Combining CrIS double CO₂ bands for detecting clouds located in different layers of the atmosphere

Abstract Detection of clouds within certain vertical layers of the atmosphere from satellite infrared instruments is challenging, especially of those optically thin clouds due to their small thermal contrasts to the background. This study develops a new method for cloud detection by using the Cross‐track Infrared Sounder (CrIS) hyperspectral radiances at shortwave (~4.3 µm) and longwave (~15 µm) CO 2 bands. Specifically, CrIS longwave channels are first paired with shortwave channels based on weighting function altitudes and sensitivity to clouds. A linear relationship of brightness temperatures between each paired channel is then derived for predicting the shortwave channel from the longwave channel in clear‐sky conditions. A cloud emission and scattering index (CESI) can finally be defined as the difference of the paired shortwave channel between the clear‐sky, regression model predicted and the observed brightness temperatures. Spatial distributions of such derived CESI for several paired channels in the troposphere are examined for a winter storm that occurred in the eastern part of the United States during 22–24 January 2016. It is shown that the CESI values over the storm regions with optically thin cirrus, fog, and supercooled water clouds are positively larger than those over optically thick opaque ice and overshooting clouds or in clear‐sky conditions. Of particular interest is that an area of fog and water clouds over Gulf of Mexico, which are indicated by the Visible Infrared Imaging Radiometer Suite day and night band observations, is identified by the CESI. The global distribution of CESIs derived from CrIS double CO 2 bands with weighting functions peak around 321 hPa agrees well with the distribution of ice cloud optical thickness from the Atmospheric Infrared Sounder version 6 cloud product data set in both daytime and nighttime.