Long‐Term Trends of High Aerosol Pollution Events and Their Climatic Impacts in North America Using Multiple Satellite Retrievals and Modern‐Era Retrospective Analysis for Research and Applications version 2

Abstract

Mean magnitudes and temporal trends in aerosol optical depth (AOD) from satellite observations and an aerosol reanalysis exhibit a negative‐positive east‐northwest dipole across the contiguous United States with large magnitude negative trends over the eastern United States while small magnitude positive trends over the northwestern states. Based on analyses of Modern‐Era Retrospective Analysis for Research and Applications version 2, the AOD reduction over the eastern United States appears to be largely attributable to reductions in aerosol sulfate, while there have been marked increases in aerosol‐organic and elemental carbon over almost all of the contiguous United States and particularly the northwestern states. Long‐term trends of high aerosol pollution events (HAPEs; days with AOD over the long‐term local 90th daily AOD percentile) during 2000–2017 also indicate that over the eastern United States, both the frequency and spatial scale of summer HAPEs exhibit significant negative trends, while those of moderate aerosol events (days with AOD between the local 30th and 70th AOD percentiles) exhibit weak upward trends. Opposing trends, of smaller magnitude, are derived in the northwestern United States and southwestern Canada. Net and shortwave solar radiation show positive trends at the surface and top of atmosphere under clear‐sky conditions over the eastern United States consistent with the reduction in AOD. However, skin temperatures do not indicate significant trends during all days or HAPEs, indicating that the aerosol‐induced radiation trends are not sufficient to manifest trends in surface temperature. Precipitation during HAPEs appears to have increased during 2000–2017 over the eastern and central United States, indicating that the reduction in aerosol may already be enhancing the precipitation through aerosol‐cloud interactions.

Extreme Aerosol Events Over Eastern North America: 1. Characterizing and Simulating Historical Events

“…The summertime clustering of extreme AOD is likely primarily attributable to three key factors: a higher frequency of stagnation during this season (or alternatively a lower frequency of passages from mid‐latitude cyclones and thus a reduction in wet deposition and dispersal) (J. Wang & Angell, 1999), higher oxidant production due to higher downwelling solar radiation and higher emissions of volatile compounds of both biogenic and anthropogenic origin (Kim et al., 2015). Also consistent with previous research that has shown disproportionate declines in the upper percentiles of AOD over the eastern USA (Jin & Pryor, 2020), analysis of the entire MERRA‐2 and MODIS data sets during 2003–2018 showed that no events of equal or greater magnitude after 2007.…”

“…Figure 2 shows the spatial patterns of daily mean AOD and SSA from MERRA‐2 (left and middle panels) and daily mean AOD from MODIS (right panel, where the white pixels are the cloud screening mask) on the centroid day of each extreme AOD event. Consistent with previous analyses of the seasonality in mean AOD, the mean AOD above the 90th percentile values, and the frequency of exceedance of the long‐term p90 (AOD) (Crippa et al., 2019; Jin & Pryor, 2020), all of the events occurred in the summer. The summertime clustering of extreme AOD is likely primarily attributable to three key factors: a higher frequency of stagnation during this season (or alternatively a lower frequency of passages from mid‐latitude cyclones and thus a reduction in wet deposition and dispersal) (J. Wang & Angell, 1999), higher oxidant production due to higher downwelling solar radiation and higher emissions of volatile compounds of both biogenic and anthropogenic origin (Kim et al., 2015).…”

“…This research focuses on the eastern half of North America because: (i) Although aerosol radiative forcing is not exceptional in a global context relative to, for example, parts of China (Li et al., 2010), India (Tripathi et al., 2005) and Africa (Ichoku et al., 2003), historically it exhibited relatively high aerosol optical depth (AOD) and the direct radiative effect from anthropogenic aerosols over this region was estimated to be −2 Wm −2 during the 1980s (Leibensperger et al., 2012). (ii) AOD and net surface radiation in this area have substantially evolved over the last 30 years, due in large part to implementation of emission control policies under the Clean Air Act (Chin et al., 2014; Jin & Pryor, 2020; Sullivan et al., 2017). (iii) The upper percentiles of AOD have exhibited disproportionately large trends in this region over the last two decades (Jin & Pryor, 2020), and the co‐occurrence probability of extreme AOD and near‐surface PM 2.5 has declined (Jin et al., 2020).…”

“…(ii) AOD and net surface radiation in this area have substantially evolved over the last 30 years, due in large part to implementation of emission control policies under the Clean Air Act (Chin et al., 2014; Jin & Pryor, 2020; Sullivan et al., 2017). (iii) The upper percentiles of AOD have exhibited disproportionately large trends in this region over the last two decades (Jin & Pryor, 2020), and the co‐occurrence probability of extreme AOD and near‐surface PM 2.5 has declined (Jin et al., 2020).…”

Extreme Aerosol Events Over Eastern North America: Part 2. Responses to Changing Emissions

“…The same is true from all five of the other events (Figure 5b–5f). The peak AOD during these events as simulated using WRF‐Chem and emissions for 2015 are reduced to less than twice the regional mean 90th percentile AOD for the three NCA regions as computed using MERRA‐2 output for March 2000 to February 2018 (Jin & Pryor, 2020). This implies that if these meteorological conditions were repeated in the current decade they would no longer generate “exceptional” AOD values due to the reduction in anthropogenic emissions.…”