Why does it seem like tornado seasons of late have been kinda bleh? Wrong pattern alot during peak season in recent years.
— Landon (@landon1988) May 31, 2017
Landon’s question warranted a response more thorough than what we could provide in a simple Q&A or even a single article for that matter. Anecdotally speaking, this sentiment of disappointment is common among storm chasers, as gripes from tornado junkies ostensibly flood every social media platform whenever a spring ridge builds over the Great Plains. Are seemingly unseasonable stable air-masses becoming seasonable, or are new perceived factors contributing to a pessimistic view of relatively steady tornado statistics?
Let’s begin Part I of AtmoLife’s investigation by examining the full suite of tornado data, detailing every American tornado on record from 1950-2016, made publicly available by our friends at NOAA’s Storm Prediction Center (SPC).
Over this 67-year data span, we’ve witnessed a drastic increase in tornado observations. Today’s “disappointing” 5-year average of nearly 1,000 annual tornadoes exceeds far beyond the 1955 5-year average around 400. These observations however, do not indicate that tornadoes themselves are becoming more frequent, but rather, fewer tornadoes are occurring unreported. Our other friends at NOAA’s National Climatic Data Center (NCDC) note that “Today, nearly all of the United States is reasonably well populated, or at least covered by NOAA’s Doppler weather radars.” NCDC’s rationale for less tornadoes slipping through the cracks is supported by increasing observations of EF-0 tornadoes with no significant trend in the reporting of stronger tornadoes (EF-3 – EF-5), which in their words, “would have likely been reported even during the decades before Doppler radar use became widespread and practices resulted in increasing tornado reports.”
Even with improved reporting capabilities, the US has experienced a noticeable decline in the overall tornado count over the past 5 years. Does this translate to disappointing tornado seasons?
As an editorial aside, we at AtmoLife believe it is not unethical to wish for tornadoes because 1) we understand that you are not wishing for these tornadoes to harm anyone, and 2) your wishing isn’t going to do a damn thing.
In order to properly gauge the disappointment level that a storm chaser may have regarding a given tornado season, we need to analyze beyond the annual overall tornado count. In the age of chasecations, identifying the number of tornadoes likely to be witnessed by a large collective of storm chasers is key to measuring the opinion of said collective. This means filtering the tornado data by date, time, and location.
We’ll consider tornado season in America as March through June, outside of which tornadoes are already too infrequent to promote disappointment. As for the time of day, noon through 9 PM is the ideal storm chasing window, as anything later is probably too dark to observe and anything earlier is likely occurring within a high-precip convective system. In general, it would be convenient to filter out any rain-wrapped or otherwise poorly visible tornado, but unfortunately we are limited by the lack of a “How cool did it look?” field in the SPC data.
The ideal chasecastion region hinges on where tornadoes most frequently occur during peak season (i.e. Tornado Alley) as well as where the terrain is most compatible for tornado viewing. Sorry Dixie Alley, nobody (sane) wants to chase through your piney forests. Although definitions of Tornado Alley vary drastically, evident through a quick google image search, storm chasers consistently flock to a specific region of the country, which we will name Chasecationland. Satisfaction levels from the general storm chasing populous are determined by the success of storm chases in Chasecationland, and to identify the boundaries of Chasecationland, look no further than the old Big XII conference.
We sort through our tornado data in a simple manner by crudely discretizing Chasecationland as the states of Colorado, Iowa, Kansas, Missouri, Nebraska, Oklahoma, and Texas. Applying this regional filter in conjunction with our temporal filters yields annual counts of tornadoes tailored more specifically to severe weather enthusiasts.
In addition to our filtered tornado count, chaser disappointment can be quantified by identifying the number of successful chase days, or rather, lack thereof. Storm chasers tend to be, also anecdotally speaking, “what have you done for me lately” type people. To many chasers who spend money and valuable PTO days on a week-long chasecation, seeing 3 tornadoes in 3 days is far more exciting than seeing 3 tornadoes in a single afternoon while wasting the other 2 days sitting in a Dairy Queen parking lot waiting for the cap to break.
With this chaser mentality in consideration, we seek the number of tornado season days in which our filtered tornado count is at least 5. Although this number is somewhat arbitrary, it’s indicative of a day-1 severe setup that produced multiple opportunities for someone in Chasecationland to spot a tornado if they correctly played their weather forecasting cards that day.
Our graph of Quality Tornado Days (QTD) aligns nearly identically with the overall tornado count implying that the overall tornado count in Chasecationland is rarely skewed by a single major outbreak but rather comprised of the full seasonal weather patterns. It is important to note that our QTD metric equates a 50-tornado day to a 5-tornado day, a comparison that no storm chaser would agree with. In its defense however, most storm chasers would rather witness 10 5-tornado days than a single 50-tornado day (along with 9 bust days), because a higher seasonal QTD reduces the number of bust days and increases the overall number of tornadoes a single chaser is likely to witness (Nobody is going to witness all 50 tornadoes on that outbreak day). To estimate the overall quality of a storm season in Chasecationland, we multiply the observed tornado count by QTD to create a weighted Tornado Quality Index (TQI).
With the observed tornado count and QTD graphs nearly aligning, it is no surprise than out TQI index combing the two functions also takes a similar form. In every graph pertaining to the morale of tornado chasers, there’s a well pronounced drop over the last 5-years. Landon’s hypothesis that the “wrong pattern” is occurring more frequently is strongly supported by the data, especially the QTD result, leading to a handful of follow-up questions. What is causing this recent decline of tornadoes? Are non-meteorological factors amplifying the effects of chaser disappointment? Are people planning their chasecations to the most ideal locations? How do these tornado statistics look with actual mapping data and not a crude outline of the old Big XII conference? Should we expect this tornado decline to continue? AtmoLife will continue to analyze whatever data we can find and provide you with our findings at our convenience.