Temps and Dewpoints on days with F-4/EF-4 and F-5/E-F5 Tornadoes

[Hannah.Taylor]

I had a strange brain storm the other day, but I have no idea where to start looking for the data.

Is there any correlation at all whatsoever in the temps and dewpoints on days that produced F-4/EF-4 & F-5/EF-5 tornadoes? (i.e. Greensburg, Wichita Falls 1979, Joplin, Jarrell, Andover, Murfreesboro 2009, Ruskin Heights, Kansas City 2003, Moore 1999, Moore 2013, Grand Island 1980).

Do those dewpoints and temps vary in any aspect from EF-1 to EF-3 and F-1 to F-3 rated tornadoes? In such a way that perhaps those numbers would suggest a tornado might be stronger or weaker?

 

[Shawn Schuman]

This paper should at least get you started, as it looks at various parameters (including T/Td) associated with tornadoes in various regions and seasons.

http://www.spc.noaa.gov/publications/grams/sigt-env.pdf

There's obviously no real threshold since a whole host of other factors come into play, but certainly a higher dewpoint would be more favorable for a significant tornado with all else being equal. Seems I've read another paper on this topic that dealt particularly with discriminating between significant and non-significant tornadoes, but I can't find it off-hand. I'll post it if I can track it down.

 

[Jeff Duda]

I think you'll find any correlation to be pretty loose (low). Generally, one would expect higher theta-e (combination of higher temperature and higher dewpoints) on days where top-end tornadoes occur because you need very strong vertical motion/acceleration to get enough vertical vorticity to generate an EF-4 or EF-5 tornado, and to get that much vertical motion/acceleration you need a ton of CAPE. There are two ways to get very high CAPE:
1) Steep lapse rates (e.g., from an EML)
2) High theta-e (very warm and moist near the surface)
High temperatures and dewpoints will get you (2). However, there are more factors than just surface temp/dewpoint. There are many examples of days where there is 3000-6000 SBCAPE and great shear, but also a cap strong enough to keep storms from forming. That's pretty much a daily occurrence in Iowa in June. So, there is no hard and fast threshold that distinguishes weak environments capable of producing weak tornadoes from those capable of producing violent tornadoes. Although I have noticed many recent EF4/EF5 tornado cases occurred in the presence of > 3000 J/kg SBCAPE, but again, I don't think that is a physically justified threshold.

Another very relevant paper to this issue (doesn't directly address surface temperature and dewpoint, but just about everything else under the sun) is
Thompson et al. (2003), WAF

There are a number of similar papers out there, and maybe even some more updated than this. But I believe most of the informative text for the products in the SPC mesoanalysis came from this paper. You can find other papers at
http://www.spc.noaa.gov/publications/

Also, check the tornado environment browser feature SPC now has:
http://www.spc.noaa.gov/exper/envbrowser/

 

[calvinkaskey]

Seems like dewpoints in mid 60's or higher and temps 80's to 90's usually occur with violent tornadoes.

 

[Hannah.Taylor]

This paper should at least get you started, as it looks at various parameters (including T/Td) associated with tornadoes in various regions and seasons.

http://www.spc.noaa.gov/publications/grams/sigt-env.pdf

There's obviously no real threshold since a whole host of other factors come into play, but certainly a higher dewpoint would be more favorable for a significant tornado with all else being equal. Seems I've read another paper on this topic that dealt particularly with discriminating between significant and non-significant tornadoes, but I can't find it off-hand. I'll post it if I can track it down.

a little too broad for what I am looking for. i was looking for exact data on the tornadoes mentioned in my initial post.

Seems like dewpoints in mid 60's or higher and temps 80's to 90's usually occur with violent tornadoes.

thats what I am thinking, just looking for the actual evidence to support that.

 

[Shawn Schuman]

Oh, gotcha. You can use this to access ASOS/AWOS obs that should probably cover most of the tornadoes you're looking for (or at least a site in reasonably close proximity). Just choose the appropriate state network and then your date.

http://mesonet.agron.iastate.edu/request/download.phtml

You can also use Plymouth State's archived data, which is one of my favorites. The sfc products only go back to August 1998 though, and it obviously won't be as precise as actual obs.

http://vortex.plymouth.edu/u-make.html

I'm sure there are others I'm forgetting as well. To your and Calvin's point, I'd say 60º Td is a rough threshold for strong/violent tornadoes, though it's obviously not a hard-and-fast rule. It especially doesn't apply during HSLC days, where mediocre thermodynamics can be overcome by great kinematics, boundary interactions, etc. For instance, the Tri-State tornado was one of the most extraordinary events in history and yet T/Tds were only in the 60/56 or 62/58 range at various points along its path. Some of the tornadoes on 3/2/12, including strong/violent ones, occurred with dewpoints below 60º and temps in the low 70s.

It'd be an interesting dataset, though, and I'd certainly suspect that a majority of violent tornadoes occur with 80/65º+ T/Tds.

 

[Rick Schmidt]

As a side note, I could be mistaken, but I believe that the dewpoint for the Hesston tornado in 1990 (F5) was around 60, or just under.

 

[Shawn Schuman]

I'm pretty sure dews near Hesston were in the low 60s with temps in the low/mid 70s, but I'd have to check. Anyhow, this may also be of interest (though not surprising). From this blog post by Jon Davies.

Significant tornadoes with surface dew points below 60 deg F are rather rare (in my database of 2001-2005 events, only 17 of 255 significant tornado occurred with a dew point below that!).

 

[Elaine Spencer]

At least two violent Illinois tornadoes, the Plainfield F5 (8/28/1990) and the Roanoke F4 (7/13/2004) occurred on hot and sticky summer days with insanely high CAPE values, I believe in excess of 8000 J/kg in Plainfield and in the 6000+ range in Roanoke.

That said, if high temps/dewpoints were essential to violent tornado formation, I would think that violent tornadoes would occur much more often in midsummer than they do. Instead, tornados are much LESS frequent in July and August than they are earlier or later. (The Plainfield tornado is, to date, the only recorded F5/EF5 to strike the U.S. in the month of August.)

 

[Paul Knightley]

Looking for exact values in meteorology is not really appropriate, although it can be interesting, of course. It's the blend of ingredients which is important, rather than focussing in on any one thing. The relationship of dewpoint to air temperature, and the depth of the low-level moisture is more important than the exact values of dewpoint temperature. Of course, a higher dewpoint means more latent heat, and all other things being equal in an unstable environment, a more powerful updraught will ensue - this is hypothetical, of course, as all other things are never equal when dealing with the weather.

 

[Bob Schafer]

I had some time to kill today, and I liked this question about violent tornadoes and T/Td's. So, I did a little digging around and came up with the following data set. All these events are particularly interesting to me for one reason or another, but this is not intended to be any sort of representative sample beyond that, by any stretch. I used the SPC Reports page to narrow down the times of day and the locations, and in each case I tried to find the *best* raw metar data, though sometimes finding a perfect metar ob left a little to be desired (without getting too anal about it). I used this page for the data: http://vortex.plymouth.edu/sa_parse-u.html

I added the year and month to each for clarity.

In chronological order:

KOKC 990503_2328Z 14017G24KT 9SM -TSRA BKN019 BKN150 OVC250 23/20 A2959 RMK AO2
TSB28RAB28 PRESRR TS S-W MOV NE OCNL LTGICCC P0000
May 3, 1999 "Bridge Creek-Moore" OK.

KBGD 010529_2351Z AUTO 10008G14KT 060V140 10SM VCTS -RA FEW023 BKN030 OVC050
23/20 A2980 RMK AO2 PK WND 16031/2258 LTG DSNT ALQDS TSE30B39RAB17
SLP056 P0002 60002 T02280200 10289 20228 55004
May 29, 2001. An F4 struck near White Deer TX (which I barely missed due to being an incompetent noob).

KIAB 040612_2356Z 15018G25KT 7SM -TSRA FEW003 BKN030CB BKN060 BKN250 26/23 A2972
RMK OCNL LTG ICCCCG TS OHD MOV NE SLP052 WR// WND DATA ESTMD 60000
55017
June 12, 2004 "Mulvane KS" tornado. While *only* an F3, I'd bet any chaser who was there will admit to this storm being one of their career favorites. Mind-blowing day.

KP28 070505_0156Z AUTO 16011G21KT 26/19 A2961 RMK AO2 PK WND 17028/0115 SLP011
T02560189 PWINO TSNO
May 4, 2007 Greensburg KS. Whereas this ob sports the ONLY sub-20ºC Td in this whole lot, it must be pointed out that the station was WAY too far east to record what the tornado was actually ingesting. That event featured a narrow fetch of elevated theta-e, perhaps only 50 or so miles wide, and the DL was retreating at the time.

KJLN 110522_2053Z 19013G19KT 7SM FEW032 SCT045 BKN060CB 27/21 A2968 RMK AO2 PK
WND 20027/2011 SLP039 T02670211 56013
May 22, 2011. Joplin. I would have preferred the 2153Z ob, but the Plymouth site only offers 2053 and 2253 (which was post-event).

KOKC 110524_2128Z 14022G32KT 4SM TS HZ FEW034CB BKN044 OVC150 26/21 A2944
May 24, 2011, EF5 that crossed I-40 in west-central OK.

KCHK 110524_2155Z AUTO 15018G23KT 2 1/2SM +RA SCT011 BKN043 OVC070 25/22 A2943
RMK AO2 VIS 1 1/2V5 P0012
May 24, 2011 EF4 that tracked near Chickasha OK.

KOUN 130520_2035Z AUTO 19003KT 10SM TS 26/22 A2968 RMK AO2 LTG DSNT ALQS
May 20, 2013 Moore OK tornado.

KSLN 130528_2153Z 16027G33KT 10SM FEW042 SCT050 30/22 A2957
May 28, 2013, over-an-hour in one freakin' spot Bennington KS EF4 wedge.

KOKC 130531_2252Z 15018G23KT 10SM TS FEW050CB BKN150 BKN300 28/22 A2957
May 31, 2013 El Reno OK.

Apparently, if you want a violent tornado, you probably ought to bring a Td of at least 68ºF or so to the table. Other than really hot days, I don't think there's much point in focusing on T's, because they change fast in a storm environment.

Anyone want to dig up some more examples?

 

[Shawn Schuman]

Random selection of events, both within and outside of the Plains at various times of year.

March 13, 1990 Hesston, KS F5
KHUT 900313_132200Z 17018G23KT 10SM SCT015 BKN050 20/17

March 2, 2012 Henryville, IN EF4
KSDF 021956Z 19013G22KT 10SM -RA FEW044 BKN110 23/14
KLOU 021953Z 19008G21KT 10SM FEW042 SCT049 SCT110 24/15

March 2, 2012 Crittenden, KY EF4
KCVG 022106Z 14011KT 5SM -TSRA BR FEW009 SCT040CB OVC090 14/13

June 5, 2010 Millbury, OH EF4
KTDZ 060305Z AUTO 18009KT 2 1/2SM +TSRA BR BKN029 BKN042 OVC080 23/22

February 10, 2009 Lone Grove, OK EF4
K1F0 110050Z AUTO 17011G15KT 10SM SCT022 SCT027 22/18

June 17, 2010 Wadena, MN EF4
KADC 172145Z AUTO 19017G36KT 3/4SM +TSRA 21/19

February 5, 2008 Clinton, AR EF4
KRUE 052248Z AUTO 25006KT 2 1/2SM +TSRA BKN028 OVC034 21/18

May 25, 2008 Parkersburg, IA EF5
METAR KALO 252154Z 14014KT 5SM VCTS HZ SCT019 BKN023 BKN028 24/21

June 24, 2003 Manchester, SD F4
KHON 242311Z 03010KT 9SM TS SCT029 28/23

June 22, 2007 Elie, Manitoba F5
METAR CYPG 222100Z 22012KT 170V240 15SM FEW042TCU FEW100 31/20

So I suppose it comes down to whether you're talking about violent tornadoes in general or specifically in the Great Plains during chase season. I may poke around more later, but I doubt you'll find many violent tor cases during the typical chase season with dews lower than, say, 65-68º. That's a pretty good rule of thumb. At other times of year and/or in other areas of the country, however, you certainly can have violent tornadoes with significantly lower dews so long as there are other compensating factors.

ETA: Of course there's also the caveat that Bob hit on re: Greensburg; moisture can vary pretty significantly over small temporal and spatial scales at times, so these obs may not represent the actual near-storm environment. Still, it ought to be close enough for the most part.

 

[JeremyS]

I would think the closer the better for temp/dew pt spreads with violent tornadoes, and of course just tornadoes overall. If you have temps in the 90s and dews in the mid 60's that is a pretty large spread and you are going to be looking at pretty high LCL's. I seem to remember in the past on most days like that, the better tornado potential will come in the evenings when the surface temps cool into the 80s and the dews rise to 70 or even higher.

Seems like dewpoints in mid 60's or higher and temps 80's to 90's usually occur with violent tornadoes.

 

[Paul Knightley]

I suppose you also need to consider the t/td spread (and the values, per the original question) of the inflow region of a tornadic storm. It may be that the values you look up are typical of the overall moist sector, but the tornado you're interested in formed after the parent supercell crossed a mesoscale boundary of some kind.

 

[Rob H]

I would think the closer the better for temp/dew pt spreads with violent tornadoes, and of course just tornadoes overall. If you have temps in the 90s and dews in the mid 60's that is a pretty large spread and you are going to be looking at pretty high LCL's. I seem to remember in the past on most days like that, the better tornado potential will come in the evenings when the surface temps cool into the 80s and the dews rise to 70 or even higher.

Dews rising in the evening also happens with a strengthening low-level jet, which isn't a bad thing to have for violent tors.

If you look at Davies' chart comparing significant tornadoes to EHI, you'll see some of the well-known violent tors happening with 3000+ MLCAPE. You need to have pretty high surface dewpoints to achieve that given "standard" thermodynamic profiles in near-storm environments.