Comparing MRMS Data to Single Radar Data to Improve Severe Thunderstorm Warnings

The Multi-Radar/Multi-Sensor System (MRMS) developed at the National Severe Storms Laboratory (NSSL) has been used operationally by NWS forecasters since 2016. By rapidly integrating data from multiple platforms including radar, satellite, observational data and numerical weather prediction models, the MRMS can provide valuable and robust severe weather products to NWS forecasters. An independent sample of hail-producing thunderstorms that occurred across central North Carolina and central South Carolina from 2017, 2018, and 2019 were examined. A statistical analysis of MRMS products (NSSL Archive), including the Maximum Expected Size of Hail (MESH), was performed on the data. A statistical analysis was also completed on the GR2 Analyst Hail Algorithm (Gibson Ridge Software) using level two radar data retrieved from the NCEI archive. A Contingency Table of Absolute Frequencies was made using the forecasted hail size from each product and the observed hail size. The Probability of Detection (POD), Critical Success Index (CSI), False Alarm Rate (FAR), Hit Rate, and Bias were calculated, and a comparison of these calculations from the MESH product and the GR2 Analyst Hail Algorithm was performed. As expected, it was determined that the MESH is more successful in accurately forecasting severe hail. The purpose of the analyses was to quantify the potential statistical correlations between the products and the occurrence of severe hail, determine product severe hail forecast skill, and any biases. Determination of the discriminating product values between non-severe and severe hail may be ascertained leading to improved severe thunderstorm warning lead times.