Global Warming and Tornadoes: What Can We Say?

I know, it's probably merely anecdotal, but I rarely remember hearing about tornadoes in the winter until about 1990. Not that they never happened, but it was rare. It just seems like there's something of a trend towards tornadoes all year round.

This was a huge system of tornadoes, including some that stayed in contact with the ground for over and hour, and cut swaths through several counties.

Of course, skeptics can say - "find the coal plant in China that caused tornado X," and you can't do it.

Nevertheless, the trends definitely fit the theoretical scenarios.

More intense, but perhaps fewer of them.

Anyone got a per-month count over time?

Tornado climatology is a little tricky. The number of observed tornadoes has doubled since the 50s. However, this is a consequence of the fact that tornadoes are very short-lived, small scale phenomena compared to hurricanes or large-scale storm systems and are less likely to be observed. Furthermore, at least in the US, tornadoes are most common over rather sparsely populated areas, and prior to modern Doppler radar systems and networks of storm spotters many tornadoes likely went unreported and would cause no damage.

However, there are certain large-scale environments that are favorable for large outbreaks of tornadoes, as we had on Tuesday. If climate change makes such environments more or less common, we might expect the number of tornadoes to change accordingly. Also, certain large-scale circulation changes, such as a poleward shift of the jet stream, could make tornadoes more common in some areas.

As to whether tornadoes could become more intense, that's a much more difficult question, even than for hurricanes. It could be tied to more intense thunderstorms, but there are other factors that are important, the most important of which is the magnitude of directional wind shear.

Over at Dot Earth, Andy Revkin has put up a graph from NOAA's National Severe Storms Lab that shows total number of reported tornadoes and storms of F2-F5 from 1950-2006. It looks the number of severe storms has gone down over time, while the total number of storms have gone up - although both of these trends can be explained by more and better data gathering.

The mean annual number of days with tornadoes in the US for the entire year and just in January and February for 5 year periods beginning in 1956 (from the Storm Prediction Center database):

Period Ann. Jan-Feb
1956-60158.410.2
1961-65160.2 8.8
1966-70166.2 7.2
1971-75196.215.8
1976-80178.410.6
1981-85174.6 7.4
1986-90164.610.0
1991-95187.413.2
1996-00198.419.4
2001-05175.4 7.4

The 70s and 90s were big for winter tornadoes, the 60s and 80s weren't. 2001-2004 had the fewest tornado days in January and February of any 4-year period in the record that goes back to 1950, even without adjusting for better reporting. 1998 was the biggest winter on record in terms of tornado days, but after that, it's 1975, 1999, and 1971. Comparing total numbers of tornadoes is very difficult because the answer you get will depend on how you do the adjustment for reporting changes.

As far as theoretical expectations of tornado changes in a warmer world, some of the ingredients for tornadoes will probably get more favorable (thermodynamic) and others (deep tropospheric vertical wind shear) will get less favorable. For others, such as near-surface vertical wind shear, we don't have much of a handle on how they'll change. On balance, the sign of changes is difficult to know. If you consider the distribution of probability of strong tornadoes given a specified combination of environmental parameters currently, and then assume that the thermodynamic parameters get more favorable, while the wind shear gets less favorable, the change in occurrence looks like it would be pretty small. The first climate model results of storm environment changes seem to support that. You can make a much stronger case for an increase in large hail, where the increase in thermodynamic conditions appears to be more important than the decrease in shear.

As far as projected intensity changes are concerned, our ability to discriminate intensity now is limited. We can discriminate environments that will support strong tornadoes (F2 or greater) compared to those that won't to some extent, but we can't predict intensity beyond that even to the level of having much confidence in the intensity of the next tornado from a thunderstorm that has already produced a tornado.

Harold,

Why did you throw the big wet blanket of empirically based science on the global warming party? The idea is to toss out speculative conjecture that can't be verified but also can't be completely refuted either, while making emotional statements about our "children" and evil oil companies.

Are you some kind of "denialist"?

Lance says:

"The idea is to toss out speculative conjecture that can't be verified but also can't be completely refuted either, while making emotional statements about our "children" and evil oil companies."

Huh??

Are we reading the same post?

You should really have your eye's checked, Lance.

Dark Tent,

With my contacts my eyes are 20/20. I re-read Harold Brooks' post to try to understand your remark.

First he shows data that demonstrates no statistically significant increase in winter time tornadoes over the last fifty years. He then discusses this fact in the following paragraph.

The 70s and 90s were big for winter tornadoes, the 60s and 80s weren't. 2001-2004 had the fewest tornado days in January and February of any 4-year period in the record that goes back to 1950, even without adjusting for better reporting. 1998 was the biggest winter on record in terms of tornado days, but after that, it's 1975, 1999, and 1971. Comparing total numbers of tornadoes is very difficult because the answer you get will depend on how you do the adjustment for reporting changes.

He then says,

As far as theoretical expectations of tornado changes in a warmer world, some of the ingredients for tornadoes will probably get more favorable (thermodynamic) and others (deep tropospheric vertical wind shear) will get less favorable.

He then concludes with,

As far as projected intensity changes are concerned, our ability to discriminate intensity now is limited. We can discriminate environments that will support strong tornadoes (F2 or greater) compared to those that won't to some extent, but we can't predict intensity beyond that even to the level of having much confidence in the intensity of the next tornado from a thunderstorm that has already produced a tornado.

Sounds to me like no evidence exists to support the contention that winter time tornadoes have increased and there isn't any empirically based reason to expect them to increase significantly even if temps were in fact to go up.

Maybe you should see your optometrist.

Lance,

If it's not your eyes, maybe there's some dirt on your computer monitor or something.

Who above was "making emotional statements about our "children" and evil oil companies"?

Oh, OK fair enough Dark Tent. There hadn't been any appeals to emotion in this thread. I guess I went off half cocked.