Statistical Comparison of Mean Areal Precipitation Estimates from WSR-88D, Operational and Historical Gage Networks

The Sacramento Soil Moisture Accounting (SAC-SMA) model has been used to conduct hydrological predictions in the National Weather Service (NWS). The model continuously simulates soil moisture condition and discharge using Mean Areal Precipitation (MAP) as one of inputs. Three MAP estimates respectively derived from Weather Surveillance Radar 1988-Doppler (WSR-88D), River Forecasting Center's operational gage network and National Climatic Data Center's historical gage network were statistically analyzed.

Eight basins in the region near the Oklahoma-Arkansas-Missouri state boundaries are selected for this study. The study period is from June 1, 1993 to December 31, 1997. The radar-based MAP values (MAPX) are derived from the gridded NEXRAD stage III precipitation estimates before June 15, 1996 and thereafter from a mixed use of the stage III and P1 processing algorithms, whereas the operational and historical MAP values are computed by a calibration preprocessor using Thiessen polygon weighting. In terms of long-term averages, MAPX are in very good agreement with both operational and historical MAPs. The overall average ratios of MAPX to operational and historical MAP values over the eight basins are 0.988 and1.008, respectively. MAPX is close to historical MAP in term of precipitation intensity. However, the MAPX values are strongly dependent on the processing algorithms. The underestimation in a range of 3~6% was found for MAPX values before June 15, 1996 while the overestimation was noted for MAPX values after June 15, 1996. The MAPX in winter (November, December, January, February, and March) were more accurate than in summer due to the adjustment made on MAPX for winter seasons. The conditional analysis revealed that the MAPX values were underestimated for all basins when precipitation was observed by gages. When radar and gage predicted the same amount of precipitation, the radar estimates tended to be more intense and less spread out. Effect of three MAP estimates on SAC-SMA model results was also analyzed over one of the selected study basins.