Recent developments in numerical guidance and forecasting strategies for the convective scale

Over the last decade convection-allowing model (CAM) configurations have been used for numerical weather prediction with increasing frequency and numerical guidance from these models is now routinely available for forecasting operations in the U. S. Many forecasters have embraced CAMs because they provide direct information on attributes and related hazards from explicitly simulated convective storms, which is a significant step forward from more traditional modeling systems that parameterize deep convective processes. However, CAMs suffer from the same fundamental set of problems that plagues coarser resolution models: inherent errors in initial conditions, imperfect model formulations (especially model physics), and inadequate spatial resolution. The potential utility of CAMs has been examined and developed by forecasters and research scientists during annual Spring Forecasting Experiments (SFEs), led by the Storm Prediction Center (SPC) and the National Severe Storms Laboratory (NSSL) and conducted by the NOAA Hazardous Weather Testbed, since 2004. These collaborative experiments have resulted in visualization of numerous innovative high-resolution output fields, new post-processing and data mining strategies, and development of CAM-based ensemble forecast systems. The deficiencies of CAMs have also been explored in annual SFEs, including sensitivities to model formulations, physical parameterizations, grid resolution, and assimilation of radar data. These HWT activities have been focused on topics that could directly impact forecasting operations within the subsequent 2-3 years. This has resulted in many new tools and concepts being implemented in the SPC. In parallel, research scientists at the National Severe Storms Laboratory, together with collaborators at other institutions, have been working on a longer-term research project known as convective-scale Warn on Forecast (WoF). The goal of this project is to develop the scientific understanding and technological capability to initialize individual storms in very high resolution CAM-based ensembles and use the ensembles to predict the detailed evolution of storms, including the development of severe weather phenomena such as tornadoes.

The talk would provide an overview of lessons learned from recent SFEs, recent developments in WoF research, and the implications for advancing forecasts for small scale, high impact weather.