,
19 tweets,
5 min read
Read on Twitter
I want to talk about the hot takes I've seen after the March 3rd tornadoes in S AL, the FL Panhandle, and GA – specifically, that “the storms had warning and thus there should not have been deaths” or “the NWS did everything it could, we just can’t save everyone.” Thread 1/17ish
This is a cynical viewpoint that is simply untrue. We are on the cusp of a revolution of our weather warning system, brought forth by the increasing number of studies and experiments combining social science with meteorological knowledge.
Through these studies, we have learned of unique societal and economic vulnerabilities that exist in the Deep South that have yet to be addressed – and, if we can tackle these weaknesses, we may be able to lower the current casualty numbers significantly.
In their books “Economic and Societal Impacts of Tornadoes” and “Deadly Season: Analysis of the 2011 Tornado Outbreaks”, Dr. Kevin Simmons and Dr. Daniel Sutter discuss some of these vulnerabilities – with several particularly relevant to this case.
First, tornadoes are 31% deadlier on weekends. Simmons and Sutter suggest this may relate to the fact that 31.1% and *43.2%* of tornado fatalities occur in permanent and mobile homes, respectively. Schools and businesses account for only 9.9% of fatality locations. 
Mobile homes in particular are highly vulnerable. 70% of EF1 fatalities and 60% of EF2 fatalities occur in mobile or manufactured homes, vs 8% and 23% for permanent homes at those tornado strengths.
The thing to consider here is that stronger tornadoes generally have larger wind fields - so the area of EF1 winds that can hurt people in mobile homes in an EF4 tornado is much larger than that of an EF1 tornado, multiplying the risk strong tornadoes pose.
This problem has been further investigated by @StephenMStrader, who has produced maps of mobile home locations in the Southeast. He has added the rotation track of the EF4 Lee County tornado to one of these maps here, revealing the extreme vulnerability:
To make matters worse, community shelters are rare in this part of AL. @fcst70 shows her work on this here - note that requiring community shelters is not necessarily a solution, as many mobile homes in the Southeast are not within parks.
There is also the issue of folk science – as Klockow, Peppler, and McPherson (2014) explain (s3.amazonaws.com/academia.edu.d…). Widely held ideas about tornado behavior exist in AL and MS that do not match real world forecast communication.
These folk science conclusions range from believing that hills, mountains, or waterways stop tornadoes, to increased risk due to a new highway construction. There is also the insidious effect of Optimism Bias – that “home” will always be ok, and the tornado will happen elsewhere.
Importantly, these beliefs are strongly held by many, and are a critical part of an individual’s sheltering decision process, but people are not “stupid” for holding them, they are trying to understand a complex phenomenon using the evidence at hand. You may have similar beliefs!
The VORTEX-SE project underway in N AL has produced a treasure trove of work seeking to better understand these vulnerabilities, and it is likely that an even better picture of the range of societal vulnerabilities will come to light over the next few years.
For now, however, the Southeast is incredibly vulnerable to strong tornadoes – fatalities are 31% higher in the region than the rest of the US as a whole (this suggests that, if the April 2011 tornadoes had occurred elsewhere in the US, there would have been 70 fewer deaths).
Furthermore, within the Southeast, Lee County is particularly vulnerable. Simmons and Sutter found that, through a statistical model of worst case EF5 fatality scenarios, Lee County ranked 17th in the nation, with a worst case of *627* fatalities from a single tornado.
So what does this tell us? As my thread yesterday showed, the warning system was incredibly accurate well in advance of the storms (), but we still need to solve this vulnerability problem.
I believe that, with solutions like improvements to science literacy education, building code management, community tornado planning, and increased shelter access, these vulnerabilities can be reduced, and lives can be saved.
More research needs to be done to prove these solutions. But it can, and should, be done. It is our responsibility as meteorologists, as scientists, as members of the community, to do what we can to help those at risk from tornadoes. /end
An update from today - more work from @StephenMStrader discussing the Beauregard EF4. Really recommend following his work, I've learned a lot from his AMS presentation and tweets about this event!
