Over the last two decades, the satellite remote sensing datasets have provided continuous and valuable information to constrain the climate models’ simulation of global aerosol extinction (scattering + absorption), i.e. aerosol optical depth (AOD). However, the climate models simulation lacked the observational constraints of such important properties as aerosol absorption, because the commonly used single-viewing photometric satellite sensors have weak or no sensitivity to aerosol absorption. The scientific community has anticipated to address this issue by the use of multi-angular polarimetric (MAP) remote sensing that has high potential for characterizing detailed aerosol properties, including absorption and size. However, until recently there was no reliable aerosol absorption product from MAP due to apparent challenges of the retrieval community to fully explore the high potential of high information content and generate extensive and accurate MAP aerosol products. Meanwhile, the accuracy of aerosol absorption related products (aerosol absorption optical depth - AAOD; single scattering albedo - SSA) retrieved from remote sensing is highly dependent on the information content, in other words, the lower AOD the higher uncertainty of AAOD. Hence, the direct comparison of climatological AAOD between simulation and observation is not straightforward and inconclusive.
In this regard, this study was motivated by the recent enhanced aerosol product generated by the GRASP algorithm (www.grasp-open.com) from the MAP measurements for the entire record of POLDER observations. This POLDER/GRASP product includes spectral AOD and AAOD at six wavelengths from the shortwave visible to the near infrared (VIS-NIR) that were used for constraining aerosol emissions in the GEOS-Chem model. For that purpose we have established an inverse modeling framework based on the adjoint GEOS-Chem for assimilating the POLDER/GRASP products into GEOS-Chem and constrain/retrieve emissions of three major absorbing aerosol species (black carbon - BC, organic aerosol - OA and desert dust - DD) by direct fitting of the observed spectral AOD and AAOD. The obtained MAP-constrained absorbing aerosol emission database (MACE) was used to constrain global aerosol absorption and to pinpoint associated all aerosol instantaneous direct radiative forcing (DRF) (alter incoming solar radiation flux due to the presence of atmospheric aerosol).
Figure 1: Global distribution of aerosol absorption constrained by MAP (POLDER/GRASP) satellite observations. (a) Modern-era global AAOD at mid-visible (550 nm) averaged from 2006 to 2011; (b) Global distribution of anthropogenic fraction of AAOD.
Figure 1 shows the global distribution of aerosol absorption simulation based on the emissions constrained by POLDER/GRASP products. The globally averaged modern-era (2006-2011) AAOD is estimated at 0.0070 in the mid-visible spectrum (550 nm), with the 95% confidence interval [0.0068, 0.0073]. The three major contributors to the value of AAOD (0.0070) are BC (0.0055), OA (0.0007), and DD (0.0008). The anthropogenic AAOD fraction, which corresponds to the AAOD fractional value due to anthropogenic emissions, is estimated globally 72.9% [58.2% to 79.8%] and BC AAOD anthropogenic fraction is 80.3% [65.9%, 88.9%]. The Northern Hemisphere (NH) has a higher anthropogenic fraction of aerosol absorption than the Southern Hemisphere (SH), and the map also indicates that the AAOD over the Arctic region is largely (>85%) from anthropogenic origins. Our estimate suggests a ~1.3-1.8 times stronger global aerosol absorption than the 5th and the recent 6th Intergovernmental Panel on Climate Change (IPCC AR5 and AR6) assessments. Hence, this suggests that elevated aerosol absorption is largely affected by human activities and is among the crucial factors of pollution induced climate change effects.
Furthermore, we calculated the Top-of-Atmosphere (TOA) all-sky aerosol instantaneous radiative effects (DRF) using GEOS-Chem (v11-01) coupled with a radiative transfer mode (RRTMG) (Figure 2). Our best estimate of modern-era aerosol DRF is -0.14 W/m2 with a 95% confidence interval of [-0.25, +0.01] W/m2, which indicates 45-60% greater towards warming than previous estimates due to higher presence of BC and in general aerosol absorption. BC DRF is +0.33 W/m2 with a 95% confidence interval [+0.17, +0.54] W/m2. Interestingly, the estimated increase of BC DRF was only about a factor of 1.3 higher than the multi-model assessments integrated into IPCC, while the associated global AAOD was 1.8 times higher than in the IPCC reports. Based on our analysis that seems associated with 3 factors. First, our satellite derived emissions have higher abundances of BC in industrial areas where surface reflectance is not very high that diminishes BC DRF. Second, the absorbing aerosol below clouds is potentially high over industrial regions that also diminishes the TOA all-sky BC DRF. Third, the aerosol abundance over industrial regions is generally high and DRF is affected stronger by non-linear multiple scattering effects that leads to saturation of the absorption influence and results in lower DRF efficiency. Overall, we found the spatial distribution of aerosol absorption obtained from satellite remote sensing is crucial for adequate constraining of the aerosol climate effects owing to significant regional variation of forcing efficiency.
Figure 2: Modern-era TOA all-sky aerosol instantaneous direct radiative forcing (DRF) and BC DRF constrained by POLDER/GRASP. (a) Spatial distribution of modern-era all sky aerosol DRF averaged from 2006 to 2011. (b) Spatial distribution of modern-era all sky BC DRF averaged from 2006 to 2011. (c) Modern-era aerosol DRF intercomparison. (d) Modern-era BC DRF intercomparison.
Thus, our results demonstrate an interesting insight for climate studies by constraining the large-scale spatial distribution of aerosol absorption derived using advanced satellite observations, such as from multi-angular polarimeters (MAP). While this is challenging it is of crucial importance to combine the satellite observation and model simulation tightly in order to understand global aerosol absorption and its related climate effects.