Although tornadoes can occur in any month of the year the ‘tornado season’ is usually thought of as being the months of March through to June, that’s when there’s a greater chance of cooler continental air meeting humid warmer air from the Gulf of Mexico, and hence a much greater chance of thunderstorm development. So far 2018 has been a remarkably quiet year for tornado activity across the United States (fig 1). In fact it could be that April 2018 has one of the lowest tornado counts on record (fig 2).
Here’s a news item from Matt Taylor of the BBC weather team about the slow start to the tornado season in the United States that spurred me into action (fig 2).
I’ve just downloaded the latest updated US tornado data from 1950-2016. First off, this is a big data series, and it’s not easy to navigate and it’s complicated. It’s complicated because for some of the more severe events that run across more than one county, there are multiple entries that share the same common ID. These ID’s are not always stored sequentially in the dataset which makes counting using the EF categories very tricky. Personally, I think this database is in need of a rethink. Maybe by splitting the data into two, with one series of single unique entries for each tornado, and another separate series linked to the first, to store multiple date time and location coordinate data, in much the same way as the HURDAT2 data series does with Hurricanes. Having said that it would be a big task, because there are already over 62,000 entries, but I think it’s long overdue. So with the caveat that the information that I present in the following graphics are done to the best of my programming ability!
This table (fig 1), is a complete list of the occurrence of EF 5 tornadoes from 1950 to the end of 2016, the most recent EF 5 tornado happened over 4 years ago on the 20th of May 2013.
This is a map of the start point of each EF 5 event (fig 2), hopefully they should tally with the table above.
The annual number of tornadoes has shown some decline in the last decade or so (fig 3), even with the better reporting of tornadic events in recent years, made possible by the rise in the use of mobile devices, and the effectiveness of social media reporting. I don’t have a clue why this is, or what it’s linked to, but I hope it’s not down to the jet stream.
Finally, I don’t do much as much as I could with this Tornado application that I’ve written to visualise this data series, because there is so much information regarding tornadoes on the NOAA Storm Prediction website already, and I can’t match the superb graphics that they generate.
In fact – shooting myself in the foot completely – there is now a GIS web application (fig 4) that does much of what I do in a desktop application available on the Midwestern Regional Climate Center [MRCC] website. This is a great resource, and their site is also full of interesting information regarding the destructive power of tornadoes, and the climate of the midwest. Web apps are great, but as a programmer I still prefer the flexibility of writing my own desktop applications, then again I do realise, that you can’t beat the reach of a web application such as this.
The Great Thunderstorm of Widecombe-in-the-Moor on Dartmoor, according to the Wikipedia article took place on Sunday, 21 October 1638, when the church of St Pancras was apparently struck by ball lightning during a severe thunderstorm. An afternoon service was taking place at the time, and the building was packed with approximately 300 worshippers. Four of them were killed, around 60 injured, and the building severely damaged as the tower and roof were ripped off. It’s hard to believe now, when you stand and look at the church as I did just a few days ago, that such a calamity ever befell it, but then again it did happen almost 378 years ago.
According to an article entitled “Jan Reynolds and the Devil” on the website Legendary Dartmoor website says that Widecombe has been called the ‘valley of thunderstorms’, and that the number of dead might have been as many as 21. The late great L C W Bonacina has an account of it in an early edition of the Weather Magazine (Volume 1, Issue 4, August 1946, Pages: 123–125), and it may be as well as being hit by a severe thunderstorm and lightning, Widecombe may have also have laid in the track of a tornado due to the amount of damage done to the church and surroundings. The weather setup at the time sounds a lot like a 15th century Spanish Plume event, with severe thunderstorms, wind and hail sweeping right across the southwest of England that day.