Tornadoes of the Upper Cumberland

With tornadoes touchdowns in my neighborhood this last week, I am reminded of this very nice dataset regarding tornadoes. I thought that a focused regional analysis may be appreciated. Below you can see tracks of tornadoes from 1950 - 2017 that are displayed by EF rank (tornado intensity). The tracks are approximations based on a start and end point, which is why they all appear perfectly strait. Note that the path of most of the points is from south-west to north-east. This follows the prevailing winds of the region.



Many of the following charts I show will have some minor errors in them that result from data smoothing. Please take these with a grain of salt. The first three charts show temporal trends in tornado behavior in the UC.

Yearly Tornado Activity



From this chart one can see that there have been more lately, but it is hard to draw long term conclusions from this data. The data tapers towards our current date as a result of the smoothing process. Other considerations towards the apparent rise is recent tornadic activity may be accounted for by sampling bias. Increasing population make the likelihood of observing a tornado, as well as a report being made about a tornado is more likely to be done in a more interconnected communication oriented society.

Daily Tornado Activity Throughout the Year (by month)



About 60% of tornadoes in the UC occur in March, April, and May.

Hourly Tornado Activity



Peak tornado hour is shown to be 4pm (Central) by this data.

Length of Tornado Tracks in Miles



A logarithmic relationship with regard to the frequency of of tornado track lengths (R² = 0.661) where the longer the track is the less frequently they occur.

Histogram of EF Scales



Azimuthal Direction



A surprising 21% of tornadoes travel a 39° track from the data. It's noteworthy that the data only includes beginning and ending points. The tracks are never a perfectly strait line.

Software: ArcMap 10.7, Google Sheets.