Session Presentations

Yusuf Alizade Govarchin Ghale, Ismail Sezen, Goksel Demir, Ali Deniz and Alper Unal

The environmental problems around the world have been increasingly becoming larger-scale and planetary problems because the pollutants emitted are crossing borders and they affect different regions of the Earth that are actually not emitting these pollutants. The Middle East countries, such as Iraq and Syria suffer from air pollution, especially dust. Natural hazards such as drought and desertification have intensified this problem in recent years. The transported dust from these areas affects the air quality of neighboring countries such as Turkey. In this study, ground-level PM10 data and retrieval satellite products derived from MODIS, OMI and CALIPSO were used for aerosol type detection and understanding the source of aerosol pollution in the southern part of Turkey from 2010 to 2020. Based on the results of this study, transported dust from the Arabian Desert affects the air quality of Turkey and causes high aerosol concentration in the southern part of the country. The results indicated that dust storms mostly occur during spring and summer and local emission sources like fuel combustion account for air pollution, especially during winter.

Farzaneh Taksibi, Maryam Zare Shahne, Hossein Khajehpour

Particulate Matter with a diameter of 2.5 micrometers or smaller (PM2.5) pollution in megacities is a major environmental issue in developing countries. To manage this issue, pollutant dispersion modeling is an effective tool for simulating the distribution of the pollutant concentration and predicting the effectiveness of pollution control policies and measures. However, a prerequisite for a valid pollutant dispersion modeling is to consider the natural concentration of pollutants from non-anthropogenic emission sources. Though, due to lack of in-site measurements or source apportionment results, this information is merely available. In this research work, the remote sensing technique is used to estimate the background PM2.5 concentration from natural dust in eight megacities in Iran: Tehran, Isfahan, Ahvaz, Tabriz, Shiraz, Mashhad, Karaj, Kermanshah. The concentrations are estimated in points around the case studies which are located upstream of the anthropogenic emission sources via two methods: 1) calibration of the AOD values received from (MODIS) (combined Dark Target and Deep Blue) and (MISR), based on satellite Terra and 2) comparison of PM2.5 estimates from MODIS, MISR and SeaWiFS Aerosol Optical Depth (AOD) with GWR product. The monthly average concentrations are validated for Tehran through comparison with dust concentrations from source apportionment studies. The derived calibration factors are then applied to the remote sensing-based estimates of the other case studies. As a result, for the first time, the seasonal background PM2.5 concentrations are reported for the main megacities in Iran. This may serve as a valuable input for future pollution dispersion modelings in these case studies.

Andreas F. Grauer¹; Débora L. Perazzoli¹; Jacqueline A. Schraier¹; André L. Malheiros¹; Emílio G. F. Mercuri²; Lucas M. Malheiros¹

¹ EnvEx Engineering and Consulting, Curitiba-PR, Brazil

² Federal University of Paraná - UFPR, Curitiba-PR, Brazil

    Fugitive emissions of inhalable particles (PM10) from a grain conveyor belt were measured by a set of six low-cost sensors from the SDS011 model coupled to a Raspberry Pi microcomputer, which operate on an optical principle. These sensors were mounted downstream and upstream of the conveyor belt considering the predominance of the local wind direction. There were 4 sensors on a rod in two positions downstream of the belt, at distances of 5.35 and 7.35 m respectively, a sensor on the reference equipment side (TEOM) and an upstream sensor 3.25 m from the center of the source. At the same time, a meteorological station was installed to have information on the local wind speed and wind direction.

    In this configuration, the experiment recorded for 5 days the concentrations of PM10, meteorological data and operational data of the conveyor belt, that is, the operational periods and the quantities of grain transported. The SDS011 sensor readings were calibrated by the equation obtained by the sensor operating nearby the TEOM. Next, the dataset, usually available every minute, was summarized into 30-minute averages.

    The fugitive emissions quantification method was based on the Gaussian equation where the emission rate depends on the wind speed, the difference between the upstream and downstream concentration and the horizontal dispersion coefficients σy and σz. For the σy coefficient, a fixed value was adopted, corresponding to a source width of 1m. In this approach, the entire conveyor belt is considered a sequence of individual sources 1 m wide each. Thus, fugitive emissions were quantified per meter of belt in mg/(ms). The σz coefficient was calculated according to the ISC3 atmospheric dispersion model manual as a function of the atmospheric stability class.

    For the application of the fugitive emissions quantification algorithm, also known as the upwind-downwind method, it was necessary to filter the situations of the wind direction perpendicular to the conveyor belt, as this was the position of the rod with the sensors mounted. These situations where the wind lined up with the bar within a tolerance of ±45 degrees were filtered out. Filtering reduced the amount of data from 7056 measurements to 2682 measurements. Then, averages of thirty (30) minutes were calculated for standardization, which reduced the amount of data to 107 values corresponding to 53.5 hours.

    The average PM10 emission rate was 36.6 mg/(ms). In the same period, 5561.3 tons of grain were transported, which means an emission factor of 1266 mg/(tm). Statistical analysis of the data showed that the emission rate depended on the amount transported by the belt as well as on the intensity of the wind. Excluding the two highest and lowest emission rates significantly improved the correlation between the emission rate and transported mass and wind speed.

Md Hasibul Hasan, Yi Li, Haofei Yu

Low-cost air quality sensors are being extensively used nowadays for community air quality monitoring. Most studies focused on installing low-cost sensors in a single geographical location for calibration and testing. How the performances of low-cost sensors vary across different geographical locations still remain largely under-investigated. In this study, performances of low-cost electrochemical air quality sensors installed in six different cities (Atlanta GA, New York City NY, Sacramento CA, Riverside CA, Portland OR, and Phoenix AZ) with different climatic and geographical conditions were compared. Several mathematical models were applied for calibration and testing in both local and other areas. The models selected include linear regression, polynomial regression and random forest. We also investigated the potential cross-interference of CO signal to NO₂ and O₃ in low-cost electrochemical gas sensors. We found that calibration parameters that are developed in one location are generally not suitable for other locations. We also found machine learning methods, when applied for calibration, performs generally better than conventional linear regressions. The incorporation of CO sensor data in calibration models increased model accuracy which suggested a possible cross-interference of CO-related gas to NO₂ and O₃.

Congmeng Lyu, Drexel University, Civil, Architectural, and Environmental Engineering, Philadelphia, Pennsylvania, USA

Shannon Capps, Drexel University, Civil, Architectural, and Environmental Engineering, Philadelphia, PA, USA

Mark Shephard, Environment and Climate Change Canada, Toronto, Ontario, Canada

Daven Henze, University of Colorado, Mechanical Engineering, Boulder, Colorado, USA

Matthew Lombardo, Johns Hopkins University, Baltimore, Maryland, USA

Shunliu Zhao, Carleton University, Civil and Environmental Engineering, Ottawa, Ontario, Canada

Amir Hakami, Carleton University, Civil and Environmental Engineering, Ottawa, Ontario, Canada

Steven Thomas, University of Melbourne, School of Earth Science, Melbourne, Victoria, Australia

Jeremy Silver, University of Melbourne, School of Mathematics and Statistics, Melbourne, Victoria, Australia

Peter Rayner, University of Melbourne, School of Earth Science, Melbourne, Victoria, Australia 

The Community Multiscale Air Quality (CMAQ) model calculates the impact of emissions on atmospheric composition, including inorganic aerosols, while considering the transport and reactions of chemical constituents. Adjusting emissions by comparing modeled concentrations with observations is justified when the science processes are well understood as is the case for inorganic species such as ammonia (NH3). The Finite Difference Mass Balance (FDMB) method and four-dimensional variational (4D-Var) data assimilation leverage differences in simulated and actual observations to revise estimates of emissions with spatial specificity. In this study, we evaluate the capability of a CMAQ-based data assimilation system to improve NH3 emissions, which are relatively uncertain given the diversity of emissions processes in the agricultural sector. To do so, the iterative FDMB and a Python-based four-dimensional variational framework (py4dvar) are integrated with CMAQ and its adjoint to constrain NH3 emissions with observations from the satellite-based Cross-track Infrared Sounder (CrIS). Observing System Simulation Experiments (OSSEs) are conducted with the CrIS observation operator to evaluate the extent to which emissions are expected to be recovered with the hybrid assimilation framework. The OSSE conducted with the 2007 modeling platform and 2016 CrIS data on a regional domain in Georgia results in promising recovery of the true emissions. The framework is then ported to a 2017 modeling platform for assimilation of 2017 CrIS NH3 observations to mitigate the mismatch between modeling platform and satellite observation years. Three suitable periods are selected from April through October 2017 for assimilation. Independent surface measurements are used to evaluate posterior modeled concentrations.

Taheri Ahmad, Panahifar Hossein, Shahidzadeh Hossainreza

Tehran (∼ 35.55◦–35.83◦ N, ∼ 51.09◦–51.59◦ E) is the capital and the most populated city in Iran. The current study intends to determine the share of ventilation coefficient and anthropogenic emission in the PM2.5 concentration rise in the atmosphere above Tehran. Many kinds of research have been done regarding urban air pollution in Tehran, but above question is rarely addressed in previous studies. The research includes in-situ PM concentration measurements supported by synoptic and reanalysis meteorological (ECMWF ERA5) as well as space-borne remote sensing data. The ventilation coefficient has been quantified based on measurements of mixing layer height and wind profiles for two successive years (2018-2019). The ground-based PM2.5 concentration measurements, as well as NO2 measurements by OMI and TROPOMI satellite, are used as an indicator to determine the air pollution episodes. The space-born CALIOP lidar measurements are used to derive the vertical profile of PM2.5 mass concentration from the backscattering coefficient profile. Meanwhile, the occurrence of heavy pollution in Tehran and the transport of pollutants from the densely industrial regions located on the west of Tehran is investigated. Finally, by the comparison of transport flux for heavily polluted episodes of each year, the impact of anthropogenic emission has been investigated. The results of this research are of broad interest to both the scientific community and policy-makers.

Jia Jung¹, Yunsoo Choi^1*, Sayedali Mousavinezhad¹, Daiwen Kang², David C. Wong², Arman Pouyaei¹, Jincheol Park¹, Mahmoudreza Momeni¹, Masoud Ghahremanloo¹, and Hyuncheol Kim³

¹Department of Earth and Atmospheric Sciences, University of Houston, TX, USA

²US Environmental Protection Agency, Research Triangle Park, NC, USA

³Air Resources Laboratory, National Oceanic and Atmospheric Administration, College Park, MD, USA

*Corresponding author: ychoi23@central.uh.edu

To investigate the changes in ozone (O₃) chemical production regime over the contiguous United States (CONUS) with accurate knowledge of ambient concentrations of its precursors, we applied an inverse modeling technique with the Ozone Monitoring Instrument (OMI) tropospheric nitrogen dioxide (NO₂) and formaldehyde (HCHO) retrieval products in the summer of 2011, 2014 and 2017. The inclusion of dynamic chemical boundary conditions provided by hemispheric model simulation and the lightning-induced nitric oxide emissions significantly account for the contribution of the background sources in the free troposphere. Satellite-constrained nitrogen oxide (NO_x) and non-methane volatile organic compounds (NMVOC) posterior emissions significantly mitigate the discrepancy between satellite-observed and modeled columns; nation-wide increases in tropospheric NO₂ column by 34.95 – 48.18%, and the decrease in HCHO column over the southeastern US by -6.77 – -14.85%. Model-derived HCHO/NO₂ column ratio with threshold values to identify O₃ production regime classification (i.e., NO_x-limited, NO_x-saturated, and transition) showed gradual spatial changes in O₃ production regime near urban cores during the study period, as well as apparent shifts from NO_x-saturated regime to transition regime (or transition regime to NO_x-limited regime) over most of the major cities in the western US between 2014 and 2017. In contrast, rural areas, especially in the southeastern US, exhibited an increased HCHO/NO₂ column ratio by -1.30 ± 1.71 between 2011 and 2017, implying the change toward the NO_x-saturated regime. The results of this study show that incorporating satellite observation into numerical modeling could be helpful to implement appropriate emission control policies for O₃ air quality in time.

Disclaimer: The views expressed in this abstract are those of the authors and do not necessarily represent the views or policies of the U.S. Environmental Protection Agency (EPA).