Aerosol mass extinction analysis and constraint by assimilation of multispectral UV-VIS retrievals from DSCOVR EPIC

Global models of atmospheric transmission in visible wavelengths require accurate constraint of scattering and absorption by aerosols from both natural and anthropogenic sources. Numerical models greatly reduce the complexity, representing hugely diverse aerosol particle properties generally with only a handful of numbers. For forecasting applications, numerical data assimilation has become a crucial tool for prediction of visibility-reducing aerosols, but operational use has largely been restricted to a single aerosol-related measurement, such that the data assimilation increment simply scales the prior aerosol mass.

This paper describes a study using a sophisticated aerosol retrieval the Multi-Angle Implementation of Atmospheric Correction (MAIAC), applied to an advanced satellite remote sensor, the Earth Polychromatic Camera (EPIC) on board the NASA Deep Space Climate Observatory (DSCOVR). The MAIAC team processed multiple years of EPIC MAIAC, which were then used to diagnose the baseline spectral absorption and extinction simulated by the Navy Aerosol Analysis and Prediction System (NAAPS) global aerosol model. Experiments using sequential assimilation of multiple wavelengths and including assimilation of absorption aerosol optical depth were used to demonstrate the impact on NAAPS speciation in mixed aerosol cases. The implications for spectral mass extinction coefficients, source function biases, and the interactions with satellite retrieval biases were considered.

This study demonstrates how advanced satellite retrievals can be applied to both dynamic constraint of aerosol properties for forecast applications, and for long-baseline diagnosis of source and representation errors to improve simulations over longer time scales and for climatological studies.