Impact of Low-Altitude Radar Data on Storm-Scale Numerical Weather Prediction

An outbreak of severe weather occurred across the central United States during the afternoon of June 4, 2005. The outbreak included several modes of convection, including squall lines, supercells, and multicells, and produced a wide spectrum of severe weather including large hail, damaging winds, and some tornadoes. Low-level features, which were particularly critical in detecting localized severe weather in some of the storms, were difficult to identify due to the nature of the convection and its occurrence in areas of relatively sparse radar coverage.

A network of radars has been created by the Center for Collaborative Adaptive Sensing of the Atmosphere (NetRad) for the purpose of sensing weather features low to the ground, while complimenting current radar technology in place (NEXRAD).

A series of model runs have been generated by the Weather Research and Forecasting (WRF) model to investigate the impact of the new NetRad technology in storm-scale numerical weather prediction. Each model run generates a 13 hour forecast and is initialized using a low-resolution (24km) initial background field and one of four different combinations of radar data: 1) A network of NetRads across Wisconsin, 2) NEXRAD across the entire forecast domain, 3) all available NEXRAD data and NetRad data from the Wisconsin network, and 4) no radar data at all. Each run is then compared to the “nature” run, which is a 30 hour WRF forecast generated during the NOAA SPC 2005 Spring Experiment to determine the usefulness of adding radar data to the WRF.

Preliminary results show that NetRad data do provide a better analysis of small-scale convection across Wisconsin in initial hours. However, the value of NetRad and NEXRAD data in forecasting convection in later forecast hours is hindered significantly when compared to the “nature” run, primarily due to differences between the background field and the “nature” run.