TY - JOUR
T1 - Towards Understanding the dependency of AOD-PM association on the aerosol vertical distribution
AU - Sorek-Hamer, Meytar
AU - Chatfield, Robert B.
AU - Strawa, Anthony W.
AU - Broday, David M.
AU - Banerjee, Tirthankar
AU - Helena, Furman Krasnov
AU - Katra, Itzik
AU - Itai, Kloog
N1 - Funding Information: This work has been funded by Universities Space Research Association (USRA) NPP fellowship and the New England Foundation at the Technion, Israel. The research was done jointly with the Technion Center of Excellence in Exposure Science and Environmental Health (TCEEH). D.B. and I.K. were supported by a Ministry of Science, Technology and Space, Israel, grant #3-13142. We thank Johanna Lepeule for providing data over France, and our colleagues at NASA GSFC for providing the MAIAC AOD data.
PY - 2017/1/1
Y1 - 2017/1/1
N2 - Since the launch of Aqua and Terra satellites, the use of satellite observations to estimate ground particulate matter (PM) concentrations has been a growing trend. Numerous studies used the satellite-based Aerosol Optical Depth (AOD) as an explanatory variable in different statistical models associating the AOD with PM concentration observed at ground level. Although AOD has a better spatial coverage than PM measurements, it sets many constraints on the construction of PM predictive models: a coarse temporal resolution of the Moderate Resolution Imaging Spectroradiometer (MODIS)-based AOD, a large number of missing data due to cloud presence, and an unknown vertical distribution. In conjunction with a long record of vertical aerosol distributions from CALIPSO (Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation), this work analyzed the AOD relationship to surface respirable PM concentrations while searching for the optimum conditions where AOD best represents PM in relation to the location of the detected aerosols in the vertical column. Data from CALIPSO, high resolution satellite-based AOD, and the planetary boundary layer (PBL) thickness were investigated in Northeast USA, Indo Gangetic Plain-India, France, and Israel (2006-2013). We found considerable variability among the different geographic regions and along the years with respect to where, on average, (i.e. below or above the PBL) the aerosols reside. A relatively low number of cases with aerosols detected only within the PBL was found in all locations. Accounting for the vertical location of the aerosols did not result in improved relationships between AOD and PM observations using either Linear regression or Mixed effects models. Although when the aerosols were detected within the PBL the RMSE (estimation error) was relatively low, the explained variability was inconsistent. Understanding the conditions under which the satellite-based AOD represents well the ground PM is still a challenge, this work shades light and should contribute towards this goal.
AB - Since the launch of Aqua and Terra satellites, the use of satellite observations to estimate ground particulate matter (PM) concentrations has been a growing trend. Numerous studies used the satellite-based Aerosol Optical Depth (AOD) as an explanatory variable in different statistical models associating the AOD with PM concentration observed at ground level. Although AOD has a better spatial coverage than PM measurements, it sets many constraints on the construction of PM predictive models: a coarse temporal resolution of the Moderate Resolution Imaging Spectroradiometer (MODIS)-based AOD, a large number of missing data due to cloud presence, and an unknown vertical distribution. In conjunction with a long record of vertical aerosol distributions from CALIPSO (Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation), this work analyzed the AOD relationship to surface respirable PM concentrations while searching for the optimum conditions where AOD best represents PM in relation to the location of the detected aerosols in the vertical column. Data from CALIPSO, high resolution satellite-based AOD, and the planetary boundary layer (PBL) thickness were investigated in Northeast USA, Indo Gangetic Plain-India, France, and Israel (2006-2013). We found considerable variability among the different geographic regions and along the years with respect to where, on average, (i.e. below or above the PBL) the aerosols reside. A relatively low number of cases with aerosols detected only within the PBL was found in all locations. Accounting for the vertical location of the aerosols did not result in improved relationships between AOD and PM observations using either Linear regression or Mixed effects models. Although when the aerosols were detected within the PBL the RMSE (estimation error) was relatively low, the explained variability was inconsistent. Understanding the conditions under which the satellite-based AOD represents well the ground PM is still a challenge, this work shades light and should contribute towards this goal.
KW - Air quality
KW - CALIPSO
KW - PM
KW - Satellite-based AOD
UR - http://www.scopus.com/inward/record.url?scp=85039150683&partnerID=8YFLogxK
M3 - Conference article
SN - 1052-6102
JO - Proceedings of the Air and Waste Management Association's Annual Conference and Exhibition, AWMA
JF - Proceedings of the Air and Waste Management Association's Annual Conference and Exhibition, AWMA
T2 - Air and Waste Management Association's 110th Annual Conference and Exhibition: Bridging Environment, Energy and Health
Y2 - 5 June 2017 through 8 June 2017
ER -