Over the past decade, the Global Precipitation Climatology Project (GPCP) has emphasized the importance of obtaining accurate estimates of global precipitation for improvements in general circulation models. As part of this research thrust, GPCP endowed the National Climatic Data Center (NCDC) with the responsibility to develop validation data sets for satellite-estimated rainfall. The result was the Surface Reference Data Center (SRDC).
The SRDC collects raingauge and radar data from nearly a dozen "test-sites" in various parts of the globe, representing different climatic regimes and different challenges for satellite estimation of precipitation. Raingauge data are assembled into area-estimates for 2.5 x 2.5 degree latitude-longitude grids for five day "pentad" totals and monthly totals. Higher resolution 0.5 x 0.5 degree grid estimates are also created for the same temporal resolution. Several different techniques are used to obtain these estimates, including Thiessen Polygons and the PRISM model. Radar-Raingauge syntheses are used in the UK Paragon and Japan AMeDUS regions.
Recently, the GPCP has requested smaller spatial and temporal scales, introducing new problems into the estimates for these regions. As temporal scales are shrunk, time-of-day factors for daily precipitation totals become more pronounced. In the five-day pentad accumulations, for example, daily accumulations for the first and last days are partitioned into 24 equal values, and those values for the portion of the observation period coinciding with the GMT day are included in the pentad estimates. Three of the five days in the accumulation period (and 28 or 29 days in the monthly totals) are not subject to these effects. However, these factors can no longer be ignored when producing daily precipitation estimates.
The Environmental Verification and Analysis Center (EVAC) at The University of Oklahoma is taking up the challenge of providing such estimates. EVAC is building upon the initial work of NCDC in order to (1) compare estimates and error characteristics from several techniques, (2) improve the spatial and temporal resolution of the estimates to correspond with the needs of the GPCP community, and (3) identify regions less prone to error resulting from sharp topographical changes or coastal effects.
This presentation will present EVAC's plans for enhancement of the SRDC, initial results, and a discussion of limitations encountered in producing finer-resolution estimates of precipitation