Aerosol Absorption Over Land Derived From the Ultra-Violet Aerosol Index by Deep Learning

Quantitative measurements of aerosol absorptive properties, e.g. the absorbing optical depth (AAOD) and single scattering albedo (SSA), are important to reduce uncertainties climate radiative forcing assessments. Currently, global retrievals AAOD SSA mainly provided by ground-based Aerosol RObotic NETwork (AERONET), whereas it is still challenging retrieve them from space. However, we found AERONET has a relatively strong correlation with satellite Ultra-Violet Index (UVAI). Based on this, numerical relation built Deep Neural Network (DNN) predict over land long-term UVAI record (2006 2019) Ozone Monitoring Instrument (OMI) onboard Aura. The DNN predicted absorption satisfying for samples AOD at 550 nm larger than 0.1 model performance better smaller aerosols (e.g. smoke) ones mineral dust). validation predictions shows high coefficient 0.90 root mean square 0.005 AAOD, 80% within expected uncertainty (0.03).