Ruc Model-based Convective Probability Forecasts

The low predictability of warm season convective storms remains one of the outstanding challenges to the safety and efficiency of aviation travel. In particular, two unique aspects of thunderstorms combine to make them perhaps the most significant weather-related challenge confronting the aviation community: 1) the rapidity with which they develop and 2) the extreme hazard they pose to aircraft. In contrast to large winter storms, which have a reasonable degree of predictability at a 1-2 day lead time, inherent uncertainties in convective forecasting severely limit the usefulness of explicit thunderstorm predictions beyond about 2 h. In spite of these thunderstorm prediction difficulties, the increasing utilization of the National Air Space has led to a growing need for longer lead-time (2-6 h) thunderstorm likelihood information. This information is needed as guidance to aviation meteorologists and traffic flow managers as they work together to make strategic aircraft routing decisions to optimize air traffic relative to developing thunderstorm clusters. In addition to the obvious safety benefits from improved long lead-time thunderstorm likelihood guidance, significant improvements in air travel efficiency are likely from such information. To address this need, NOAA Forecast Systems Laboratory has developed a convective probability forecast product based on the Rapid Update Cycle (RUC) model (Benjamin et al 2004a,b). Known as the RUC Convective Probability Forecast (RCPF), this product was first tested in a real-time mode during the summer of 2003, with verification of 2-, 4-, and 6-h forecasts performed within the RealTime Verification System (RTVS, Mahoney et al 2002). D2 status within the Aviation Weather Technology Transfer (AWTT, Knapp et al. 2002) process __________________ * Corresponding author address: Stephen S. Weygandt, NOAA/FSL, R/FS1, 325 Broadway, Boulder, CO 80305, [email protected] was obtained during the spring of 2004. A series of improvements have since been made to the product and a statistical comparison of the 2003 and 2004 versions is ongoing. In this paper, we first discuss the rationale for a probabilistic forecast of convection, then describe the techniques employed to create this forecast from the RUC model output. Next, results from the summer 2003 season are summarized, followed by a description of the 2004 accomplishments. We conclude with a discussion of the plans for further development and operational implementation of this product.