Abstract
The purpose of my project was to investigate verification statistics such as lead times and false alarms between tropical cyclone tornadoes and non-tropical tornadoes. The motivation to better understand any major differences between ten tropical tornadic events and ten non-tropical tornadic events. The data was obtained from the storm-based verification database, also known as the ISU IEM Cow, made by the people from Iowa State University. The database allowed me to view warnings and storm reports issued by the Weather Forecast Office (WFO) for any period after June 2005. IEM Cow attempts to automatically provide verification numbers for the warnings issued, and using these verification numbers like lead times and false alarms, I compared them between the tropical tornadic events and the non-tropical tornadic events. We found that tropical tornado events typically have shorter lead times than non-tropical tornado events. The figures included in this study indicate that our original hypothesis is defensible – it indeed appears that tropical tornadic events are associated with shorter lead times. There are several theories involving warning operations and meteorological conditions that may support the results of the study. Generally speaking, tropical tornadoes are more difficult to detect on doppler radar than non-tropical supercells. In many cases, this would yield lower probability of detection, and a higher false alarm rate. Regarding the storm verification results, the average verified warnings percentage for tropical events are more than double the non-tropical tornadic events. Non-tropical tornadic events are generally more intense and longer duration than tropical tornadic events. Additionally, there are some human elements involved such as ones willingness or ability to provide a storm verification report during high impact region-wide tropical events.