Daytime Cloud property retrievals over the Arctic from multispectral MODIS data

Improving climate model predictions for the Arctic requires a complete knowledge of polar cloud microphysics. Passive satellite remote sensing techniques can be used to retrieve the microphysical properties of polar cloud systems. However, variable snow and ice cover, strong surface-based temperature inversions, and mixed-phase clouds make it difficult to obtain complete and accurate retrievals. For this study, a polar cloud mask developed for the Clouds and the Earth’s Radiant Energy System (CERES) is used to assign cloudy pixels in Moderate-Resolution Spectroradiometer (MODIS) satellite imagery. A solar-infrared, infrared, near-infrared technique (SINT) is then applied to MODIS data to retrieve cloud properties over snow and ice covered surfaces. For times when snow and ice are not present, a visible-infrared solar-infrared split-window technique (VISST) is used instead. The MODIS-retrieved cloud properties include phase, height, optical depth, ice crystal diameter, liquid droplet radius, ice water path, and liquid water path.

Modifications will be made to the existing CERES cloud detection algorithm to capture more clouds over cold snow and ice surfaces in low sun conditions. An attempt will be made to discriminate mixed-phase clouds from cloud systems containing either all ice or all liquid. Certain layered cloud systems such as thin cirrus overlapping liquid or mixed-phase cloud decks will be included in the mixed-phase category. CERES/MODIS cloud amounts are validated by comparing them to micropulse lidar measurements over the ARM-NSA Barrow site. The CERES/MODIS cloud properties are validated by comparing them with a similar set of retrievals from the millimeter-wave cloud radar, microwave radiometer, and AERI based in Barrow.