Relationship between dewpoint/heat index and CAPE

[Elaine Spencer]

I apologize if this question has been asked before or seems a little silly to the more experienced people on this forum, but I'm curious.

I know that CAPE is a measure of instability or potential energy and that the higher it is the more potential for severe weather including tornadoes. Heat and humidity are, of course, some of the major ingredients for CAPE.

For the better part of two weeks now we have been experiencing really, really hot and sticky weather in central Illinois -- highs in the mid to upper 90s combined with dewpoints in the upper 70s and even cracking 80 at times, yielding heat indices in the 105-110 or higher range. There have also been some storms to our north and even some outflow boundaries passing through our area, but so far, nothing severe here (Springfield area).

Now I can recall other days like these, in terms of heat and humidity, in northern and central Illinois that produced strong to violent tornadoes, such as 7/13/04 (Roanoke F4) and 8/28/90 (Plainfield F5). It is my understanding that the CAPE levels in both cases were extremely high, over 6000 j/kg in '04 and perhaps over 8000 at the time of the Plainfield tornado.

Furthermore I recall reading somewhere that during the deadly Chicago heat wave of 1995, CAPE levels soared to something like 8000 or 9000 j/kg, but no storms developed because there was nothing around to trigger them (although there were several waves of severe storms/derechos in the upper Midwest at the time).

So I would think that right now, CAPE levels in central and southern Illinois would be off the charts, but from what I've seen on the SPC mesoscale discussions and whatnot, they don't seem to be all that high, "only" about 1500 j/kg at most. Can anyone explain this?

 

[rdale]

The CAPE at 7pm (00Z) in Springfield IL was about 3500.

http://w1.spc.woc.noaa.gov/exper/soundings/10081200_OBS/ILX.gif

That's fairly high. On the other hand, you are only looking at the surface features - and CAPE is computed using the entire atmosphere. So you want colder air aloft to "enhance" the CAPE.

 

[Elaine Spencer]

Thanks. I was going off a couple of earlier CAPE readings that showed us well under 2000. 3500 is plenty high for producing storms but still not as high as I would have expected it to be, and, there are no storms forecast tonight. Must be lack of cold air aloft like you said, plus, lack of sufficient triggers such as fronts, low pressure areas, etc.

Another "just curious" question: anyone know what the highest CAPE reading ever recorded was? Has it ever been higher than 10,000 j/kg anywhere?

 

[Jeff Duda]

Like Rob said, there are more things than surface temperatures that control CAPE. CAPE is the vertical integral of the difference between the atmospheric temperature and that of a parcel lifted from a given layer, aka CAPE = int(g(T-Tbar/Tbar)*dz,z=LFC,z=EL), where LFC = level of free convection and EL = equilibrium level and each point corresponds to the levels of "positive area" on a sounding. See http://www.theweatherprediction.com/habyhints/305/ for reference on the definition. Basically, although surface temperature and dewpoint (and really, theta-e), determine the path the parcel will follow (for the computation of surface-based CAPE...there's also mixed-layer CAPE and most unstable CAPE), the key is how much warmer the parcel is than the environmental temperature at each level in the atmosphere. During the middle of the summer, although surface theta-e is generally at its seasonal highest, so are temperatures at levels above ground such as 850 mb, 700 mb, and 500 mb. Thus, the temperature difference may not be that great despite the high temperature and dewpoint.

I do not know the greatest CAPE value ever measured or analyzed. However, on July 14th of this year, RUC analyzed surface-based CAPE briefly topped 8000 J/Kg in a small area of central Iowa, seen here. The RUC did predict that > 10,000 J/kg would exist on this day, but it typically overdoes surface moisture, and thus surface-based CAPE as well. One of the highest CAPE values ever seen in a sounding came from Davenport, IA last year on June 18th, seen here. Values like these are rare, but if you look often enough during this time of year you will probably see it at least once during each year in a small area. I recall seeing an area of RUC analyzed surface-based CAPE exceeding 8000 J/kg on a midsummer day in North Dakota in 2007 as well. If you really want to look, there is an archive of SPC mesoanalysis graphics at http://www.spc.noaa.gov/exper/ma_archive/images_s4/ that goes back to 2005 you can check.

 

[jshields]

here in omaha, as in most of the upper midwest we've had a good 4 days of extreme heat and humidity, i.e. temps above 95 and heat indexes over 110. i always like to take a look at the sounding just to see how much CAPE is available each evening. this was the sounding from this past sunday(you'll have to click on omaha):
http://w1.spc.woc.noaa.gov/exper/soundings/10080900_OBS/

surface cape 7006 AND with 0-1km srh of almost 200! man it would have been nice to get something going rooted near the surface. it also shows a temp/dewpt of 94/79.

a couple of days later this was the 0Z sounding(again click on omaha):
http://w1.spc.woc.noaa.gov/exper/soundings/10081100_OBS/
note the obs of 92/77, a difference of 2 degrees, but the surface CAPE was "only" 5537. the mid level lapse rates were a fair amount steeper on the day with the higher CAPE owing to the slightly cooler midlevel temps that day

 

[jshields]

Either those are hot links or you mislinked them because your description does not match up with the images.

sorry, the default site for the soundings page is norman, so you have to click on the star for omaha, ne, to get the correct graphics!

 

[Jeff Duda]

here in omaha, as in most of the upper midwest we've had a good 4 days of extreme heat and humidity, i.e. temps above 95 and heat indexes over 110. i always like to take a look at the sounding just to see how much CAPE is available each evening. this was the sounding from this past sunday(you'll have to click on omaha):
http://w1.spc.woc.noaa.gov/exper/soundings/10080900_OBS/

surface cape 7006 AND with 0-1km srh of almost 200! man it would have been nice to get something going rooted near the surface. it also shows a temp/dewpt of 94/79.

a couple of days later this was the 0Z sounding(again click on omaha):
http://w1.spc.woc.noaa.gov/exper/soundings/10081100_OBS/
note the obs of 92/77, a difference of 2 degrees, but the surface CAPE was "only" 5537. the mid level lapse rates were a fair amount steeper on the day with the higher CAPE owing to the slightly cooler midlevel temps that day

Okay, now that I've taken care of the idiot in me, I had a chance to look at the soundings side by side. It turns out that the just-below-500-mb temps are nearly identical in both soundings, as are the sfc-3km lapse rates. Granted, the other lapse rates in the sounding with only 5500 CAPE are slightly lower, but only by 0.2 to 0.5 K/km. But, the 5500-CAPE sounding has a slightly lower LFC. I think the big difference in CAPE between these two soundings is due to the tropopause-level temps. Between 100 and 200 mb, temperatures are much different between the two soundings.

 

[jshields]

here we go from tonight: http://w1.spc.woc.noaa.gov/exper/soundings/10081300_OBS/

again you'll have to click on omaha, as norman is the default location. that's quite the sounding!! surface CAPE at 7679 with a temp/dewpt of 93/81!! lapse rates were higher yet(over 8.0 C/km). i think that's the highest CAPE i've seen on a sounding this summer. again 0-1 km srh was near 200, and 0-1 shear was 25 kts. very impressive again for a whole lotta nothing going on

 

[Jeff Duda]

here we go from tonight: http://w1.spc.woc.noaa.gov/exper/soundings/10081300_OBS/

again you'll have to click on omaha, as norman is the default location. that's quite the sounding!! surface CAPE at 7679 with a temp/dewpt of 93/81!! lapse rates were higher yet(over 8.0 C/km). i think that's the highest CAPE i've seen on a sounding this summer. again 0-1 km srh was near 200, and 0-1 shear was 25 kts. very impressive again for a whole lotta nothing going on

The lack of activity may be due to a general lack of forcing and the 22 SBCIN. If only that amount of CAPE was associated with a sounding that had a well mixed boundary layer. Then we'd be in business.

 

[davecarlsen]

Another "just curious" question: anyone know what the highest CAPE reading ever recorded was? Has it ever been higher than 10,000 j/kg anywhere?

I can attest that on July 25, 2007, Winnipeg had a calculated SBCAPE of about 10,000 J/kg (I can't remember the exact number). The T/Td was 34/27 (93/81). That was the same day that the Canadian humidex record was set in Carman, just southwest of Winnipeg. The temperature was 92F (33C) and the dewpoint 86F (30C). Their SBCAPE was similar.

And by the way, a storm went that day. Unfortunately (or fortunately, depending on which way you look at it) there was about 10 knots of deep shear. Storm went up, went down, and was done, all in 45 minutes.

 

[jshields]

The lack of activity may be due to a general lack of forcing and the 22 SBCIN. If only that amount of CAPE was associated with a sounding that had a well mixed boundary layer. Then we'd be in business.

yeah without a doubt the problem was lack of forcing. there was a prefrontal trough that hung out in central nebraska yesterday, but nothing close to here. i know there are numerous days like this throughout the summer everywhere, where there just isn't enough forcing to get something going. what a shame!

 

[Chris Vagasky]

I can attest that on July 25, 2007, Winnipeg had a calculated SBCAPE of about 10,000 J/kg (I can't remember the exact number). The T/Td was 34/27 (93/81). That was the same day that the Canadian humidex record was set in Carman, just southwest of Winnipeg. The temperature was 92F (33C) and the dewpoint 86F (30C). Their SBCAPE was similar.

And by the way, a storm went that day. Unfortunately (or fortunately, depending on which way you look at it) there was about 10 knots of deep shear. Storm went up, went down, and was done, all in 45 minutes.

There's no upper air station there...nearest one I can find is for Bismarck, and they had about 3300...RUC had estimates of >7000.

Where is the actual calculation that you speak of?

 

[afischer]

The lack of activity may be due to a general lack of forcing and the 22 SBCIN. If only that amount of CAPE was associated with a sounding that had a well mixed boundary layer. Then we'd be in business.

Using a lowest 100mb ML parcel in this case (not atypically) certainly produced different results... and stronger CIN (60 J/kg) given the richest moist layer was only ~50mb deep as you noted. And given the moist boundary layer (ML dewpoint of 75F), if one opts not to apply the virtual temp correction the MLCIN jumps up to a substantial 122 J/kg (compared to the only 22 J/kg Tv-corrected SBCIN). I sometimes question how well applying the Tv correction (though necessary to apply in theory) "represents the real world" in some of these null events/bust cases characterized by small to moderate Tv-corrected CIN and strong uncorrected CIN. For example, parcel theory assumes little to no mixing/entrainment... how realistic is this?

very impressive again for a whole lotta nothing going on

Thursday's convective evolution was pretty typical of many summertime "bimodal" severe cases in which severe convection (sometimes elevated above a sfc stable layer, though not in Thursday's event) initiates in a regime northeast of the ~12-14C 700mb isotherms; and then farther southwest in deep/hot boundary layer mixing (e.g. sfc temps 100-110F) you get high-based storms on the elevated terrain of the High Plains (amidst 15-18C 700mb temps) with sporadic wind damage and some hail. Then in the middle of these two regimes (e.g. Omaha vicinity Thurs), near the extreme instability max, you have a minimum in convection if not convective failure altogether. Granted, the prefrontal trof that you mentioned was pretty modestly convergent compared to points N and SW... and as was alluded to by Jeff, the entire event was being modulated on the synoptic scale by rather complex processes (e.g. the early band of mid-level convection passing N that sprouted into severe sfc-based storms; some weak low-amplitude s/w trofs likely approaching from the W very late; etc.) that more than likely weren't particularly favorable for upward vertical motion in the area.

More related to the topic of the original thread, there have certainly been some very moist and explosive environments sampled this summer, especially via the OAX RAOB. I believe it was the 00Z 08/10 OAX RAOB that sampled a 100mb ML dewpoint of 78F (another interesting day with an only weakly convergent remnant OFB and unfavorable background pattern in which storms waited until after dark to fire)... which is the highest boundary layer moisture I've ever seen at Omaha (though I've seen ML dews top 80F at DVN and ILX a time or two). About a month ago (00Z 07/15), one of OAX's most unstable soundings of the year (MLCAPE 5500 J/kg), just downstream of linear storms along a cold front, was also characterized by some of the highest PWATs ever observed (THE highest ever observed?) at OAX (~2.5")!

 

[Jeff Duda]

Using a lowest 100mb ML parcel in this case (not atypically) certainly produced different results... and stronger CIN (60 J/kg) given the richest moist layer was only ~50mb deep as you noted. And given the moist boundary layer (ML dewpoint of 75F), if one opts not to apply the virtual temp correction the MLCIN jumps up to a substantial 122 J/kg (compared to the only 22 J/kg Tv-corrected SBCIN). I sometimes question how well applying the Tv correction (though necessary to apply in theory) "represents the real world" in some of these null events/bust cases characterized by small to moderate Tv-corrected CIN and strong uncorrected CIN. For example, parcel theory assumes little to no mixing/entrainment... how realistic is this?

My apologies for hijacking this thread, but lots can be learned here.

Since the buoyancy force that drives convection is derived from density differences, I would assume that applying the virtual temperature correction when computing sounding thermodynamic parameters is very approprate and better than not applying it. Also, parcel theory is known for a few deficiencies, one of which being the assumption of no entrainment. Entrainment pretty much always occurs in convective plumes, and can be quite significant. I have an equation from one of my grad courses for an equation for the moist-adiabatic lapse rate when entrainment is accounted for and it's pretty large. Parcel theory also does not account for compensating subsidence outside of an updraft region.

 

[Lanny Dean]

Is there anyway we can "sticky" this thread? This is a great thread! Jeff, thanks for all your input... I really enjoyed the read and so wish we had more threads like this one!