For our Wednesday Project, we were looking at tornado data. The data provided by the National Oceanic and Atmospherics Administration covered the time period between 1950- 2018 and had various interesting metrics, including origination locations, number of deaths and injuries caused, tornado magnitude, etc.
I decided to look at the number of deaths caused by tornadoes across the years and noticed that the numbers for Alabama were higher than for any other state, although it was not even in the top 10 states for the total number of tornadoes recorded. I wanted to explore what could be the reason for that, and decided to focus my analysis on comparing Alabama, which showed the highest number of fatalities, and Texas, which had the highest total number of tornadoes.
I explored three main potential reasons
- Could it be that the storms in Alabama are stronger, and thus cause more deaths?
- Could the high number of deaths be related to the number of mobile homes in the state?
- Could seasonality have anything to do with high fatality levels?
Overall, it appeared that the answer to my first two questions was ‘no’: the percentage of high-strength storms looks very similar in both states, and the popularity of mobile homes in a state did not seem to correlate with the fatality levels. However, there seasonality analysis showed that whereas there is a more apparent ‘tornado season’ in Texas, with much higher concentration of tornadoes around the spring months, there was no obvious seasonal pattern in Alabama, which could mean that tornadoes are even less predictable.
My seasonality analysis also revealed that there was an extreme spike in tornado-related deaths in April 2011. After doing some additional research, I found out that this was caused by the so-called 2011 Super Outbreak, which was the largest, costliest, and one of the deadliest tornado outbreaks ever recorded in the US. Total fatalities recorded in Alabama that month account for 38% of all tornado-related deaths in the state since 1950.