The National Weather Service (NWS) maintains two integrals of statistics for the tornado database. The database is derived by the 121 local Weather Service Forecast Offices (WFO) and filed into the National Climatic Data Center publication Storm Data. The NWS Office of Climate, Water and Weather Services, National Climatic Data Center and the NWS Storm Prediction Center all have versions of the Storm Data database in order to track severe storm and tornado statistics.

The Warning Coordination Meteorologists at each of the 121 NWS WFOs are tasked with surveying tornado damage in their respective areas. This practice goes back in time to the early 1980s when Warning Preparedness Meteorologists surveyed tornado damage before NWS modernization. For the most part, the database has been consistent over the years with only a few changes in the past 52 years. One change occurred in the late 1980s when the mean path width was changed to using maximum path width in describing and tabulating tornadoes in the Storm Data publication.

This paper will discuss how the tornado database is derived and its characteristics and changes since 1950. It will also show the strengths and weaknesses of the Fujita Scale using information from Storm Data focusing on examples from various noteworthy tornadoes and tornado outbreaks over the last 52 years. Rating tornado damage still remains very subjective and varies according to region. Tornado damage ratings in the southeastern and Great Plains regions of the United States are typically done by more experienced people compared to areas where tornadoes are less frequent. Trends in both tornadoes and their F-Scale ratings will be analyzed revealing two noticeable changes in both tornado frequency and the Fujita ratings of tornadoes.