Cloud Property Diagnostics from EPIC Data for Longitudinal GCM Cloud Property Comparisons

Abstract:

The EPIC camera, onboard the DSCOVR spacecraft, makes narrowband spectral measurements of reflected solar radiation of the sunlit hemisphere from the Lagrangian L1 point. These narrowband measurements are converted to shortwave (SW) fluxes using MODIS/CERES regression relationships and CERES angular distribution models (Su et al., 2018). As the Earth rotates, the changing cloud distribution within the sunlit hemisphere generates a time series of the diurnal/longitudinal variability of the reflected SW radiation, which then serves as a unique observational constraint, one that is not available from standard format satellite data, for the diurnal/longitudinal variability of clouds and cloud radiative properties. For comparison to this EPIC data time series, GCM output diagnostics for the sunlit hemisphere are sampled in accord with the DSCOVR-view geometry from the Lissajous orbit about the Lagrangian L1 point. An earlier EPIC/GCM planetary albedo comparison (Carlson et al., 2020) found significant longitudinal biases in model-generated planetary albedo with excess reflection over ocean areas and too little reflection from continental land areas. In this study we examine the longitudinal differences in the seasonal and diurnal/longitudinal variability of cloud fraction, cloud optical depth, and cloud height between EPIC-derived cloud properties over the sunlit hemisphere and the corresponding sunlit-hemisphere GCM-generated cloud properties.