Evaluation of GPM precipitation products using Q3 over the CONUS

Abstract:

Given the one-year-plus, successful operation of the Global Precipitation Measurement (GPM) Mission, it is now possible to provide quantitative evaluation for a new generation of space-borne instrument measurements and retrieved precipitation products using ground-based precipitation observations with greater certainty.

This study compares three Day-1 GPM surface precipitation products derived from the GPM Microwave Imager (GMI), Dual-Frequency Precipitation Radar (DPR) and DPR-GMI CoMBined (CMB) algorithms, as well as the near-real-time Integrated Multi-satellitE Retrievals for GPM (IMERG) Late Run product, with the NOAA Multi-Radar Multi-Sensor suite (MRMS; now called “Q3”). The comparisons are conducted over the conterminous United States (CONUS) at various spatial and temporal scales with respect to different precipitation intensities, and filtered with radar quality index (RQI) thresholds and precipitation types.

Preliminary comparisons of Day-1 GPM products and Q3 are in reasonably good overall agreement. Based on the mission-to-date (expected to be March 2014 – November 2015) data from all GPM overpasses, the biases relative to Q3 for GMI and DPR precipitation estimates at 0.5^o resolution are negative, whereas the biases for CMB precipitation estimates are positive. Based on all available data (March-July 2015 at this writing), the CONUS-averaged near-real-time IMERG Late Run hourly precipitation estimate is about 33% higher than MRMS. Detailed comparison results are available at http://wallops-prf.gsfc.nasa.gov/NMQ/.

This evaluation is carried out over the CONUS. Additional work is required to determine how applicable the results drawn from this land area might be to oceanic areas and regional land sites, as the precipitation error statistics can be highly regime dependent. Accordingly, the authors plan to conduct more comprehensive comparisons over a variety of regimes as GPM continues its mission.