Session Presentations

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

This presentation continues with a report of the thread parallel performance results for CMAQ 5.3 expanded here to include a 192-day simulation. This extends by far the results in [1] and will eventually reach the goal of the entire 2016 year-long scenario. Attention is focused on the Gear, Rosenbrock, and EBI solvers in the Chemistry Transport Model (CTM), by comparing both FSparse [1], and the legacy JSparse [2] algorithms. The former implements OpenMP thread parallelism for all three solvers in the CTM. Thread parallelism in the horizontal advection science procedures (HADV) is also included. The results discussed include execution performance and numerical precision for the first two quarters of the 2016 annual CONUS scenario provided by the U.S. EPA [3]. Both the legacy (EPA) JSparse and the FSparse thread parallel versions are compared in a hybrid MPI+OpenMP version on a heterogeneous cluster of 14 nodes with a total of 192 of the available 232 cores.

[1] G. Delic, Modern Environmental Science and Engineering, Vol. 5, Nr.9, 2019, pp. 775-791. Full text available at: https://www.researchgate.net/publication/338581080_A_Thread_Parallel_Sparse_Matrix_Chemistry_Algorithm_for_the_Community_Multiscale_Air_Quality_Model

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

[3] The author gratefully acknowledges help from Kristen Foley (EPA), Ed Anderson (GDIT), and Elizabeth Adams (UNC) in providing model data and resolving implementation issues.

Roger Kwok and Sarika Kulkarni

The new Detailed Emissions Scaling, Isolation & Diagnostic (DESID) module that has been in place since the implementation of CMAQ5.3.1, offers great flexibility to CMAQ users on performing any number of model sensitivity studies by scaling multiple emission streams and exploring multiple regions/chemical species/PM size distribution combinations. Instead of allowing one single-layered, gridded emissions file with multiple inline emissions as in previous CMAQ versions, multiple emission streams together with the inline emissions can be ingested into the model. However, the arbitrary increase in the number of emission files poses disk space issues. Furthermore, windowing the emissions inventory into a subdomain has to be carried out offline, henceforth duplicating the emission files. The disk space constraint necessitates file size reduction. A conventional way to do this typically uses standalone compression tools (gzip, pbzip2, etc) on netCDF classic files before and after emission processing and/or AQ modeling, thus considerably increasing offline processing time. Therefore, research has been conducted at the California Air Resources Board (CARB) to offer a solution to the disk space problem with reduced overhead cost. This can be achieved via two one-time steps. First, recompile netCDF with additional libraries: Libz to enable compression, and Hierarchical Data Format (HDF5) to allow for more complex data structures with unlimited array size. Second, edit IOAPI source code to utilize the newly recompiled netCDF libraries, so that every Models3 program (CMAQ, MCIP, SMOKE, etc) can be recompiled with the updated libraries. Further code edits to IOAPI also provides end-users an additional option to choose the format of the netCDF outputs: either in compressed mode, or in 64-bit offset format to cater to large file sizes. The latter is particularly useful for CMAQ’s regular or instrumental runs. Initial tests show that file size becomes 10% of its netCDF classic counterpart in compressed mode. More test results will be presented.

Baek, B.H., Wang, C.T., Ma, Siqi, Tong, Daniel and Coats, Carlie

The National Oceanic and Atmospheric Administration (NOAA) and National Weather Services (NWS) have developed its operational National Air Quality Forecasting Capability (NAQFC). One of the NAQFC chemical transport models (CTMs) is the Community Multiscale Air Quality (CMAQ) model developed by the U.S. Environmental Protection Agency to forecast concentrations of ozone and particulate matter below 2.5 µm (termed PM2.5) over the contiguous continental U.S. (CONUS). One of the recent key developments in NAQFC is dynamically coupling the three major components: meteorology, emissions, and pollutant chemistry-transport. Simulating emissions inline is crucial especially for real-time air quality forecasting. It allows the model to include the influences of the most recently forecast meteorological fields on emissions from key emission sources. While biogenic emissions, bi-directional NH3 from fertilizer applications, and point-source plume rise are calculated inline in CMAQ, other important emissions sectors have little or no accounting of meteorological influences in current air quality forecasts. Sources such as onroad/off-network mobile, agricultural livestock, and fugitive dust use annual or monthly total emissions and standard weekly/daily temporal allocation profiles to disaggregate them on finer time scales for the hourly air quality forecasts. Especially, the onroad/off-network mobile emission sector is one of the most complicated and computationally expensive emission sectors in emission processing for CTMs. It requires the most computational resources and time to compute meteorologically sensitive mobile emissions from various processes (i.e., running exhaust, star-up, evaporative, idling, refueling, hot soak, etc.)

This study aims to share the recent inline meteorology-sensitive emissions (MetEmis) module development within CMAQ (CMAQ-MetEmis) that can dynamically represent onroad mobile emissions without its computational requirements that are currently prohibitive in real-time air quality forecasting applications. We will illustrate its computational enhancements and air quality simulation performances, and confirm its feasibility for NAQFC applications.

Patrick C. Campbell, Youhua Tang, Pius Lee, Barry Baker, Daniel Tong, Rick Saylor, Ariel Stein, Jianping Huang, Ho-Chun Huang, Li Pan, Edward Strobach, Jeff McQueen, Ivanka Stajner, Jamese Sims, Jose Tirado-Delgado, Youngsun Jung, Fanglin Yang , Tanya L. Spero, Robert Gilliam, Michael Neish, and Paul Makar

The next-generation National Air Quality Forecast Capability (NAQFC) in the U.S. uses the advanced Finite Volume Cubed-Sphere (FV3) Global Forecast System (GFS) Version 16 to drive the CMAQ model version 5.3.1. Development of the NOAA-EPA Atmosphere Chemistry Coupler (NACC) formed the bridge between the GFSv16 driving meteorology and CMAQ model chemical predictions (i.e., “NACC-CMAQ”). The NACC-CMAQ predictions are further extended from 48 to 72-hours for the first time, and provides scientific advances in atmospheric chemistry modeling to state and local forecasters. NACC-CMAQ became operational at NWS/NOAA in July 2021, thus replacing the previous NAQFC configuration based on the North American Mesoscale Model (NAM)-CMAQv5.0.2. Here we present recent developments of NACC-CMAQ and a comprehensive analysis against the prior operational NAQFC for predictions of summer season ozone (O3) and winter season fine particulate matter (PM2.5). Results show that the NACC-CMAQ model has significant changes to the meteorological and chemical predictions compared to the prior NAQFC, and that NACC-CMAQ has generally more realistic ground-level O3 and PM2.5 predictions and diurnal patterns. New in-canopy effects (e.g., canopy shading and modified turbulence) in NACC-CMAQ will also be assessed for their impacts on air quality predictions. Early results show that inclusion of in-canopy processes can substantially reduce model biases for North American surface O3 forecasts, especially within contiguous forested regions of the eastern U.S. We also present the progress on migrating the GFSv16 data and NACC to the Amazon Web Services (AWS) Cloud and High Performance Computing (HPC) platform, which aims to facilitate community GFSv16-CMAQ applications for any regional domain across the globe.

Arman Pouyaei, Bavand Sadeghi, Yunsoo Choi, Jia Jung, Amir H. Souri, Chun Zhao, Chul Han Song

To improve the representation of convective mixing of atmospheric pollutants in the presence of clouds, we developed a convection module based on Kain and Fritsch (KF) method and implemented it in the Community Multiscale Air Quality model. The KF-convection method is a mass flux-based model that accounts for updraft flux, downdraft flux, entrainment, detrainment, and the subsidence effect. The method is consistent with the convection parametrization of the meteorology model. We apply the KF-convection model to an idealized case and to a reference setup prepared for East Asia during the KORUS-AQ campaign period to investigate its impact on carbon monoxide (CO) concentration at various atmospheric altitudes. We investigate the impact of KF-convection on the horizontal distribution of CO concentrations by comparing it to aircraft measurements and the MOPITT CO column. We further discuss two types of impacts of KF-convection: the direct impact caused by vertical movement of CO concentrations by updraft or downdraft and the indirect impact caused by transport of lifted CO concentrations to another region. May 12 saw a high indirect impact originating from the Shanghai region at higher altitudes and a high direct impact of updraft fluxes at 1 km altitude. However, May 26 revealed an immense updraft increasing higher altitude concentrations (up to 40 ppbv) and diverse indirect impacts over the region of the study (±50 ppbv). The overall comparison shows a strong connection between differences in the amount of concentration caused by the direct impact at each altitude with the presence of an updraft at that altitude. The developed model can be employed in large domains (i.e., East Asia, Europe, North America, and Northern Hemisphere) with sub-grid scale cloud modeling to include the impacts of convection.

Xinyue Huang, Hosein Foroutan

Simulation of windblown dust emissions in regional to global scale atmospheric models requires vegetation coverage as input to account for the alteration of dust flux by vegetation. Most of the current models merely consider the coverage of photosynthetically active (green) vegetation (PV) due to the availability of various remote sensing indices for estimating PV fractions and omit the non-photosynthetic (brown) vegetation (NPV), which is prevalent in dust source regions and potentially has significant impacts on simulated dust loads. In this study, we implemented a satellite-based dataset for total vegetation, which comprises NPV and PV, into the windblown dust module in CMAQ v5.3 model. The NPV and PV fractions were derived from the MODIS Nadir Bidirectional Reflectance Distribution Function-Adjusted Reflectance (NBAR) product MCD43A4 using spectral mixture analysis (SMA). The SMA method assumed that the seven bands of MODIS surface reflectance as well as their log-transformed and interactive terms were linear combinations of the synthetic reference reflectance of three pure components, namely NPV, PV, and bare soil, and resolved the fraction of each component. To examine the effects of NPV, we also conducted a control simulation using the PV-only data (i.e. the fraction of absorbed photosynthetically active radiation (fPAR) data based on MOD15A2GFS product). Simulations were conducted for the entire year 2016 over the conterminous United States at 12 km resolution. With the inclusion of NPV, the simulated dust concentrations decrease by above 10% over most of the southwestern U.S. during spring to autumn. Reductions in dust concentrations are the largest in spring, and when compared to observations from the IMPROVE sites, overpredictions of fine soil concentrations are attenuated at over 93% of the sites in the western U.S. Further analyses indicate that directly sheltering the surface and increasing the threshold velocity through drag partitioning are major mechanisms for NPV to suppress dust emissions. On the other hand, NPV causes the friction velocity to increase by more than 10% over most erodible lands during autumn and winter, which can amplify the dust flux. This study highlights the impact of total (PV+NPV) vegetation, which may be important to other parts of the model (e.g., dry deposition module).

Zachary Lawless, Bradley R. Adams

Windblown dust events can be defined as windblown dust emitted from the Earth’s surface to the atmosphere. These events have significant impact on local air quality. Predicting the location and magnitude of these events is important for air quality assessment and planning. Previous modeling studies have focused primarily on past dust events. This work utilizes a state-of-the-science software framework based on WRF v4.2.1, OBSGRID, and the Community Multiscale Air Quality (CMAQ) v5.3.1 modeling system to predict dust events in Utah. This approach is easily applied to other geographical regions as well. The framework was verified using previous studies for dust events in April 2017 and March 2010. Once verified, the framework was used to predict the impact of future land use properties on dust events. Two scenarios were studied – shrinking of the Great Salt Lake and the addition of large-scale solar farms west of the Wasatch Front. Both showed large increases in dust concentrations near dust sources and lesser but still significant increases over populated areas using the meteorological conditions from the April 2017 dust event. Such information from future impact studies can assess potential impacts from climate change and can guide government water and land use policies to mitigate dust event impacts.

Jaime R. Green, Yuzhi Chen, Jason D. Surratt, and William Vizuete

Field and laboratory measurement data have shown the importance of an aerosol’s phase state in determining its interactions with gas-phase oxidation products derived from biogenic volatile organic compound (BVOC) emissions and on the formation and evolution of secondary organic aerosols (SOA), with implications on climate and air quality. Recent flow tube and smog chamber experiments systematically generated SOA and were subsequently collected by a particle-into-liquid sampler (PILS) for time-resolved chemical measurements. PILS samples were then analyzed for SOA constituents by either reverse-phase liquid chromatography (RPLC) or hydrophilic liquid interaction chromatography (HILIC) coupled to electrospray ionization-high resolution-quadrupole-time-of-flight mass spectrometry (ESI-HR-Q-TOFMS) and inorganic aerosol constituents by ion chromatography (IC). These experimental systems have produced new insights on how acidity, organic water, and phase state alters the formation of isoprene epoxydiols (IEPOX)- derived SOA. These SOA formation processes are not well represented in the current Community Multiscale Air Quality Model (CMAQ) model and could result in a significant under-prediction of the production of organic sulfur, affecting multiphase reactivity, phase state, aerosol growth, and reactive uptake of other chemical species. This work describes the development of the F0AM-CMAQ box model that will use these experimental data to evaluate and develop modifications to the existing CMAQ model. The F0AM-CMAQ model is based on the Framework for 0-D Atmospheric Modeling (F0AM) model and includes two critical updates. First is the integration of the inorganic aerosol thermodynamic equilibrium model (ISORROPIA 2.1) with the purpose of determining the acid-base activity of the inorganic components of the aerosols at varying percentages of relative humidity. The second is based on current implementation of SOA formation in CMAQ. Both of these modules will allow an evaluation of any explicit aerosol formation mechanism, developed based on field and laboratory data, alongside any mechanism developed for CMAQ. This model also has the capability of taking advantage of any explicit gas-phase chemical output from the F0AM model. Results presented here include the comparison of the predicted evolution of SOA mass loading between CMAQ and the observational data provided by our previous experiments using the F0AM-CMAQ model. These model runs will guide future updates to the CMAQ model related to isoprene derived SOA.

Elyse A. Pennington, Karl M. Seltzer, Benjamin N. Murphy, Momei Qin, John H. Seinfeld, Havala O.T. Pye

Volatile chemical products (VCPs) are commonly used consumer and industrial items that are an important source of anthropogenic emissions. Organic compounds from VCPs evaporate on atmospherically relevant time scales and include many species that are secondary organic aerosol (SOA) precursors. However, the chemistry leading to SOA, particularly that of intermediate volatility organic compounds (IVOCs), has not been fully represented in regional-scale models such as the Community Multiscale Air Quality (CMAQ) model, which tend to underpredict SOA concentrations in urban areas. Here we develop a model to represent SOA formation from VCP emissions. The model incorporates a new VCP emissions inventory and categorizes new SOA precursor emissions into three classes: siloxanes, oxygenated IVOCs, and nonoxygenated IVOCs. VCPs are estimated to produce 1.67 ug m-3 of noontime SOA, doubling the current model predictions and reducing the SOA mass concentration bias from -75% to -58% when compared to observations in Los Angeles in 2010. While oxygenated and nonoxygenated VCP species are emitted in similar quantities, SOA formation is dominated by the nonoxygenated IVOCs. This work suggests that VCPs contribute up to half of anthropogenic SOA and it is necessary to better represent SOA precursors from VCPs in CMAQ algorithms to predict the urban enhancement of SOA.

Havala Pye, US EPA Office of Research and Development

Karl Seltzer, US EPA Office of Air Quality Planning and Standards

Ben Murphy, US EPA Office of Research and Development

Chris Allen, General Dynamics Information Technology

Ivan Piletic, US EPA Office of Research and Development

Emma D’Ambro, US EPA Office of Research and Development

Rebecca Schwantes, NOAA/Cooperative Institute for Research in Environmental Sciences at the University of Colorado Boulder

Matt Coggon, NOAA/Cooperative Institute for Research in Environmental Sciences at the University of Colorado Boulder

Emily Saunders, US EPA Office of Chemical Safety and Pollution Prevention

Sara Farrell, US EPA ORISE/Office of Research and Development

Kristen Foley, US EPA Office of Research and Development

George Pouliot, US EPA Office of Research and Development

William R. Stockwell, University of Texas at El Paso

The chemical mechanism of an atmospheric chemical transport model like the Community Multiscale Air Quality (CMAQ) system contains a condensed set of reactions that describe the interactions between emitted organic compounds and nitrogen oxides as well their reaction products. The Community Regional Atmospheric Chemistry Multiphase Mechanism (CRACMM) builds on the history of the Regional Atmospheric Chemistry Mechanism, Version 2 (RACM2) and aims to couple gas- and particle-phase chemistry by treating the entire pool of atmospheric reactive organic carbon (ROC) relevant to present-day emissions. Here, we develop CRACMM species to represent the total emissions of ROC, considering the OH reactivity, ability to form ozone and secondary organic aerosol (SOA), and other properties of individual emitted compounds. Compared to RACM2, we reduce the number of traditional volatile organic carbon species and increase the number of oxygenated and semivolatile to intermediate volatility precursors in the mechanism. In addition, we add explicit hazardous air pollutants to better characterize exposures relevant for human health. We contrast emissions of ROC in 2002 and 2017 from the EPA’s Air QUAlity TimE Series project and their treatment in CRACMM to illustrate how precursors to ozone, SOA, and other endpoints are expected to propagate through the system. The CRACMM mechanism species will be operationalized in chemical transport models in future work.

Yang Zhang1, Kai Wang1, Xiaoyang Chen1, Daniel Schuch1, Eeshan Basu1, Siqi Ma2, Daniel Tong2,3, and Ravan Ahmadov4,5

1Department of Civil and Environmental Engineering, Northeastern University, Boston, MA 02115, USA 2 Department of Atmospheric, Oceanic and Earth Sciences, George Mason University, Fairfax, VA 22030, USA 3IM Systems Group, Rockville, MD 20852, USA 4CIRES, University of Colorado 5Earth Prediction Advancement Division, Global Systems Laboratory National Oceanic and Atmospheric Administration

Mitigation of air pollution and carbon emissions in a changing climate requires an integrated modeling system that can dynamically simulate important feedbacks between climate/meteorology and atmospheric chemistry (i.e., so-called online-coupled models) and simultaneously track the evolution of criteria air pollutants (CAPs) (e.g., SO2, NO2, O3, CO, and PM) and greenhouse gases (GHGs) (e.g., CO2 and CH4). While online-coupled weather and chemistry models (e.g., the Weather Research and Forecasting model coupled with Chemistry (WRF-Chem)) have been developed since early 2000’s, regional/urban air quality modeling that can simultaneously track the atmospheric evolution of CAPs and GHGs has not been performed. In this work, to fill in the above model deficiency, we are developing an innovative integrated model system based on WRF/Chem to simultaneously simulate the atmospheric evolution of CAPs and GHGs and account for the chemistry-meteorology/climate feedback in one simulation (referred to as WRF-Chem-GHGs). WRF-Chem-GHGs integrates WRF-Chem with full chemistry and the WRF Greenhouse Gas Model (WRF-GHG) for the CO2/CH4 tracers. It can simulate the emissions, transport, removal, and chemical and physical processes for both CAPs and GHGs and their interactions via meteorology, radiation, cloud, and chemistry. In the initial application of WRF-Chem-GHGs, anthropogenic CAPs emissions are based on the US EPA’s National Emission Inventory (NEI), and anthropogenic CO2 emissions are based on Vulcan version 3.0. Biogenic VOCs and CO2 emissions are generated online based on the Model of Emissions of Gases and Aerosols from Nature version 2 (MEGAN2) and the Vegetation Photosynthesis and Respiration model (VPRM), respectively. The simulation results will be evaluated using observations from the EPA’s surface networks for CAPs such as (e.g., the Aerometric Information Retrieval Now (AirNow), the Aerometric Information Retrieval System– Air Quality System (AIRS-AQS), and the Chemical Speciation Network (CSN)) and the tall towers datasets for CO2 from the NOAA’s Carbon Cycle Greenhouse Gases (CCGG) program. The enhanced model capability of WRF-Chem-GHGs allows simultaneous predictions of GHGs and CAPs in one simulation that is currently not possible and thus will provide a powerful tool to develop mitigation strategies for co-benefits of air quality, carbon, and climate change.