What you're looking at here is an extremely rare phenomenon: A tornado passing close to meteorological instrumentation without destroying it. Because wind sensors are not widespread and tornadoes are small, their paths rarely cross, and when they do, the instrument usually fails before reporting anything useful.

This was the El Reno Oklahoma Mesonet Station's readings last night. Their Ringling station also recorded 78 mph but there were no other gusts above 70 mph. The pressure spike in the graph (for the uninitiated) is equal to the pressure falling more than 0.25" in 30 minutes, much greater than the threshold for reporting a rapid pressure fall, which is 0.06" per hour! (Officially, the highest wind gust measured was 82 mph at Coffeyville, KS).

SIDEBAR: Even before yesterday's severe weather outbreak that caused about 50 tornado reports, I had lamented on Facebook that I am getting close to "storm burn out" which is something that happens to meteorologists (and to a lesser extent, our fans) after a series of extreme weather events one-after-the-other. I'm burning the candle at both ends, staying up late to tweet and Facebook about the active storms, and coming in early the next day to report on the devastation they wreaked. Last night around 10 PM, I was watching a tornado in Dallas, one that just left Oklahoma City, and another headed for Joplin, all at the same time. It was nuts.

But a lot of people are losing their lives and we're not sure why, so we need to continue to report on this. Here's a graph of tornado deaths from 1875 to the current decade.

Clearly, something is amiss here, because such high numbers (the current death toll for 2011 according to the SPC is 483) haven't been approached since 1953 (and before that, 1936). In fact if you look at recent tornado deaths by decade, the toll just last year and this is close to the average for the last five decades!(I didn't consider data before the era of weather radar; this graph was inspired by Ralph Fato, an AccuWeather.com Facebook Fan. An alternate version suitable for TV display can be downloaded).

So as meteorologists and weather enthusiasts, we must continue to report on these events, and, as macabre as it is, try to figure out why so many are perishing. Warning time doesn't appear to be an issue - warnings were given 20 minutes in advance for Joplin, Missouri, which has recorded more deaths than any other tornado in the modern recording era, and this poll by our AccuWeather.com Facebook users show that most people can take shelter in one minute, 91% voting for "10 minutes or under."

Is it just that we haven't seen this many severe twisters since the 1950s? Or is it just that population has grown since then? Are people becoming ttoo accustomed to sirens or is it perceived that the warnings are crying wolf?Leave me a comment below with your opinion. One of the only questions we can answer statistically is: Did the deaths happen in specific places? While there are still 277 deaths which are undetermined this year (including most of the ones this week), the SPC stats show this breakdown for the 2011 fatalities where the location was known:

Yes, the majority (58%) were killed in mobile homes, which is not unusual. But 74 people (36%) were killed in homes or other structures (people who were away from home but took shelter), which may speak to the severity of the tornadoes, and/or lack of shelter. Only 6% were killed outside or in vehicles. It will be important to see how these stats change with the data coming in this week.

Jesse Ferrell