Evaluation of 2B31 TRMM-product rain estimates for single precipitation events over a region with complex topographic features

Abstract

An evaluation of precipitation estimations on the ground for individual rainfall events was carried out by comparing the 2B31 Tropical Rainfall Measuring Mission (TRMM) product versus a high-density, rain gauge network deployed at the ground over a small (about 1000km 2) study area in a continental region characterized by complex topography and high altitude. This comparison, using categorical analysis, showed a good agreement for several skill parameters most frequently used in works of this type. In this paper, it is concluded that Odds Ratio Skill Score (ORSS) is a more reliable measure of skill for categorical statistics than other scores because it better reflects the agreement between the two data sets. Furthermore, ORSS allows one to test the significance of the results so it is possible to discriminate whether the resulting skill is due to pure chance (ORSS was significant in 70% of the cases studied). Although variance and mean analyses generally showed differences between data sets for both the amount and the distribution of rainfall rate over the study area, least squares fits indicate a very high and quite linear correlation for both the mean rainfall rate (r 2 = 0.90) and the maximum amount of precipitation at a given point (r 2 = 0.74). It is concluded that 2B31 TRMM data can be used in weather applications for the area studied here and others with complex orographical characteristics and also as a tool in the diagnosis of individual rain events in other regions where there are no other data sources available. Copyright 2012 by the American Geophysical Union.