Is There A Tornado Cycle?

[Bill Tabor]

I was just reflecting and considering the Hurricane Cycle that supposedly cycles every 20 to 30 years and is supposed to be associated with maximums and minimums in hurricane activity such as in the Atlantic. Last I heard we started a new cycle of increased activity back in 95 so we should see increased Hurricane activity in the Atlantic for some time.

My thoughts though are based around tornadoes. If hurricane activity and development is based on ocean temperature, African droughts, and various air circulations - basically weather makers then might not these same maximums and minimums create tornado activity cycles as well?

Guess I need to look up the details of exactly what these cycles are. Perhaps they only affect the Atlantic. If so then maybe it doesn't have as much affect on tornadoes because the Pacific probably plays a bigger role in west to east US weather than the Atlantic. However if both the Atlantic and Pacific are involved then I would think there could be a "tornado cycle". I wonder if anyone has ever looked at the historical data and trends and tried to determine this or tried to relate tornado activity and supercell production with hurricane cycle activity?

Edit:
Here's a link that discusses the hurricane cycle I found:
http://www.post-gazette.com/pg/04277/388989.stm

 

[Mike Johnston]

You can get somewhat of a long-term snapshot on Figure 1 in the following article of annual significant tornado days, and outbreaks:

http://www.spc.noaa.gov/publications/schne...do%20Outbreaks'

Because of the sharp inter-annual variability of tornado occurence, it is kind of a challenge to derive a cyclical pattern from this. Suppose you could construct some kind of exponential moving average line to attempt. The only observation I could feintly make out was perhaps a cyclical maximum from the mid 1950's to the late 1970's, declining since then based on lower highs and lower lows. Complicating this data, however, would be the secular increase on observational capacity over time - perhaps offset by improvements in structural integrity which may have made the "significant" threshhold higher over time since it is based on F-scale.

As for conjecture on what factors might explain any cyclicality, I suppose a starting point might be pattern shifts in the upper level jet - although factors which would, in turn, induce any such shifts would be beyond my pay grade to even attempt to guess at. Unlike hurricane patterns, which seem to be based more on "raw material" of the ocean and atmosphere, seems that tornadogensis is more related to a confluence of several specific factors. Notice there is no attempt each season for any professional forecast of tornado frequency, probably because there is no equivalent to something like the ACE index (ie - raw materials) used in seasonal hurricane forecasting.

Interesting speculation, but would be quite a challenge to establish data correlations, and even tougher to put forth causal explanations.

 

[Chris Nuttall]

First and foremost, the right ingredients have to be in place to even intiate torndo-producing storms. If the circulation patterns are set up just right preventing any S/W troughs, abundant moisture, vertical wind fields, etc. from establishing themselves, then that can do it. This would be associated more with the long-term circulations and climatology. Of course, it all boils down to the fact that we don't really have long enough records of observations (not to mention how obs of tornadoes are skewed lately since reporting and spotting has improved) to really make any determinations.

Hurricanes are one thing. They are large enough and are drivien by few primary mechanisms. Tornadoes are such small-scale phenomena it may well be impossible to nail down all of the climatological variables.

First, we've got to figure out what exacty makes Storm-A produce a tornado when Storm-B doesn't even though they appear identical. This is why I will always be able to find a job somewhere.

 

[Aaron Kennedy]

First and foremost, the right ingredients have to be in place to even intiate torndo-producing storms. If the circulation patterns are set up just right preventing any S/W troughs, abundant moisture, vertical wind fields, etc. from establishing themselves, then that can do it. This would be associated more with the long-term circulations and climatology.

Yep... pretty much. The problem with tornadoes is their parent supercells are often a biproduct of the meso/microscale. Features such as outflow boundaries, cloud cover, etc. can't be determined by climatology.

Now consider how often we rate a season by big events. If just one or two significant events occur, many times, our whole perception of the season is changed. Well how many plains tornado outbreaks are governed by just the synoptic scale? I know I can think of a couple outbreaks/busts that were determined due to mesoscale features.

Aaron

 

[Shane Adams]

There are some who believe tornado peak occurence coincides with an 11-year sunspot cycle. I've heard and read theories that both support and disallow this idea.

 

[Bill Tabor]

Originally posted by Aaron Kennedy
Yep... pretty much. The problem with tornadoes is their parent supercells are often a biproduct of the meso/microscale. Features such as outflow boundaries, cloud cover, etc. can't be determined by climatology.

Now consider how often we rate a season by big events. If just one or two significant events occur, many times, our whole perception of the season is changed. Well how many plains tornado outbreaks are governed by just the synoptic scale? I know I can think of a couple outbreaks/busts that were determined due to mesoscale features.
Aaron

I don't think I really agree. It's probably true that there can be exceptions, but I'd say on a whole that the synoptic and long wave pattern that sets up across the US combined with any blocking factors (which may affect high pressure placement), and jet placement are what set up the stage for the meso and micro scale to do their thing. Whereas it may be true that tornadoes are a smaller scale phenomenon primarily based in the meso and micro scale, I don't think these scales would play the role they do if the overall pattern wasn't strong. This is why things such as good SW flow are important. This is why I am thinking that seasonal or even longer range "decadal" tornado activity may be strongly influenced by the overall global patterns, wind flows, and moisture placement.

Imagine what the southern plains would be like if we continually throughout the year had sw flow, a strong upper jet overhead with good mid level flow as well bringing in low pressure short waves. With the Gulf opened up we'd be chasing constantly. One reason chasing down south has sucked for so many years lately is because the jet placement has gone north, and stayed north early on. That is synoptic scale.

Nice article listed by Mike Johnston by the way. Cool that it goes back so far, but it makes you wonder how reliable the data set is. Still it seems to show an increased period of tornadic activity starting about the 1920's, peaking between 1950 and 1970, and then on the decline to the present day. That seems kind of amazing because everyone always seems quick to jump on the bandwagon that global warming is causing increased tornadic activity with stronger tornadoes than in the past; however it appears that isn't true. From the '20's to 2005...that's like an 85 year cycle. It would be interesting if there was such a widespread cycle. Then again I am just now thinking too that any "cycle" may only be momentary as the Earth is in a constant state of flux. Things such as volcanic activity can affect the weather patterns and may interfere with the "cycle". Anyway...something to ponder.

 

[Glen Romine]

Originally posted by Bill Tabor

I don't think I really agree.

I do. Storms don't care what the synoptic conditions are like - they only sense the immediate atmospheric conditions around them. If the local environment for the storm is favorable, it will thrive regardless of the larger scale actions. Synoptic patterns affect broad regional priming of the atmosphere, in either positive or negative ways. While you could probably tie synoptic pattern activity to global patterns, which do appear to have some influence from long period climate oscillations, numerous studies have attempted to find correlations and none have found anything substantial above the statistical noise level that I'm aware of.

Glen

 

[Bill Tabor]

Originally posted by Glen Romine
I do. Storms don't care what the synoptic conditions are like - they only sense the immediate atmospheric conditions around them. If the local environment for the storm is favorable, it will thrive regardless of the larger scale actions. Synoptic patterns affect broad regional priming of the atmosphere, in either positive or negative ways. While you could probably tie synoptic pattern activity to global patterns, which do appear to have some influence from long period climate oscillations, numerous studies have attempted to find correlations and none have found anything substantial above the statistical noise level that I'm aware of.
Glen

Just imagine the micro and meso scale without a synoptic pattern in place...there would be no background to set up and cause the meso / micro. All global weather is related as it is a closed system.

 

[Glen Romine]

Originally posted by Bill Tabor

Just imagine the micro and meso scale without a synoptic pattern in place...there would be no background to set up and cause the meso / micro. All global weather is related as it is a closed system.

Energy in the atmopshere doesn't all flow down hill to the smallest scales from the the global pattern. When you walk outside and feel a gust of wind, that wasn't caused by a positive North Atlantic Oscillation index.

Glen

 

[Aaron Kennedy]

That arguement goes both ways... just imagine a synoptic setup without mesoscale processes occuring.

One example is the low level jet (a mesoscale process) which is often the primary transport mechanism for adequate moisture return off the gulf.

I'm not saying the synoptic pattern isn't important; it certainly is, but a tornadic cycle based off synoptic pattern will be much more noisy due to mesoscale events.

Aaron

 

[Bill Tabor]

My point is they are both related. Tornadoes aren't caused solely by synoptic or global scale, but they aren't created in a vacume on the meso / micro scale either. They are interrelated / and intertwined. It is the ultimate 'big' picture as the Sun is the ultimate weather maker on Earth which drives every other process. Because they are interrelated is why I wonder if there could be trends or long term cyclical patterns. Perhaps no one has determined what that cycle or pattern is, maybe the data set isn't big enough yet or no one has done that study yet. It may be there are enough negative feedback mechanisms to cancel any long term cycles. I'm not positive there is a cycle, but I don't think anyone is sure there isn't yet either though.

Anyway, I'll let this sit awhile and see what others opinions are.

 

[Chris Nuttall]

Originally posted by Chris Nuttall
Hurricanes are one thing. They are large enough and are drivien by few primary mechanisms. Tornadoes are such small-scale phenomena it may well be impossible to nail down all of the climatological variables.

I totally think that any tornado cycle is a combination of synoptic and mesoscale setups/interactions. Like I said before though, our data set is just now large enough, nor do we fully understand all of the variables at play, to really make any kind of climatological determinations.

The mesoscale is often in such a dynamic flux that it is prone to quick changes and very difficult to observe.

 

[Jay McCoy]

The problem is that stats to help verify any kind of cycle are tainted due to the fact that many more tornados are spotted and reported now than 30 years ago due to chasers, spotters and technology. So even if there is a cycle we wouldnt be able to tell by the data available.

Are there more tornados than 30 years ago or just more people reporting them??? I say the latter.

 

[Bryce Stone]

The most compelling argument for a cycle is the fact that there are years (1988, 2002, etc.) when tornado activity is unusally low. There are also years when the numbers are very high. While there are any number of contributing factors to offical tornado stats for any given year, it is a fact that some years (even recent years) go far above or far below the average (2002 and 2004 come to mind).

The fact that years can vary greatly from the average shows that storms are influenced by more than just the local environment, and would tend to support the idea that there is some kind of cycle. However, I don't see any evidence of an extended (20+ year) cycle, and I don't think we have enough accurate long-term data to be able to detect any long-term cycle yet.

 

[Aaron Kennedy]

The most compelling argument for a cycle is the fact that there are years (1988, 2002, etc.) when tornado activity is unusally low.

Well... statistics dictate that just because you have below/above average numbers of tornadoes in a year, you don't neccesarily have a cycle. It is possible to have the years spread out in a random distrubution.

Aaron