National Weather Service Warning Performance Based on the WSR-88D

Abstract

The National Weather Service (NWS) began operational use of the Weather Surveillance Radar-1988 Doppler (WSR-88D) system in March 1991 at Norman, Oklahoma. WSR-88D data have been available to forecasters at five additional offices: Melbourne, Florida, and Sterling, Virginia (since January 1992); St. Louis, Missouri, and Dodge City, Kansas (since March 1992); and Houston, Texas (since April 1992). The performance of the severe local storm and flash flood warning programs at the six offices before and after the availability of the WSR-88D was measured quantitatively. The verification procedures and statistical measures used in the quantitative evaluation were those used operationally by the NWS. The statistics show that the warnings improved dramatically when the WSR-88D was in operation. Specifically, the probability of detection of severe weather events increased and the number of false alarms decreased. There was also a marked improvement in the lead time for all severe local storm and flash flood events. These improvements were evident throughout the effective range of the radar. Stratification of severe local storm data by severe thunderstorms versus tornadoes revealed an improvement in the NWS's ability to differentiate between tornadic and nontornadic storms when the WSR-88D was in operation. Four individual cases are examined to illustrate how forecasters used the WSR-88D to achieve the improved results. These cases focus on the unique features of the WSR-88D that provide an advantage over conventional NWS radars.