Dynamic spatial fusion of cloud top phase from PARASOL, CALIPSO, cloudsat satellite data

Cloud phase is one of the important parameters of weather and climate research and core elements of atmospheric cloud parametric inversion. The accuracy of its recognition directly influences the inversion precision of cloud optical thickness, spherical albedo, effective particle radius, ice/liquid water content and other parameters. In this paper, we combine with cloud phase products of spaceborne multi-angle polarimetric radiometer, polarized laser radar, and millimeter wave radar. Next, we put forward a Dynamic Spatial Optimal Fusion (DSOF) algorithm. In this algorithm we construct the spatial optimal fusion rules of Cloud Top Phase (CTP) for multi-source data. Finally, we realize the cloud phase spatial fusion using these rules. We used the typhoon "Lupit" as the research object and verified this method. The standard deviations are 4.67%, 7.18%, 40.14% and 36.51%, respectively, compared the results of fused CTP to that of CALIPSO, CloudSat, POLDERS and MODIS. The fusion results are most close to the CTP of CALIPSO. The results show that this method can effectively achieve multi-source cloud phase inversion, and provide new technology for the development and data synergy of spaceborne multi-sensor satellite. (C) 2018 Elsevier Ltd. All rights reserved.