Spinning Into Consistency: Navigating the Vortex of Tornado Warning Consistency through the Implementation of Specific VROT Training and Criteria.

Tornado warning probability of detection (POD) and false alarm ratio (FAR) continue to be widely discussed. However, observations from a 2022 National Weather Service’s Operations Proving Ground testbed showcased the effectiveness of a new tornado intensity tool and highlighted potential struggles in consistency of rotational velocity (VROT) calculation and warning decisions. Consistent, science-based warnings and warning tag upgrades provide the backbone of effective life saving messaging. Inconsistencies in warning decisions can lead to a plethora of issues, including increased FAR and lower POD. The National Weather Service Warning Decision Training Division (WDTD) has supplied forecasters with baseline criteria for suggested tornado warning Issuance and tornado warning tag upgrades centered around VROT. The initial criteria was to set a 15 meters per second VROT for a base tornado tag and a 20 or 22.5 meters per second VROT for a considerable tag upgrade, with no radar couplet diameter requirement suggested in the initial guidance. This criteria was designed to serve as a recommended initial starting point for all tornado warnings. A 2022 study published by forecasters at the National Weather Service’s Storm Prediction Center in 2021 diagnosed tornado warning performance from 2016 to 2018, and suggested that the best baseline warning performance was also centered around a 15 meters per second VROT but included a 1 nm radar couplet diameter.

The emphasis on VROT’s role in tornado warning performance lays the foundation for a consistent baseline for warning decisions from forecaster to forecaster and office to office. However, a consistent baseline is only as effective as the forecaster's understanding and training. In an effort to diagnose the local understanding of VROT and consistency, the National Weather Service Weather Forecast Office (WFO) Mobile/Pensacola science team developed eight different radar case examples to test forecasters' warning strategies and consistency. Forecasters were asked to provide calculated values for VROT and mesocyclone depth for each radar case provided. Forecasters were also asked to identify any identified tornado debris signatures (TDS) and calculated TDS heights. All data was collected anonymously and analyzed for trends.

The collected data identified inconsistencies among forecaster VROT calculations with an average difference of eight meters per second. In addition, warning issuance times between forecasters varied by up to 10 minutes which resulted in differing warning lead times. Specifically, larger variations in VROT were observed when the tornado vortex signature was not gate-to-gate. These differences in calculations potentially suggest discrepancies in the forecasters' understanding of how to calculate VROT, either from prior training methods or a subjective analysis of tornado couplet tightness. While the variance in warning lead time cannot be directly attributed to the variance in VROT calculation as reasonable subjective criteria does play a role in warning decisions, the correlation of warning decision time and increased VROT spread throughout the experiment suggests a potential target of improvement.

This forecaster experiment highlighted the need for increased training targeted at the quantitative data utilized for warning issuance and tornado tag upgrades. In response to the experiment findings the WFO Mobile science team developed highly targeted training that focuses on forecasters interpretation of quantitative values such as VROT. Through the use of tools such as Google sheets, Gibson Ridge Analyst 2.0 and COBRAS radar analysis software, the training developed provides the office with a nimble and flexible training focused on the tangible quantitative data involved in the warning process. The training was also developed to promote consistency across multiple offices in an effort to support backup operations and mutual aid. It is envisioned that standardizing training and warning interrogation techniques focused on quantitative data interrogation will result in more consistent, improved severe weather warnings and improved POD and FAR. This will ultimately provide improved warning services to our partners and the public.