Retrieval of multilayer cloud properties based on VIIRS observations

Multilayer clouds (MLC), referred to as clouds with an upper ice cloud overlying a lower water cloud, frequently occur in the Earth's atmosphere, and significantly influence not only cloud property retrieval, but also estimation of cloud radiative effects. Satellite radiometers are playing an important and unique role for inferring cloud properties over a wide spatial and temporal range. Due to the natural limitation of passive instruments in atmospheric vertical structure, most radiometer can retrieve a single and homogeneous cloud. To extend the cloud property retrieval for MLC, the sensitivities of the Visible Infrared Imager Radiometer Suite (VIIRS) channels on clouds are carefully investigated. We find that the 1.6 and 2.25 micron channels show clear variations on cloud thermodynamic phases, and can be used for MLC retrievals. We develop a MLC detection algorithm by using the different features of ice and water clouds at the shortwave and longwave infrared channels. The algorithm is carefully evaluated, and can determine MLC and single-layer clouds correctly with over 70% success rate by comparing with active Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) observations. With the successful detection of MLC, cloud optical thickness and particle effective diameter retrievals from VIIRS observations are obtained, and the properties of both upper ice cloud and lower water cloud can be deduced. Our results indicate the possibility of deriving more detailed and accurate cloud properties for VIIRS.