Climatological Validation of TRMM TMI and PR Monthly Rain Products over Oklahoma

Abstract

This paper reports the results from a regional validation study of monthly precipitation products generated from sensor measurements aboard the Tropical Rainfall Measuring Mission (TRMM) satellite. The study analyzed 4 yr of precipitation estimates (1998?2001) produced from data collected by the TRMM Microwave Imager (TMI) and the precipitation radar (PR) and compared them with corresponding estimates computed using 5-min rain accumulations from 66 rain gauges in the Oklahoma Mesonet. The applied methodology estimated bulk climate-scale sampling and retrieval errors and biases for the TMI and PR at two areal resolutions: 1° ? 1° and 2° ? 5°. The approach taken in this study generated two gauge-inferred gridded estimates of monthly precipitation over the study period: 1) G0, which was computed by performing a complete integration of the monthly gauge time series, and 2) GS, which consisted of gauge-inferred rain rates subsampled to TRMM overpasses at a gridded resolution of 1° ? 1°, with monthly precipitation estimates derived from the bulk statistics collected during each month of the study. The variable GS depends on the areal swath of each sensor and so yields two sensor-dependent estimates of gauge-inferred precipitation (GTMI, GPR). The advantage of this approach is that it allows for the separation of retrieval and sampling errors, because the subsampled gauge estimates include the temporal sampling errors associated with the satellite sampling. The overall random sampling and retrieval errors for the PR exceeded the TMI errors for the study period, but the PR showed a greater reduction in the errors when the scale was increased to 2° ? 5°. Annual coefficients of variation were also generally lower for the PR than the TMI at this scale. This result was consistent with PR retrieval biases, which were positive over all 4 yr, exceeding the TMI biases in every year of the study.