Evaluation of the non-supercell tornado parameter for the green bay, Wisconsin tornado event of 07 August 2013

Abstract

The difficulty of detecting and forecasting the development of non-supercell tornadoes (NST) is well documented. The rapid, low level development of NSTs decreases the warning time as they often develop in minutes and can last only a short amount of time. Even though most NSTs are relatively weak, EF0-EF2, they still cause damage. Research for non-supercell tornadoes has been focused on the amount of CAPE, CIN, shear, vorticity, and low-level lapse of non-supercell tornadoes. These five variables were compiled into the Non-Supercell Tornado Parameter (NSTP) by Baumgardt and Cook (2006) and is currently in use by the Storm Prediction Center. The research presented focused on the evaluation of this parameter using 24 different experimental runs of the Weather and Research Forecasting (WRF) model. Here, a case study of an event on 07 August 2013 that passed through the Green Bay, WI area, between 0400 UTC and 0600 UTC which produced 6 NSTs was conducted. The predictability of the event was analyzed by using different WRF simulations with varying cumulus parameterizations and microphysical schemes at a 13-km scale. A detailed synoptic and radar analysis was conducted to further increase the understanding of NST environments and radar features. The research found that the 0000 UTC RUC initial fields could accurately forecast the tornadic event. In addition, dual-polarization radar products were used to aid in NST detection. The WRF model solutions highlighted areas of potential NST development that matched the RUC initial field when using an ensemble model approach while statistical results showed little skill. NSTs are difficult to forecast and the results here show that the NSTP can aid in situational awareness for forecasters when looking at ensemble model output.