With all the interest in tornadoes, I thought it would be helpful to provide some contextual data (focusing on US tornadoes).
Tornadoes are similar. The famous St. Peters tornado outbreak of Minnesota involved two major tornadoes from the same cell (I like to think of them as one phenomenon), which covered a great deal of ground over a very long time period, but only two people died. In 1925 a single tornado covered about 65 miles of land (same as the St. Peter tornado, roughly) in Missouri, Illinois and Indiana. It was probably an F5 (same as St. Peter). a total of about 695 people were killed, including 234 individuals in the town of Murphysboro, Illinois.
Here’s a description of the Minnesota event:
At 3:50 pm a tornado touched down 7 miles … east of Avoca, Minnesota…. As the tornado moved through Murray and Cottonwood counties, it damaged or destroyed 150 farmsteads, killed 500 dairy cattle…
At approximately 4:30 pm the twister… entered Comfrey, a town of 550 people located in both Cottonwood and Brown counties. … It moved through the center of town, destroying a grain elevator, the town hall and most of the main street businesses downtown. The town’s firehouse collapsed, and the school was heavily damaged. Fifty homes were destroyed and 100 people were left homeless. … Approximately 75% of the buildings in Comfrey were damaged or destroyed.
As the tornado moved through Brown County it achieved F4 strength where it damaged or destroyed 130 more farmsteads and killed 500 cattle. Approximately 15% of the 1000 farms in Brown County sustained damage … Northwest of Hanska a man was killed as a result his house collapsing. …. At one point the tornado had a width of 1.25 miles (2 km). After traveling across six counties for 1 hour and 25 minutes, the twister lifted back into the clouds at 5:15 pm …. With a path of 67 miles (108 km), this tornado is the fifth longest track tornado on record in Minnesota.
A few minutes later at 5:18 pm, the same supercell produced another large tornado two miles to the east of Nicollet. One mile to west of St. Peter, a 6 year old boy was killed when he was thrown from the vehicle he was riding in. At 5:30 pm the twister hit St. Peter, a town of about 10,000 people, at F3 strength. It inflicted severe damage on much of the town. … Gustavus Adolphus College sustained heavy damage after taking a direct hit from the twister, …. Officials estimated 500 homes in St. Peter were damaged or destroyed, and over 1000 trees were uprooted. Debris from St. Peter fell as far as Rice Lake, Wisconsin, 136 miles away….
A few years later, my wife showed me her dorm room, destroyed by this tornado. An iron radiator had broken loose and plowed into the very bed she may very well have been asleep in.
Except that it was spring break, the town was empty, the early warning system worked pretty well, Minnesotans tend to be cautious about tornadoes, and the rest of the tornado’s track was across sparsely inhabited farmland. So only two people (and a LOT of cattle) died in that disaster, even though there was an enormous amount of property damage.
All of the 25 deadliest US tornadoes seem to predate radar and early warning systems. Science can save your life.
So, how many tornadoes are there a year, and how deadly are they? Since tornadoes are small and short lived, counting them has been difficult and reliable data are only available for recent years. Even recent data are far from perfect. The death rates may be more accurate, but it is also possible that some earlier deaths attributed to tornadoes were actually due to other sorts of storms, given changes in the ability to verify tornadoes as opposed to, say, straight line wind storms.
Having said all that, the following chart (based on data from disastercenter.com is very interesting:
(In case you can’t make it out, the maximum on the Y-axis is 1,400.)
The steady increase in tornadoes is almost certainly due mainly to ability to spot them. This does not rule out the possibility that changes in climate (global warming) could related to tornado frequency, but this is not the first data set one might want to use to explore this issue. As I have suggested before, there is likely to be a link. What seems to be a qualitative shift in deaths is probably the implementation of warning systems and better science and safety education.