GPM Ground Validation: Strategy and Efforts
The validation of satellite products is classically defined as a ground-based observing strategy intended to assess whether satellite products meet their stated accuracy requirements and objectives. In the case of the Tropical Rainfall Measurement Mission (TRMM), this philosophy was translated to the quasi-continuous operation of four ground radar sites for which TRMM satellite sensor-based and ground-based rainfall products were compared. The findings from these four sites revealed that TRMM products generally met their stated objectives. In addition, a number of lessons have also been learned in the course of these efforts: (a) quality control and careful construction of ground validation datasets is very labor intensive, but methods that make calibration and quality control techniques more efficient continue to improve; (b) despite every effort, ground validation data has its own set of uncertainties, consisting of both biases (currently ~ 5%) and random errors that are difficult to quantify on short time/space scales such as a single satellite overpass; and (c) direct comparison between rainfall estimates from the TRMM Precipitation Radar (PR) and microwave imager (TMI) reveal that instrument differences have regional and seasonal components that require validation results to be interpreted in a similar fashion.