Relationship of TRMM precipitation radar observations with TRMM-VISR and TMI observations

Tropical Rainfall Measuring Mission (TRMM) satellite provides a unique opportunity of making the observations of raining systems by a variety of sensors viz visible/infrared sensors (VISR), microwave radiometers (TMI) and precipitation radar (PR). However the valuable measurements of vertical rain profiles by PR are available on a narrow strip of about 200-km width. This may be insufficient for some applications like data assimilation in global atmospheric models. It is worthwhile to explore if there is some relationship between the measurements by different instruments of TRMM. The objective of the present study is to assess if the reasonable information about the vertical rain structure can be obtained from visible/infrared and passive microwave observations. We used nearly two weeks of multi-sensor TRMM observations for the present study. In the first step, an unsupervised classification of cloud systems was performed using Kohenen neural network classifier. The parameters for classification were area averaged (100 X 100 km) visible (with solar correction) and infrared radiances and TMI brightness temperatures at 21.3 and 19.4 GHz. Classification was restricted to five distinct classes. Further, area averaged vertical rain profiles were compiled from PR observations, and were subjected to empirical orthogonal function (EOF) analysis, separately for each class of rain. Preliminary results from a linear regression analysis indicate that dominant modes of vertical rain structure can be correlated, with high correlation coefficients, to the information available from visible/infrared channels and passive microwave channels sensitive to water vapor and rain, using this classification based approach.