A multisensor diagnostic satellite cloud property retrieval scheme

Active sensor data, in the form of lidar and radar cloud vertical boundaries, are used as a priori information to passive sensor satellite retrievals of cloud optical depth and effective particle radius. Correct placement of cloud in the vertical eliminates the need to approximate cloud height from multispectral passive techniques and is shown to improve uncertainties in nighttime retrievals of thin cirrus in excess of 30%. The new method is exemplified by two case studies using imager data from GOES but remains valid for any passive/active remote sensing application. A strength of this method is its ability to diagnose components of the retrieval uncertainty and thereby quantify retrieval performance. Errors associated with the forward model and measurement uncertainties, and an independent validation of the retrieval, are discussed.