Cloud Type Identification Based on Macro and Micro Properties of Clouds from MODIS

Satellite cloud type product has been operationally processed in China National Satellite Meteorological Center (NSMC) for many years. But due to causes of instruments on board and methods used for cloud type identification, this product still needs improving. In 2011, American scientists proposed a new method to classify cloud types in NPOESS (national polar-orbiting operational environmental satellite system) cloud products algorithm theoretical basis documents. This method uses the satellite derived cloud optical thickness product, cloud effective radius product, cloud top height product, cloud phase product, a set of characteristic values of cloud optical thickness, cloud effective radius, cloud top height, and cloud phase for 6 cloud types to calculate distances between satellite data and characteristic parameters of 6 cloud types. Finally, a minimum distance is obtained, and the corresponding cloud type is derived.Using MODIS data, the minimum distance cloud type identification method is combined with multiple-threshold method, and cloud type identification experiments are carried out. By incorporating methods into software, and using cloud optical thickness product, cloud effective radius product, cloud top height product, cloud phase product, cloud top temperature product, and brightness temperature product of MODIS as inputs of the software, cloud type identification results are outputted for years of 2008 and 2013. Results are compared with ground cloud type observations, and two series are more than 60% consistent. Also, pictures combining satellite derived cloud types and ground hourly precipitation amount observations reflect that analyzed cumulonimbus and nimbostratus are reasonably in the zone of raining. Because the cloud optical thickness can largely reveal the water content in clouds and the vertical thickness of clouds, this cloud type identification method captures raining clouds effectively.