Separation of Convective and Stratiform Precipitation Using Microwave Brightness Temperature

This paper presents a new scheme that classifies convective and stratiform (C/S) precipitation areas over oceans using microwave brightness temperature. In this scheme, data are first screened to eliminate nonraining pixels. For raining pixels, C/S indices are computed from brightness temperatures and their variability for emission (19 and 37 GHz) and scattering (85 GHz). Since lower-resolution satellite data generally contain mixtures of convective and stratiform precipitation, a probability matching method is employed to relate the C/S index to a convective fraction of precipitation area. The scheme has been applied on synthetic data generated from a dynamical cloud model and radiative transfer computations to simulate the frequencies and resolutions of the Tropical Rainfall Measuring Mission (TRMM) Microwave (TMI) Imager as well as the Special Sensor Microwave/Imager (SSM/I). The results from simulated TMI data during the Tropical Ocean Global Atmosphere Coupled Ocean–Atmosphere Response Experiment agree very well with the ground-based radar classification maps. The classification accuracy degrades when SSM/I data is used, due largely to the lower spatial resolution of the SSM/I. The successful launch of TRMM satellite in November 1997 has made it possible to test this scheme on actual TMI data. Preliminary results of TMI derived C/S classification compared with that from the first spaceborne precipitation radar has shown a very good agreement. Further verification and improvement of this scheme are under way.