Optimization and Statistical Evaluation of GOES Cloud-Top Properties for Nowcasting Lightning Initiation

摘要

A cumulus cloud field may develop within a conditionally unstable environment, but only a fraction of the cumulus elements eventually develop into thunderstorms. Determining which of these convective elements is most likely to generate lightning-a critical need for the aviation community and Department of Defense-often starts with little more than a qualitative visual satellite analysis. To protect personnel and property, lightning nowcast tools (e.g., an automated geostationary satellite-based Lightning Initiation (LI) algorithm) require measurable research. This work quantifies the behavior of ten previously identified Geostationary Operational Environmental Satellite (GOES-12) Infrared (IR) Interest Fields (IFs) in the hour before LI. A total of 172 lightning-producing storms that occurred during the 2009 convective season are manually tracked and studied over four regions-Northern Alabama, Oklahoma, Kennedy Space Center and Washington D.C. Four-dimensional and cloud-to-ground lightning arrays provide precise lightning initiation points for each storm in both time and space. Individual tendencies are identified for the ten LI IFs. Statistical significance tests are conducted to determine the potential predictive capability and regional dependence of each IF. This study found that eight out of ten LI IFs exhibited at least 15 minutes of potential predictive capability and 35 minutes on average. Additionally, eight out of ten fields can likely be applied across a large geographical area with minimal error. Future operational applications identified and briefly explored in this work include the use of a lightning probability optimization tool.