May 22 - July 30
Comparing Historic Preliminary SPC Tornado Storm Reports to the Final Number of Verified Tornadoes
Annie V. Carlson, Harold E. Brooks, Aaron Treadway, and Somer A. Erickson
What is already known:
- Weather Forecasting Offices throughout the nation have various ways of receiving and processing tornado storm reports.
- The media and other national groups often ask the National Weather Service about the number of tornadoes on a big tornado day. Still, it is a struggle to provide an accurate estimation.
- As big tornado days get more attention, there is an increased need to balance the media with Emergency Managers’ need for the number of tornadoes. Additionally, the scientific community needs consistency in surveying tornado damage.
What this study adds:
- On a big tornado day, the number of reported tornadoes is often an overestimation of the actual verified number of tornadoes.
- Using the overall regression of combined monthly tornado totals can benefit the prediction of final verified tornado counts based on preliminary storm reports.
- Estimating the number of the final tornado count following a big tornado day can be accomplished by applying the number of unfiltered reports to the equation from the regression of the upper 10% of big tornado days.
Abstract:
Over the years, inconsistencies in reporting methods have caused issues with the reports submitted to Weather Forecasting Offices (WFOs) nationwide. WFOs use unfiltered reports to estimate the approximate number of tornadoes that occurred following a day on which tornadoes occurred. Due to the continuity issues with local storm reports, predicting the final number of tornadoes within 24 hours of a tornado day is challenging. Using data from the Storm Prediction Center’s (SPC) severe storm database (2012-2024), the purpose of this study was to analyze the relationships between the SPC’s preliminary storm reports and the confirmed tornado numbers to enhance tornado prediction estimates within a short period following a tornado day. Two statistically significant relationships were identified to predict a relatively accurate estimation for both daily and monthly final tornado counts.