR Reanalysis data with 2.5 degree spatial and 6-hour temporal resolution. The anthropogenic emissions were provided by EDGAR (Emission Database for Global Atmospheric Research) and biogenic VOC and lightning NOx emissions were obtained from GEIA (Global Emission Inventory Activity). The capability of hemispheric WRF-CMAQ model to represent the effects of radiatively active gasses and aerosol and to reproduce the observed changes in radiation attributable to changes in atmospheric aerosol levels will be examined through comparison with available long-term observations.

A Modeling Framework for Improved Characterization of Near-Road Air Quality at Fine Scales for Nationwide Exposure Assessment

Shih Ying Chang, Saravanan Arunachalam, Brian Naess, Kevin Talgo, Alejandro Valencia, Vlad Isakov, Brad Schultz and Ted Palma

Communities at the proximity of roadways are exposed to high levels of air pollution from automobile exhaust and are under potential risk of adverse health effects. To understand the relationship between air pollution and adverse health effects, exposure and risk assessment studies are needed which require the characterization of the near-road air quality. Furthermore, to locate the "hotspot" of the high-risk region, a fine scale modeling output is preferred. In this study, we develop a framework to model fine-scale near-road air quality in U.S. nationwide and demonstrate a prototype in three study areas-Detroit, Michigan, North Carolina and Portland, Maine - where there are ongoing exposure / risk assessment studies. The objective of this modeling framework is to a) Characterize near-road air quality at fine spatial resolution, and b) Explore feasibility of extending this to a national scale to support enhancing the National Air Toxics Assessment (NATA). RLINE, a new dispersion model used for modeling roadway emissions as line sources based upon a steady-state Gaussian formulation is used to model criteria air pollutants and mobile source air toxics (MSATs) at census block level. To prepare model-ready emission source data, we rely on county-level Motor Vehicle Emission Simulator 2010b (MOVES) outputs from the EPA. Since emissions factors from MOVES are sensitive to road / vehicle type and vehicle speed, traffic activity data as well as road-link data would be required to estimate actual emission rates. In this study, the traffic and road-link information is obtained from Federal Highway Administration's Freight Analysis Framework 3 (FAF3) which includes both activity and roadway link definition for the entire nation. Additionally, we estimate broad regional background at these locations using a space-time ordinary kriging (STOK) approach that uses AQS measurements from the region. We first apply the modeling framework to model near-road air quality in Detroit, MI and compare outputs with results from a previous study where similar method was implemented albeit using very detailed traffic activity and road link data. We then apply this framework to the two other study areas - North Carolina and Portland to characterize near-road air quality at fine spatial resolution. We will present results from applying this to the two study areas, and discuss feasibility of extending this to a national-scale assessment.

A Tracer Study to Assess Transport of Cruise Altitude Aircraft Emissions to the Surface at Continental and Hemispheric Scales

Lakshmi Pradeepa Vennam¹, Saravanan Arunachalam¹, B.H.Baek¹, Mohammad Omary¹, Rohit Mathur², William Vizuete¹

¹University of North Carolina, Chapel Hill, NC

²U.S. Environmental Protection Agency, RTP, NC

Aircraft emit multiple pollutants during their various modes of activity (landing and takeoff and cruise) and affect ambient air quality. The vertical and intercontinental transport of aircraft emissions during cruise activities needs to be accurately quantified to assess the impact of those emissions on surface air quality. To investigate this issue, we used a modeling framework using an application of CMAQ v5.0.1 over the Northern Hemisphere (CMAQ-NH) and Continental US (CMAQ-CONUS) and aircraft emissions during full-flight activities from the FAA's Aviation Environmental Design Tool (AEDT) for the year 2006. Considering NO_x as a passive tracer at cruise altitudes (~10 - 12km), we carried out CMAQ model simulations using different vertical advection schemes for continental and hemispheric domains at horizontal resolutions of 36-km and 108-km respectively. We zeroed out emissions from other layers of the model and boundary conditions to focus on the cruise altitude NO_x tracers. Further we also tagged the aircraft emissions from major continents in CMAQ-NH domain (U.S., Europe and East Asia) and studied the intercontinental transport of emissions from these domains to other regions of the world. We studied the processes responsible for the transport from cruise altitudes to the surface using CMAQ's Process Analysis tool. In this talk, we will present results from our assessment of the transport occurring near tropopause and upper troposphere particularly quasi-isentropic and cross-isentropic exchanges. Considering potential vorticity parameter, we will closely study the transport pathways of cruise emissions to the surface, and will also perform pollutant flux budget analysis to determine the influence of zonal and meridional circulation on the net flux of cruise altitude aircraft-related pollutants. We will present results from seasonal simulations using these two model applications and discuss potential impacts of these sources on surface air quality.

Estimating Regional Background Air Quality using Space/Time Ordinary Kriging to Support Exposure Studies

Alejandro Valencia, Saravanan Arunachalam, Yasuyuki Akita, Marc Serre, Valerie Garcia, Vlad Isakov

Use of Photochemical Grid Modeling to Quantify Ozone Impacts from Fires in Support of Exceptional Event Demonstrations

Kenneth Craig¹, Daniel Alrick¹, Yuan Du¹, Clinton MacDonald¹, Neil Wheeler¹, Tom Gross², Doug Watson²

¹Sonoma Technology, Inc., Petaluma, CA
²Kansas Department of Health and Environment

Each spring, ranchers and farmers in the Flint Hills region of Kansas burn approximately two million acres of grasslands during a six- to eight-week period. These annual burns have been a longstanding land management practice in the Flint Hills. However, air quality impacts on the public from smoke generated by these fires have gained more attention in recent years. In April 2011, widespread smoke from numerous fires in the Flint Hills and from other large fires in Texas and Mexico impacted air quality in Kansas metropolitan areas. The smoke coincided with several exceedances of the National Ambient Air Quality Standards (NAAQS) for ozone at air quality monitors in eastern Kansas. Fires from the Flint Hills were particularly extensive on April 6, 12, and 13.

Using a weight of evidence of findings from modeling and observational analyses, we established a clear causal relationship between the elevated ozone concentrations and emissions from prescribed burns in the Flint Hills and demonstrated that the NAAQS exceedances would not have occurred "but for" the Flint Hills fires. We evaluated synoptic meteorological data and trajectory output, analyzed historical ozone data on meteorologically similar days in April with no observed smoke impacts, and performed Community Multiscale Air Quality Model (CMAQ) simulations with and without emissions from Flint Hills fires. To support the CMAQ modeling analysis, we developed a fire emissions inventory for the Flint Hills using local county-level fire and fuels data, a refined spatial allocation approach, and the BlueSky Smoke Modeling Framework.

We found that the observed ozone concentrations in eastern Kansas on April 6, 12, and 13 were historically unusual, and established a clear causal relationship between the elevated ozone concentrations and emissions from the Flint Hills fires. The modeling analysis was key to establishing and quantifying this causal link. The smoke plumes generated in CMAQ using data from the Flint Hills fires matched observed smoke plumes reasonably well. The modeled impacts on peak 8-hr average ozone on the exceedance days ranged from 5 to 30 ppb at the Kansas monitoring locations, and up to 49 ppb within the modeling domain. Modeled impacts on daily average PM_2.5 were as high as 36 μg/m³ within the modeling domain. The Kansas Department of Health and Environment presented these findings to the U.S. Environmental Protection Agency in an exceptional event demonstration package.

Modeling as an exposure estimation approach for use in epidemiologic studies. Part 1: Guidance in choosing the appropriate model

Lisa K. Baxter, Kathie.L. Dionisio, Janet Burke, and Halik Ozkaynak

Many epidemiologic studies of the health effects of exposure to ambient air pollution use measurements from central-site monitors as their exposure metric. However, measurements from central-site monitors may lack spatial and temporal resolution required to capture exposure variability in a study population. This presentation will aim to explain how air quality modeling and exposure modeling can be used as exposure estimates to enhance health effect estimates obtained from an epidemiological study of the health effects of exposure to air pollution. Fundamental epidemiologic study designs will be described, so that the modeling community can be better informed when collaborating with epidemiologists. The advantages and disadvantages of air quality modeling vs. human exposure modeling, in the context of their use as exposure estimates in a health effects study, will be discussed. Topics that should be addressed when epidemiologists, exposure scientists, and modelers are planning an epidemiologic study will also be touched on. These include but are not limited to: the planned epidemiologic study design, the pollutant(s) to be studied, and the health outcome to be studied. Guidance will be provided that, together with the answers to these questions, will help inform researchers on what exposure estimation approach is best suited for the epidemiologic study in question, and will help modelers to guide epidemiologists to the modeling approach that will provide the most added value to the epidemiology study, but with the most efficient use of resources.

Changes in U.S. Regional-Scale Air Quality at 2030 Simulated Using RCP 6.0
Chris Nolte, Tanya Otte, Rob Pinder, Jared Bowden, Greg Faluvegi, and Drew Shindell

Recent improvements in air quality in the United States have been due to significant reductions in emissions of ozone and particulate matter (PM) precursors, and these downward emissions trends are expected to continue in the next few decades. To ensure that planned air quality regulations are robust under a range of possible future climates and to evaluate possible policy actions to mitigate climate change, it is important to characterize and understand the effects of climate change on air quality. Recent work by several research groups using global and regional models has demonstrated that there is a "climate penalty," in which climate change leads to increases in surface ozone levels in polluted continental regions. One approach to simulating future air quality at the regional scale is via dynamical downscaling, in which fields from a global climate model are used as input for a regional climate model, and these regional climate data are subsequently used for chemical transport modeling. In this work, regional climate simulations created by downscaling the NASA/GISS Model E2 global climate model are used as input for the Community Multiscale Air Quality (CMAQ) model. CMAQ simulations over the continental United States are conducted for two 11-year time slices, one representing current climate (1995-2005) and one following Representative Concentration Pathway 6.0 from 2025- 2035. Anthropogenic emissions of ozone and PM precursors are held constant at year 2006 levels for both the current and future periods. We examine the changes in ozone and PM concentrations, with particular focus on exceedances of the current U.S. air quality standards, and attempt to relate the changes in air quality to the projected changes in regional climate.

Modeling as an exposure estimation approach for use in epidemiologic studies. Part 2: Example applications

K.L. Dionisio, L.K. Baxter, V. Isakov, S.E. Sarnat, J.A. Sarnat, J. Burke, H. Ozkaynak

Recent studies have explored the use of modeling as an exposure estimation approach for use in epidemiologic studies of the health effects of exposure to air pollution. This presentation will provide examples from a case study in Atlanta, GA covering 169 ZIP codes, which analyzed the spatial and temporal variability introduced by modeling as compared to ambient measurements. The study explored the use of both air quality modeling (both local- and regional-scale) and human exposure modeling for use in a related epidemiology study. The spatiotemporal variability added from modeling as compared to ambient measurements will be demonstrated. The use of these alternative exposure estimates in a study of respiratory disease, asthma/wheeze, and cardiovascular disease will be described, to demonstrate the differences in health effect estimates and related confidence intervals when exposure estimates with different levels of spatiotemporal refinement are used.

Simulating the Impacts of Marine Organic Emissions on Global Atmospheric Chemistry and Climate using an Online-Coupled Meteorology and Chemistry Model

Brett Gantt, Timothy Glotfelty, Nicholas Meskhidze, and Yang Zhang

Department of Marine, Earth, and Atmospheric Sciences

North Carolina State University, Raleigh, NC 27695

The cloud droplet number concentration (CDNC) in remote marine regions has been shown to be very sensitive to changes in aerosol number concentration. This sensitivity leads to a strong dependence of the model-predicted aerosol indirect effects on the prescribed background aerosol number concentration. Despite the importance of background aerosols to climate predictions, most models either overlook or have incomplete emission inventories for number concentration and chemical composition of natural marine aerosols. Currently, most models include emission of sea-salt particles and dimethyl sulfide, an aerosol precursor gas. Two major aerosol sources neglected from most models are ocean-derived biogenic volatile organic compounds (VOCs) such as isoprene and primary organic aerosol (POA) associated with sea spray, which have been shown to contribute to the surface aerosol mass concentration and cloud condensation nuclei (CCN) in remote marine regions. The global-through-urban Weather Research and Forecasting model with chemistry (GU-WRF/Chem) is an online coupled meteorology and chemistry to simulate air quality and its interactions with meteorology and climate, but it does not include marine organic emissions. In this work, we implement the online emissions of ocean-derived isoprene and size-resolved marine POA in GU-WRF/Chem and apply it to realistically simulate the impact of marine isoprene and POA on chemistry-aerosol-cloud-radiation-precipitation-climate interactions.

The GU-WRF/Chem simulations show that the net effect of marine organic emissions is an increase in the surface mass concentration of organic aerosols, total aerosol number concentration, aerosol mean diameter, and CCN in remote marine regions. Increased isoprene concentrations associated with marine emissions decrease hydroxyl radical levels in the marine boundary layer and increase ozone concentrations in VOC-limited regions and decrease ozone in nitrogen oxide-limited regions. With the inclusion of ocean-derived VOC and POA emissions, GU-WRF/Chem predictions are in better agreement with surface observations of marine isoprene and organic aerosol concentrations and the aerosol number size distribution in remote marine regions. Increased organic aerosol concentrations lead to subsequent increases in global average CCN/CDNC and decreased net solar radiation at the surface. Consistent impacts from the inclusion of marine organic emissions in GU-WRF/Chem and another online-coupled model (Community Atmosphere Model version 5) suggest (with relatively high confidence) that these natural aerosols impact model predictions of anthropogenic aerosol indirect forcing on a global scale.

Comparing Exposure Metrics for the Effects of Fine Particulate Matter on Emergency Hospital Admissions

Elizabeth Mannshardt, Katarina Sucic, Wan Jiao, Francesca Dominici, H. Christopher Frey, Brian Reich, and Montserrat Fuentes

A crucial step in an epidemiological study of the effects of air pollution is to accurately quantify exposure of the population. In this paper, we investigate the sensitivity of the health effects estimates associated with short-term exposure to fine particulate matter with respect to three potential metrics for daily exposure: ambient monitor data, estimated values from a deterministic atmospheric chemistry model, and stochastic daily average human exposure simulation output. Each of these metrics has strengths and weaknesses when estimating the association between daily changes in ambient exposure to fine particulate matter and daily emergency hospital admissions. Monitor data is readily available, but is incomplete over space and time. The atmospheric chemistry model output is spatially and temporally complete, but may be less accurate than monitor data. The stochastic human exposure estimates account for human activity patterns and variability in pollutant concentration across microenvironments, but requires extensive input information and computation time. To compare these metrics, we consider a case study of the association between fine particulate matter and emergency hospital admissions for respiratory cases for the Medicare population across three counties in New York. Of particular interest is to quantify the impact and/or benefit to using the stochastic human exposure output to measure ambient exposure to fine particulate matter. Results indicate that the stochastic human exposure simulation output indicates approximately the same increase in relative risk associated with emergency admissions as using a chemistry model or monitoring data as exposure metrics. However, the stochastic human exposure simulation output and the atmospheric chemistry model both bring additional information which helps to reduce the uncertainly in our estimated risk.

Study of regional extreme climate and its impact on air quality in US

Estimation of human exposure to PM2.5 components in metropolitan areas across the U.S. using network measurements and a chemical transport model

Prakash V. Bhave, Mary K. McCabe, Valerie C. Garcia

Future prediction of tropospheric ozone over South and East Asia in 2030

S. Chatani¹, M. Amann², A. Goel³, J. Hao⁴, Z. Klimont², A. Kumar³, A. Mishra³, S. Sharma³, S. X. Wang⁴, Y. X. Wang⁴ and B. Zhao⁴

¹ Toyota Central R&D Labs., Inc., Nagakute, Japan

² International Institute for Applied Systems Analysis, Laxenburg, Austria

³ The Energy and Resources Institute, New Delhi, India

⁴ Tsinghua University, Beijing, China

We have developed a regional air quality simulation framework including WRF, CMAQ, and emissions to simulate tropospheric ozone over South and East Asia. Simulated ozone and related species were validated by comparing with observation data of surface monitorings, ozone zondes, and satellites. The simulation showed acceptable performance on ozone over South and East Asia in a regional scale. We predicted future energy consumption, CO2, NOx and VOC emissions under three scenarios in 2030. One of them assumed a business-as-usual (BAU) pathway, and other two considered additional energy and environmental strategies to reduce energy consumption, CO2, and pollutant emissions in China and India. Future ozone under these three scenarios was predicted. The simulation indicated future ozone is significantly increased around India for a whole year, and around northeastern China in summer. NOx was a main driver to increase ozone, but VOC was also important on annual average of ozone in East Asia. Warmer weather around India was also preferable in ozone formation. Additional energy and environmental strategies were effective to reduce future ozone over South and East Asia. It is desired that such starategies will be implemented to reduce ozone as well as energy consumption and CO2 emission.

Including marginal health damage information in air quality decision-making

S. Morteza Mesbah¹, Amanda Pappin¹, Amir Hakami¹, Stephan Schott²

¹Department of Civil and Environmental Engineering, Carleton University

²School of Public Policy and Administration, Carleton University

The marginal damage (MD) or Benefits Per Ton (BPT) of emissions refers to the dollar value of damage (or benefit) to the environment and human health caused by a unit increase (or decrease) in emissions. Because of the nature of NO_x and its contribution to the formation of secondary pollutants, ozone and particulate matter (PM), the MD of NO_x emissions varies both temporally and spatially. Calculation of source-specific MDs is important in policy design as MDs can be used for prioritizing emission reductions at different locations and times.

We discuss various methods through which MD information can be integrated into the decision-making process. To calculate source-specific MDs of NO_x emissions, we use the adjoint of the gas-phase CMAQ. The adjoint cost function is defined as a health damage function based on ozone (and/or PM) exposure. We calculate source-specific MDs of NOx emissions for different baseline emissions and different sectors (i.e., mobile sources and point sources). We perform multiple adjoint runs in a North American domain to evaluate the nonlinearity in total and marginal damage estimations. We find that MDs vary significantly both temporally and spatially, and they also change with baseline emissions. Furthermore, we find that inclusion of such MDs in policy and economic instruments can significantly enhance the performance of NO_x control policies. Finally, we show that the specific behavior of NO_x marginal damage curves differ from the traditional view held in air pollution economics, and discuss how these peculiarities in NO_x MD can affect the decision-making process.

Impact of biomass burning aerosols on regional climate over Southeast US

Peng Liu, Yongtao Hu, Athanasios Nenes, Armistead Russell

Georgia Institute of Technology

Great interest has been aroused in how important aerosols may play a role in the cooling of the southeastern USA. A major contributor to the particulate matters over this region is biomass burning, which is rich in black carbon and organic compounds, and may have significant feedbacks to regional climate through the direct and indirect effects.

In this study, the cloud droplet activation parameterization of Kumar et al.(2009), which considers the competition between soluble and in soluble aerosols for water vapor during cloud droplet formation in ascending air parcels, is implemented in coupled WRF-CMAQ. The water uptake properties of the biomass burning aerosol (required for predicting optical depth for direct radiative forcing, and CCN activity for indirect effects) are constrained using observations of fresh and aged biomass burning aerosol sampled during the 2008 ARctas campaign (Jacob et al., 2010). In order to isolate the direct and indirect effects, we first couple the aerosol module only with radiation module to estimate direct effect, and then with microphysics module for the indirect effect.

References:

Kumar, P., Sokolik, I.N., and Nenes, A. (2009) Parameterization of Cloud Droplet Formation for Global and Regional models: Including Adsorption Activation from Insoluble CCN., Atmos. Chem. Phys., 9, 2517-2532

Jacob, D. J., Crawford, J. H., Maring, H., Clarke, A. D., Dibb,J. E., Emmons, L. K., Ferrare, R. A., Hostetler, C. A., Russell, P. B., Singh, H. B., Thompson, A. M., Shaw, G. E., McCauley, E., Pederson, J. R., and Fisher, J. A. (2010): The Arctic Research of the Composition of the Troposphere from Aircraft and Satellites (ARCTAS) mission: design, execution, and first results, Atmos. Chem. Phys., 10, 5191-5212, doi:10.5194/acp-10-5191-2010

Winter Ozone Formation Sensitivity to Surface Albedo, Heterogeneous Chemistry and Precursor Emissions

Gail Tonnesen, Kirk Baker, Golam Sarwar, U.S. Environmental Protection Agency

Bernhard Rappenglueck, Department of Earth and Atmospheric Sciences, University of Houston

High concentrations of ambient ozone in winter have been observed in the Upper Green Basin in southwest Wyoming and the Uintah Basin in northeast Utah, with maximum 1-hour and 8-hour average ozone concentrations exceeding 160 ppb and 140 ppb, respectively. Several field studies have shown that winter ozone episodes are associated with: 1) emissions of VOC and NO_x from oil and gas production; 2) snow cover and increased surface albedo; and 3)stagnant winds, cold temperatures and strong, persistent cold pool inversion conditions. Additionally, extremely high daytime concentrations of nitrous acid (HONO) have been observed in both the Upper Green Basin and Uintah Basin, with peak daytime HONO values exceeding 3 ppbv. In contrast to summertime ozone episodes in which high HONO occurs at night, peak HONO concentrations during winter ozone episodes were observed from late morning to noon. We performed a base case CMAQ model simulation for February 2011 for a domain that includes both basins, and found that the model significantly underestimated ambient ozone levels with negative bias in the range of 60 to 80% and model simulated ozone levels at or below regional background ozone levels. A diagnostic analysis of the base case model identified three causes for poor ozone performance. The CMAQ version 5.0 inline photolysis rate algorithm underestimated surface albedo and had limited enhancement of photolysis rates in areas with high snow cover. CMAQ did not reproduce the observed high daytime HONO concentrations. CMAQ was also biased low for VOC precursors with largest negative bias in the Upper Green Basin. We performed model sensitivity experiments to evaluate the effects of uncertainty for each of these processes. We modified the inline photolysis rate calculation to set surface albedo as 90% of the fractional snow cover in each grid cell. We added a model HONO source in the surface layer proportional to the model nitric acid concentration to represent a possible heterogeneous source of HONO in the snow. We also performed a model sensitivity simulation with a factor of five increase in VOC emissions from oil and gas production. We found that increasing the surface albedo and adding a heterogeneous HONO source resulted in large increases in model ozone in the Uintah Basin, with maximum ozone increases above the base case of 34 ppb and 30 ppb, respectively. These sensitivity experiments resulted in much smaller increases in ozone in the Upper Green Basin, consistent with the larger negative bias for precursor concentrations in that basin. The factor of five increase in VOC emissions in the oil and gas sector resulted in significant increases in modeled ozone in both basins. While the sensitivity study reproduced the observed HONO concentrations, the model sensitivity experiments used in this study probably do not accurately characterize the presumed heterogeneous HONO source or the VOC emissions inventory, and these model experiments were designed only to estimate model sensitivity to these parameters. There is large uncertainty in both the presumed heterogeneous HONO source and VOC emissions inventories, and additional research is required to represent these processes accurately in photochemical air quality models. The analysis of the CMAQ inline photolysis rate found that CMAQ does not accurately treat surface albedo in areas with extensive snow cover, and the algorithm should be modified if CMAQ is used to simulate air quality in areas with snow cover.

UPDATE ON A NEW PARALLEL SPARSE CHEMISTRY SOLVER FOR CMAQ

George Delic, HiPERiSM Consulting, LLC, P.O. Box 569, Chapel Hill, NC 27514

This report is an update on the report at CMAS in 2012 [1] on the new CMAQ sparse solver (FSparse) that has been implemented in the Rosenbrock algorithm to replace the JSparse sparse solver method based on the work of Jacobson and Turco [2]. The new FSparse solver is a Fortran implementation of Gaussian Elimination procedures from the CSparse library of Davis [3]. This update extends the previous report to include comparison of higher thread count parallel results with the serial results of the standard U.S. EPA 4.7 version of CMAQ. Preliminary indications are that with 6 threads the FSparse version of CMAQ is 1.3 times faster than the standard release of CMAQ using the Portland Group compiler. The FSparse version with 6 threads completes in 20.8 hours versus 27.0 hours for the EPA version in a 24-hour simulation for a domain of 2.3 million grid cells. The detailed analysis reveals that that there are numerical differences in progress of times step iterations in the chemistry solver for the two methods. These differences result in FSparse completing significantly more chemistry time step iterations than does JSparse. This difference is related to the numerical precision achieved in the two methods and the way this affects the time step increment calculation. This time step calculation is based on the Rosenbrock residual error estimate in each iteration. Thus it is important to determine which of the two methods produces the greater accuracy. In this investigation a side benefit of FSparse is that various norms may easily be computed as the computation progresses to monitor numerical stability whereas such estimates are more difficult to do in the JSparse method. Nevertheless, this is the subject of current investigation. If FSparse is more accurate then relaxation of the time step criterion could result in even greater reduction of wall clock time.

[1] G. Delic, 11^th Annual CMAS conference, October 15-17, 2012.

[2] M. Jacobson and R.P. Turco (1994), Atmos. Environ. 28, 273-284

[3] T.A. Davis, Direct Methods for Sparse Linear Systems, SIAM, Philadelphia, 2006.

High-resolution meteorological simulations over a metropolitan area with and without an urban canopy model

Jared H. Bowden

Sarav Arunachalam

The WRF model is used to simulate the meteorology at a 1-km horizontal resolution over the Dallas-Fort Worth area for a winter and summer time week to explore the benefits of using urban morphology data for air quality simulations. Simulations were performed comparing traditional meteorological techniques with different land surface models (PX and NOAH) relative to the Multi-level Building Environmental Parameterization urban canopy model in WRF. The WRF urban canopy model uses urban morphology data from the National Urban Data Access Portal Tool (NUDAPT). Relevant air quality parameters including PBL height, wind speed and direction are compared with observations and across simulations to demonstrate model differences.

Study the non-linear response of ozone concentration in Taiwan using CMAQ-HDDM

Fang-Yi Cheng, Meng-Ching Wu, Ka-Wa Chan, Soontae Kim

Elevated ozone (O₃) concentration has been an important environmental issue in Taiwan over the past decades. In order to understand the source contributions and to build an effective emission control strategy to reduce Taiwan's high O₃ concentrations, we applied CMAQ-HDDM to study the non-linear responses of the O₃ concentration with respect to its precursors, nitrogen oxides (NO_x) and volatile organic compounds (VOC_s) emission rates. The objective is (1) to investigate O₃ sensitivities with respect to its emissions; (2) to quantify the source contributions of the high O₃ concentrations; (3) to identify the source-receptor relationships in the Taiwan area.

The modeling was conducted for a high O₃ event (October 17~25, 2011) in Taiwan. The CMAQ-HDDM was run with the Taiwan Emission Data System version 7.0 (TEDS 7.0) emission inventory and WRF 3.4.1 meteorological simulation results.

The preliminary results show that the maximum O₃ concentration occurs in different locations dependent on atmospheric conditions. O₃ isopleths constructed by CMAQ-HDDM sensitivities demonstrate that near the emission source regions most O₃ peaks reveal a NO_x-VOC transitional condition; however, downwind areas exhibit mostly NO_x-limited condition. Sometimes maximum O₃ concentration occurs in the downwind areas due to the transport of emission source coming from the upwind side. The result from this study would improve our understanding of O₃ formation in the area of Taiwan.

Dynamical Evaluation of Model Suitability for a Retrospective Analysis of Ozone Formation

The Lower Fraser Valley at the West Coast of British Columbia, Canada, frequently experiences high ozone concentrations during summer fair weather conditions. Onshore winds and the interaction of nitrogen oxide emissions from vehicle traffic with biogenic and anthropogenic emissions of volatile organic compounds lead to high ozone concentrations in the eastern part of the valley. More than two decades of substantial reductions of precursor emissions lead to a decline of ozone concentrations in the western half of the valley but no significant change in the eastern half. We performed a retrospective study with a WRF-SMOKE-CMAQ modeling system for several ozone episodes over two decades and different local circulation patterns and evaluated the ability of the modeling system to reproduce the observations. It proved useful for diagnostic purposes, for example the identification of sensitivity regime changes and the general distribution of the ozone field. The modeling system was capable of reproducing the temporal response of ozone to the observed reduction in precursor emissions and provided insights into the causes and mechanisms for its inhomogeneous (west-east) spatial response.

Updates to CAMx Plume-in-Grid Model to Improve NOy Chemical Processing and Computational Speed

Chris Emery, Bonyoung Koo, Tan Sakulyanontvittaya, Greg Yarwood

ENVIRON International Corporation

The CAMx photochemical grid model includes a Lagrangian Plume-in-Grid (PiG) sub-model to treat the physical and chemical evolution of point source plumes prior to reaching grid scale. CAMx offers two options to treat plume chemistry: GREASD PiG treats only the early stages of NOx-rich plumes when oxidant production is suppressed; IRON PiG is used to simulate plumes in which oxidant chemistry is active. The first option limits the lifetime of puffs to the inorganic chemistry stage, after which the puff mass is passed to the grid model; no such lifetime constraints are in place for the second option. Both PiG options employ a full representation of photochemistry (such as Carbon Bond), and both employ an "incremental chemistry" approach that solves chemical evolution of puff concentrations relative to their background, i.e., the grid concentrations. Incremental chemistry approaches are computationally expensive and we have found that they can lead to chemical instabilities such as nitrogen non-conservation. Plume processing of NOx emissions to less reactive forms of NOy has important effects on downwind ozone and PM distributions, and is intricately tied to interactions among primary emissions, ambient oxidant conditions, and plume dispersion rates. We describe revisions to the GREASD PiG that include a chemical mechanism with only 25 reactions that describe NOy processing and sulfate formation to improve speed, holding constant the puff background concentrations during each PiG time step to improve chemical stability, and revising nighttime plume dispersion rates to be consistent with in-situ measurements. Test results will be presented for two historical ozone episodes in Texas to assess how the GREASD PiG updates change model predictions and computational speed. Comparisons to ozone and NOy measurements from aircraft and surface monitoring sites will be presented to assess model performance.

The severe haze over the Southern Hebei area, China in 2013: what we learn from modeling and observation

Litao Wang, Pu Zhang, Xiujuan Zhao, Jing Yang, Zhe Wei, Jie Su, Fenfen Zhang, Chenchen Meng

Department of Environmental Engineering, Hebei University of Engineering, Handan, Hebei 056038, China

Hebei has been one of the most air polluted provinces in China. In Jan. 2013, continuous, very severe haze pollution happened in east and central China. In Beijing, only five days were not fog and haze days during Jan. 2013. It is reported that the daily fine particulate matter (PM_2.5) concentrations in Beijing and Shijiazhuang, the capital of Hebei Province, has been over 500 μg m^-3, which is 6.7 times of the new China National Ambient Air Quality Standard (CNAAQS). In the statistics of the Ministry of Environmental Protection of China (MEP), during this month, the ten most polluted cities are Xingtai, Shijiazhuang, Baoding, Handan, Langfang, Hengshui, Jinan, Tangshan, Beijing and Zhengzhou city, out of the reported 74 key cities all over China. Seven of the top ten cities are within Hebei Province and five of them locate in the southern area of Hebei. In this paper, we pursued a modeling study on the North China for Jan. 2013 using the MM5-CMAQ modeling system, to understand the severe haze pollution over the Southern Hebei cities. The observations of PM_2.5, PM₁₀, BC, SO₂, NO_x concentrations and the visibility obtained from a monitoring site at the Handan city, the most southern city in Hebei, were analyzed and discussed as well to further understand the develop of the severe haze episode.

Lagrangian Photochemical Modeling of Ozone Formation and Aerosol Evolution in Biomass Burning Plumes: Toward a Sub-grid Scale Parameterization

M. J. Alvarado¹, R. J. Yokelson², S. K. Akagi², E. Fischer³, K. Travis⁴, T. Soni⁵, J. S. Craven⁶, J. W. Taylor⁷, G. R. McMeeking³, I. R. Burling², S. P. Urbanski⁴, C. E. Wold⁸, J. H. Seinfeld⁶, H. Coe⁷, and D. R. Weise⁹

¹Atmospheric and Environmental Research (AER)

²University of Montana, Department of Chemistry

³Colorado State University, Department of Atmospheric Science

⁴Harvard University, School of Enginerring and Applied Science

⁵Massachusetts Institute of Technology, Department of Environmental Engineering

⁶California Institute of Technology, Division of Chemistry and Chemical Engineering

⁷University of Manchesxter, Centre for Atmospheric Science

⁸United States Forest Service, Fire Sciences Laboratory

⁹United States Forest Service, Pacific Southwest Research Station, Forest Fire Laboratory

Biomass burning is a major source of atmospheric trace gases and particles that impact air quality at urban, regional, and global scales. Within minutes after emission, rapid, complex photochemistry within a smoke plume can cause large changes in the concentrations of ozone and fine particles (PM2.5). Being able to understand and simulate this rapid chemical evolution under a wide variety of conditions is thus a critical part of forecasting the impact of these fires on urban and regional air quality. The Aerosol Simulation Program (ASP) has been previously used within a Lagrangian parcel model to simulate the formation of ozone and secondary organic aerosol (SOA) within several African and North American plumes. In this work, we will present ASP simulations of the chemical evolution of a young biomass burning smoke plume sampled over California during the 2009 San Luis Obispo Biomass Burning campaign. We will discuss the sensitivity of the model simulations to uncertainties in the emissions, dilution rate, and gas- and particle-phase chemistry. We will then present our initial work in using the ASP model to develop a sub-grid scale parameterization of the near-source chemistry of biomass burning plumes for use in regional and global air quality models.

CMAQ validation of optical parameters and PM10 concentrations based on LIDAR experimental campaign in the Metropolitan Area of Vitria - Brazil

Taciana T. de A. Albuquerque¹, Uma Shankar², Erick G. S. Nascimento¹, Fabio Lopes³, Gregori Moreira³, Renato M. Sartario⁴, Nadir Salvador¹, Ayres Loriato¹, Alexandre Magalhies¹, Eduardo Landulfo³, Gehard Held⁵, Neyval C. Reis Jr.¹, Davidson M. Moreira¹.

1. Environmental Engineering, Federal University of Espirito Santo, Vitoria, ES, Brazil.
2. University of North Carolina-Institute for the Environment, Chapel Hill, NC, USA.
3. Nuclear and Energy Research Institute (IPEN / USP), Sao Paulo, Sao Paulo, Brazil.
4. Geography Department, Federal University of Espirito Santo, Vitoria, ES, Brazil.
5. Meteorological Research Institute (IPMet / UNESP), Bauru, Sao Paulo, Brazil.

taciana.albuquerque.ufes@gmail.com

This study aims to evaluate the PM10 levels and optical parameters in the Metropolitan Area of Vitoria (MAV), Esperito Santo State, Brazil, using the LIDAR, SODAR, and satellite measurements to evaluate CMAQ model prediction. For the first time ever in the MAV, a ground-based LIDAR was deployed for one week to enable the assessment of the planetary boundary layer height, classification of atmospheric constituents, particulate matter concentration and sources. The LIDAR (Light Detection and Ranging) was co-located with a SODAR (Sound Detection and Ranging) to simultaneously observe the vertical distribution of wind (u, v and w components) up to 800 m. The campaign lasted for one week from 23 - 30 July 2012 with approximately 72 hours worth of LIDAR measurements. The SODAR was operating continuously. Vitoria is the home of two of the largest shipping ports in Brazil: the Port of Vitoria and the Port of Tubarao. The latter is owned by the company Vale and is primarily used to export iron ore. Also located next to the Port of Tubarao is the Arcelor-Mittal Tubarao steel manufacturing plant. The LIDAR and SODAR were located on the Federal University of Espirito Santo Campus (UFES), with the weather station at the airport being is directly east of it. The results of the instrumental campaign show a regular occurrence of the sea breeze circulation during high pressure synoptic conditions with light synoptic winds and clear skies. High PM10 concentrations were found mainly in the first days of the campaign because of the weather conditions. In July 23 was observed by LIDAR a plume over the MAV around 2km in altitude. The HYSPLIT model results show a backward trajectory from the central area of Brazil indicating that it is arriving in Vitoria region.

Enhanced Approach to Model Air Quality Impacts of Aircraft Operations in and around an Airport for Surface Movement Optimization Research

Saravanan Arunachalam¹, Jared Bowden¹, Elizabeth Adams¹, Mohammad Omary¹, Alejandro Valencia¹,

Prakash Karamchandani², Brian Kim³, William Chan⁴

¹Institute for the Environment

University of North Carolina at Chapel Hill, NC

²Environ, Novato, CA

³Wyle Laboratories, Inc., Atlanta, GA

⁴Aviation Systems Division,

National Aeronautics and Space Administration, Ames, CA

Aircraft emissions represent a non-negligible source of the total emissions in the vicinity of an airport that affects ambient air quality - specifically O3 and PM. As aircraft operations continue to grow and the impact on local air quality is increasingly scrutinized, the advancement of tools to assess and reduce emissions from aircraft movements becomes progressively necessary. It is envisioned that such tools would allow users to analyze various what-if surface scenarios to reduce the burden on local air quality. To support NASA's continued research on airport surface movement optimization, air quality modeling tools are beneficial to predict the environmental impact of airports on local and regional air quality. To accomplish this, we developed a comprehensive framework using the WRF-SMOKE-CMAQ at a nested resolution of 12/4/1-km for the Dallas Fort-Worth (DFW) region for a summer and winter episode. A detailed airport-level emissions inventory is developed using the EDMS modeling system. We augment the framework by using urban morphology around the airport region from the National Urban Data and Access Portal Tool, and its implementation in WRF called WRF-NUDAPT for the 1-km simulation. We compare and contrast assessing the airport impacts on ambient air quality using WRF/CMAQ application at 12/4/1-km versus WRF/CMAQ enhanced for subgrid scale treatment using the Advanced Plume Treatment (APT) module, but only at 12/4-km resolution. We will present results from these two different modeling approaches for the DFW region, and present pros and cons with a focus on practical implications for coupling these tools with the surface movement optimization algorithms.

Application of WRF/Chem over the continental U.S. under the AQMEII Phase II: Comprehensive Evaluation and Chemistry Feedbacks to Meteorology
Khairunnisa Yahya1, Kai Wang1, Masilin Gudoshava1, Shiang-Yuh Wu2, Timothy Glotfelty1, and Yang Zhang1

1Department of Marine, Earth, and Atmospheric Sciences, North Carolina State University, Raleigh, NC 27695, U.S.A.
2Department of Air Quality and Environmental Management, Las Vegas, NV 89155, U.S.A.

Chemistry-meteorology interactions represent one of the most challenging yet important research topics in replicating the real atmosphere and projecting future climate and air quality changes. Online-couple meteorology-chemistry models such as the Weather Research and Forecasting model with Chemistry (WRF/Chem) provide a powerful tool to study such interactions and advance the model's representations of important interactions. As part of the Air Quality Model Evaluation International Initiative (AQMEII) Phase 2, a new chemistry and aerosol option has been implemented in WRF/Chem version 3.4, which uses the 2005 Carbon Bond mechanism (CB05) for the gas-phase chemistry and the Modal Aerosol Dynamics Model for Europe (MADE) with the Volatility Basis Set (VBS) secondary organic aerosol module for aerosol treatments (referred to as CB05-MADE/VBS). WRF/Chem-CB05-MADE/VBS is applied to the full years of 2006 and 2010 over continental U.S. to evaluate the model performance and study the feedbacks of chemistry to meteorology. The results are evaluated in terms of seasonal and annual means using observations from available surface networks, sounding, and satellites.

WRF/Chem-CB05-MADE/VBS simulations of 2006 show a good performance in most seasons for most meteorological variables such as shortwave radiation, 2-m temperature and relative humidity, wind speed and direction, and precipitation. The model also shows an overall good performance in terms of seasonal and annual means for carbon monoxide, sulfur dioxide, nitrogen dioxide, ozone, and fine particulate matter (PM2.5). For example, the normalized mean biases for annual mean concentrations of maximum 8-h ozone and PM2.5 are -12.8% to -3.8% and -9.2% to 2.2%, respectively. However, large biases exist for surface concentrations of some species such as nitrogen monoxide, sulfate in winter and fall, ammonium in fall, and nitrate in winter, aerosol and cloud properties and tropospheric ozone residual. The likely causes for model biases are being analyzed. Comparing the WRF simulation without chemistry, chemistry feedbacks treated in WRF/Chem affect meteorology in a variety of ways, e.g., decreasing downward shortwave radiation, 2-m temperature, planetary boundary layer height, increasing surface longwave radiation, cloud condensation nuclei, and cloud droplet number concentrations, and changing 10-m wind speed, precipitation, and cloud fraction in either directions. These results demonstrate the model's skill in reproducing the real atmosphere and its potential as a regulatory model for the development of integrated emission control strategies to improve air quality yet mitigate adverse climate changes.

Development of an Ozone Screening Tool for the Midwest

Alexander Cohan, Greg Yarwood, Kirk Baker, Brian Mallis, Scott Leopold, Randall Robinson

Evaluation of local ozone impact for new source permitting (NSP) and prevention of significant deterioration (PSD) can be an onerous task often requiring full regional photochemical modeling. In this study, we present the development of an easy to use ozone screening tool for a case study in Illinois. The tool is based on a parametric model (Morris, 2012) that uses CAMx higher-order direct decoupled method (HDDM) to predict the sensitivity of ozone to emission changes. A fixed-effects model is employed to relate HDDM sensitivities with emission rates, stack height, and source location. The screening tool is evaluated with brute force CAMx simulations.

Evaluation of AQMEII Phase 2 Coupled WRF/CMAQ Simulations over North America

Christian Hogrefe¹, George Pouliot¹, Shawn Roselle¹, Rohit Mathur¹, Paul Makar², Christoph Knote³, Yang Zhang⁴, and Stefano Galmarini⁵,

¹Atmospheric Modeling and Analysis Division, National Exposure Research Laboratory, Office of Research and Development, U.S. Environmental Protection Agency

²Modeling and Integration Research Section, Environment Canada

³Atmospheric Chemistry Division, National Center for Atmospheric Research

⁴Department of Marine, Earth, and Atmospheric Sciences, North Carolina State University

⁵European Commission Joint Research Center

We present an evaluation of two annual simulations performed with the coupled WRF/CMAQ system over the continental U.S. as part of the second phase of the Air Quality Model Evaluation International Initiative (AQMEII). Activities in this phase are focused on the application and evaluation of coupled meteorology-chemistry models over both North America and Europe using common emissions and boundary conditions for all modeling groups. The presentation includes an overview of these common input datasets and observational datasets for model evaluation. The AQMEII Phase 2 WRF/CMAQ simulations over North America were performed for both 2006 and 2010. This time period was characterized by a 35% reduction in U.S. SO2 emissions and 20% reduction in U.S. NOx emissions, providing an opportunity for dynamic model evaluation by investigating the impact of emission reductions on ambient concentrations and aerosol/radiation feedback effects. We present results of this dynamic evaluation with a particular emphasis on the role of meteorology vs. emissions in driving interannual changes and the ability of coupled modeling systems to discern these effects. We also present a brief comparison of these WRF/CMAQ simulations with WRF-Chem and GEM-MACH simulations performed for the same time period and domain as part of AQMEII Phase 2.

Development and Application of Interpollutant Trading Ratios to Account for PM2.5 Secondary Formation in Georgia

James Boylan and Byeong-Uk Kim

Georgia Department of Natural Resources (GA DNR)

Facilities applying for PSD air permits are required to model the impact of direct PM2.5 emissions using AERMOD. In addition, these facilities must account for the impacts from precursor emissions (NOx and SO2) on secondary formation of PM2.5. Since AERMOD does not contain chemistry or aerosol formation modules, the secondary formation of PM2.5 cannot be modeled directly in AERMOD. However, PM2.5 interpollutant trading ratios (also called PM2.5 offset ratios) can be used to account for secondary formation of PM2.5 in AERMOD and other dispersion model. This paper describes a technical approach for developing PM2.5 interpollutant trading ratios in Georgia using the CAMx photochemical grid model with flexi-nesting (12 km/4 km/1.333 km). Then, this paper describes how these interpollutant trading ratios can be used to predict secondary PM2.5 in AERMOD as a function of: (1) distance from the source, (2) season of the year, (3) stack height, and (4) emission rates.

Global and Regional Process Comparison

Model resolution and ozone sensitivity to emissions changes in the Northeastern US

Heather Simon, Kirk Baker, Norm Possiel, Pat Dolwick

Previous studies have shown that the chemical mix and intensity of local emissions affect the sensitivity of ozone concentrations to perturbations in NOx and VOC levels. It is well known that core urban areas are more likely to be VOC limited meaning that ozone concentrations decrease in response to decreasing VOC emissions but increase in response to local NOx emissions increases. Conversely, many suburban and rural areas are NOx limited meaning that ozone concentrations decrease in response to decreasing NOx emissions but are relatively unresponsive to VOC emissions reductions. However, this chemistry is affected by complex interactions between meteorology and spatial heterogeneity in emissions sources. In this work we use the CMAQ photochemical model (v5.0.1) to examine the magnitude and frequency of these conditions in the Northeast corridor of the US which includes multiple high population cities (Washington D.C., Baltimore, Philadelphia, N.Y., and Boston) as well as surrounding suburban and rural areas. We apply modeling with emissions from 2007 and emissions with across-the-board regional NOx reductions to characterize the spatial extent and temporal variability in NOx limited and VOC limited conditions over a 7 month period. We run the model using both 12km and 4km grid resolutions to identify when and where model resolution affects conclusions about ozone sensitivity. Finally, we compare results obtained by applying brute-force emissions changes to those obtained using an alternate method of analyzing indicator species ratios for both 4 and 12 km simulations.

Assessment of PM2.5 retrievals using a combination of satellite AOD and WRF PBL heights in comparison to WRF/CMAQ bias corrected outputs

Evaluating the Impact of Increasing Horizontal Resolution on Air Quality Modeling Systems Over Big Metropolitan Areas in Spain

Maria Teresa Pay¹, Gustavo Aravalo¹, Jose Maria Baldasano^1,2

¹Earth Science Department, Barcelona Supercomputing Center, Jordi Girona 29, Edificio Nexus II, 08034 Barcelona, Spain

²Environmental Modeling Laboratory, Technical University of Catalonia, Barcelona, Spain

When applying an air quality model system, a precise definition of the horizontal grid resolution is an important point to take into account, especially over complex terrains as the Iberian Peninsula (Spain). CALIOPE is an air quality forecasting system (CALIOPE AQFS, Baldasano et al., 2011) for Spain running at 4 km x 4 km horizontal resolution. The meteorological model is the WRF-ARW model (version 3.2.1) initialized by the FNL/NCEP data. The emissions are estimated by means of a bottom-up approach implemented in the High-Elective Resolution Modeling Emission System (HERMES version 2.0). The Chemical Transport Model (CTM) is the CMAQ (version 5.0.1). The present contribution evaluates the influence of increasing the spatial resolution from 4 km to 1 km on the CALIOPE AQFS in terms of the main atmospheric pollutants (O₃, NO₂, SO₂, and PM10) over the two biggest cities in Spain, Barcelona and Madrid, and over Andalucia region. Modelled concentrations are analysed and compared against monitoring stations (urban, suburban and rural) distributed within the domain on an hourly basis for April 2013. The analysis is performed by means discrete and categorical statistics. The results indicate that the horizontal grid influence highly depends on the environment (from urban to rural) and the studied pollutant. Overall, the increase of the resolution from 4 km to 1 km improves the general model performance at stations near large emission sources such as the urban agglomerations. Although differences between both resolutions in terms of monthly statistical is relatively low (MB and RMSE less than 3 μg m^-3, and r less than 0.1) the analysis of the concentration maps during a typical pollution episode revels that NO₂ concentrations are better allocate at 1 km than at 4 km, increasing the concentration (~20 μg m^-3) in Barcelona urban conglomeration and the harbour with high agreement with observations. Based on the present results, the CALIOPE AQFS provides forecast at 1 km x 1 km for the studied subdomain, urban areas of Madrid and Barlcelona and Andalucia region.

Application and evaluation of the WRF-CMAQ modeling system to the 2011 DISCOVER-AQ Baltimore-Washington D.C. study

K. Wyat Appel, James Godowitch, Shawn J. Roselle, Jon E. Pleim, David C. Wong and Rohit Mathur

Atmospheric Modeling and Analysis Division, National Exposure Research Laboratory, Office of Research and Development, U.S. EPA, Durham, NC 27711

The DISCOVER-AQ project (Deriving Information on Surface conditions from Column and Vertically Resolved Observations Relevant to Air Quality), is a joint collaboration between NASA, U.S. EPA and a number of other organizations with the goal of characterizing air quality in urban areas using satellite, aircraft, vertical profiler and ground based measurements (http://discover-aq.larc.nasa.gov). In July 2011, the DISCOVER-AQ project conducted measurements in the Baltimore, MD and Washington, D.C. area in the eastern U.S. To take advantage of these unique data, the Community Multiscale Air Quality (CMAQ) model, coupled with the Weather Research and Forecasting (WRF) model is used to simulate the meteorology and air quality in the same region using 12-km, 4-km and 1.33-km horizontal grid spacings. The goal of modeling exercise is to demonstrate the capability of the coupled WRF-CMAQ modeling system to simulate air quality at fine grid spacings in an urban area. A number of updates to the modeling system have been or will be made that will enhance the capability of the modeling system to accurately represent the magnitude and spatial distribution of pollutants at fine model resolutions. Using the data collected during the DISCOVER-AQ campaign, results from the WRF-CMAQ modeling system will be extensively evaluated at both the surface and aloft. The results of these analyses will be presented, along with other interesting results from the modeling exercise, including results from sensitivity runs examining the impact the various updates to the modeling system have on the model estimates.

Characterizing single source contribution to urban-scale ozone and PM2.5

Kirk Baker

Kristen Foley

James Kelly

U.S. Environmental Protection Agency, Research Triangle Park, NC

Single source impacts on secondary pollution including ozone and PM2.5 are becoming increasingly important for facility permit reviews under the Prevention of Significant Deterioration (PSD) and New Source Review (NSR) regulatory programs. The impacts of single source emissions on air quality were previously demonstrated for primarily emitted and secondarily formed pollutants for a 2 week period of July 1999 using two different photochemical transport models and 3 different approaches for estimating downwind contributions. Here, source sensitivity (brute force) based approaches are used to further characterize downwind secondary impacts from single sources using the Atlanta metropolitan area as a basis for these preliminary assessments.

The Community Multiscale Air-Quality Model (CMAQ) version 5.0.1 (www.cmaq-model.org) simulates the formation and fate of photochemical oxidants, ozone, primary and secondary PM concentrations over regional and urban spatial scales for given input sets of meteorological conditions and emissions. In this study, gas phase chemistry is simulated with the Carbon-Bond 05 mechanism with toluene updates (CB05-TU) and aqueous phase chemistry treats sulfur and methylglyoxal oxidation in clouds. The AERO6 aerosol chemistry module includes ISOYYOPIAII inorganic chemistry and partitioning. The CMAQ model is applied for a summer and winter month in 2007 for a domain covering the Atlanta metropolitan area with 4 km sized grid cells. Secondary impacts on PM2.5 and ozone are assessed based on hypothetical sources with varying stack parameters and locations. In addition, these CMAQ simulations are used to test statistical based approaches for estimating ozone and PM2.5 impacts from single sources.

Evaluation of an ozone attribution diagnostic analysis tool implemented in CMAQ

Roger Kwok, Sergey Napelenok, Kirk Baker

Ozone source attribution has been used to support various policy purposes including interstate transport (Cross State Air Pollution Rule) by U.S. EPA and ozone nonattainment area designations by State agencies. Common scientific applications include tracking intercontinental transport of ozone and ozone precursors and delineating anthropogenic and non-anthropogenic contribution to ozone in North America. As in the public release due in September 2013, CMAQ's Integrated Source Apportionment Method (ISAM) attributes PM EC/OC, sulfate, nitrate, ammonium, ozone and its precursors NOx and VOC, to sectors/regions of users' interest. Although the peroxide-to-nitric acid productions ratio has been the most common indicator to distinguish NOx-limited ozone production from VOC-limited one, other indicators are implemented in addition to allowing for an ensemble decision based on a total of 9 available indicator ratios. Moreover, an alternative approach of ozone attribution based on the idea of chemical sensitivity in a linearized system that has formed the basis of chemical treatment in forward DDM/backward adjoint tools has been implemented in CMAQ. This method does not require categorization into either ozone regime. In this study, ISAM will simulate the 2010 North America ozone using all of the above gas-phase attribution methods. The results are to be compared with zero-out difference out of those sectors in the host model runs. In addition, ozone contribution will be shown for sectors and regions to illustrate the utility of the tool for different types of ozone attribution assessments.