Comparing the NSSL-WRF Model and Convection-allowing Versions of UKMET's Unified Model during the 2013 and 2014 NOAA/HWT Spring Forecasting Experiments

Beginning in 2013, NOAA's National Severe Storms Laboratory (NSSL) and Storm Prediction Center (SPC) established a collaboration with the United Kingdom Meteorology Office (Met Office) as part of annual NOAA/HWT Spring Forecasting Experiments (SFEs). For this collaboration, the Met Office provided once daily forecasts to 48 h over the continental US from convection-allowing 4.4- and 2.2-km grid-spacing versions of their Unified Model (UM) during the 5-week long period of the 2013 and 2014 SFEs (ICs/LBCs were derived from the 25-km grid-spacing global version of the UM). Additionally, several Met Office researchers and forecasters participated in the SFEs, while monitoring UM data flow and forecast products. So far, this collaboration has been extremely beneficial. The Met Office has been able to implement some of the unique storm-scale diagnostics developed at NSSL/SPC like simulated reflectivity and updraft-helicity, as well as examine forecast quality over a much more geographically diverse region than the United Kingdom. Meanwhile, NSSL and SPC have been able to examine forecasts of convection from a high-resolution modeling system completely independent of the WRF model and other US modeling systems. Also, owing to the Met Office's different physical treatment of boundary layer processes in the UM and their importance in the evolution of the pre-convective environment, the NSSL/SPC were particularly interested in the quality of forecast low-level vertical profiles from the convection-allowing versions of the UM since this is well-known weakness in US models.

To gauge the quality of the convection-allowing UM forecasts, daily subjective comparisons of simulated reflectivity were made to the 4-km grid-spacing NSSL-WRF and corresponding observations. The NSSL-WRF is a real-time modeling framework that has been used to provide storm-scale guidance to SPC forecasters since 2006 and is generally highly regarded. Thus, the NSSL-WRF serves as a useful baseline against which to compare the UM forecasts. In addition, forecast soundings from the NSSL-WRF and UM were compared.

Overall, the UM compared very favorably to the NSSL-WRF. In fact, for the majority of the HWT/SFE cases during 2013 and 2014, the UM was rated subjectively as better than the NSSL-WRF. Also, a striking difference between the NSSL-WRF and UM was noticed for forecast vertical profiles of temperature and moisture when capping inversions were present. The UM oftentimes very accurately depicted the sharp gradients in temperature and moisture observed at the interface of the boundary layer and elevated mixed layer, while the NSSL-WRF and other high-resolution WRF model simulations in general predicted more smoothed temperature/moisture gradients and weaker inversions at this interface. This poster will present a summary of results from the Met Office and NSSL/SPC collaboration, as well as show examples of the different forecasts of simulated reflectivity and boundary layer profiles from the NSSL-WRF and UM.