Here at The Guardian Canary, we have begun a research project that involves examining National Weather Service (NWS) Tornado Warnings and the verification of these warnings. Looking at the timeframe of 2014 to 2020, we are trying to see if any correlations exist between the amount of false alarms issued by a National Weather Service office and the number of tornado fatalities that occur in that same office’s territory. In the coming weeks, we will be releasing information on the data that we have gleamed from our research.
The main focus for today’s blog entry will be the Verification Rates for all NWS Tornado Warnings and the resultant false alarms for 2014-2020. For a tornado warning to be deemed verified, a tornado must have been observed or have produced damage for the area and time frame that the tornado warning was in effect. Local Storm Reports (LSRs) contain these tornado observations and damage reports, and they are produced by the same NWS office that issued the tornado warning. All information for this study comes from NWS Tornado Warnings and NWS Tornado LSRs. None of this information is created by The Guardian Canary. 2014-2020 NWS Tornado Verification Statistics were compiled through the use of the Storm Prediction Center’s Severe Weather Event Summaries and Iowa State’s IEM COW (NWS Storm Based Verification) webpages.
The NWS Overall Tornado Verification Rate over the timespan of 2014-2020 was just over 25%. That means 3 out of every 4 Tornado Warnings resulted in a false alarm for the selected years. 3 out of every 4 times that the weather industry and emergency management said “Hey, Look out! A tornado is headed your way!”, they were wrong. 3 out of every 4 times the strongest short-fused warning, outside of a Tsunami Warning and an Extreme Wind Warning, was issued, it was wrong. Perhaps we are being too harsh here. Perhaps others believe 25% is ok for this type of warning. But, how many times can they cry wolf before people stop taking action?
While 2020 was an improvement over 2019, the number of tornado warnings that were false alarms is staggering. 1589 times people were alerted by sirens, alerted on their phones, alerted by their weather radios, had their programming interrupted on television to be told there was a tornado coming and one did not materialize. 1589 times people were told to take life saving action for a tornado that wasn’t there. 1589 times people were essentially lied to about the weather. To think that this isn’t creating complacency in society is wrong and dangerous.
Upcoming Blog Entries
In the coming weeks, we will be discussing how all of this may be affecting the public and their responses. We’ll take a look at individual NWS Offices’ Performance, the good, the bad, and the ugly. In addition to services such as ITAS that are available through The Guardian Canary, we’ll discuss different options and ideas that could be implemented to reduce the number of false alarms. One thing that should be pointed out, this research and this blog are not designed to cast National Weather Service meteorologists in any negative light. NWS meteorologists are the best of the best and should be highly respected by our society. They warn on everything they deem a tornado based off of their training and experience. Their mission statement is to protect life and property. Ours, however, is not. We are only concerned with saving lives and feel that current practices may be hurting that effort during times of severe weather.