What is CAPE??

[Zach Bailey]

Another good place to find various cape values...and lots of other super-useful information (probably the single most useful chaser site on the web), SPC's Mesoanalysis:

http://www.spc.noaa.gov/exper/mesoanalysis/

 

[Mike Smith]

Here is a "meteorologically incorrect" thought from someone who has been forecasting severe storms for 35 years. CAPE is the most over- analyzed parameter in meteorology.

What, exactly, do you forecast differently when the CAPE is 3,325 versus 3,000? The answer is, nothing. Nor should you. In the late 1960's to around 1970, NSSL had a mesonetwork of hourly rawinsondes and found there were very signficant spatial and temporal variations in lifted index (more in a moment) in central and western Oklahoma during severe weather situations. With today's rawinsondes 250 miles apart and twice/per day, trying to recalculate CAPE to the tens and units places (and I have seen forecasters use programs to do that on many occasions!) is silly.

What, exactly, is the advantage of CAPE versus the lifted and Showalter indices? The latter two can be easily calculated by hand. Supposedly CAPE is "more precise." OK, but what do you forecast differently for a CAPE of 3,500 versus a lifted of -7? Again, given the documented spatial and temporal variations in these parameters, what does the "more precise" CAPE yield in terms of forecast value?

If the day comes (and I hope it will) where satellite or profiler technology will allow much more frequent soundings at much closer intervals than we have today, that are more accurate than current technology allows, we may well learn that using CAPE is a signficant step forward. In the meantime, you can save a lot of time and suffer from less information overload by just doing a quick LI and SI calculation, especially when in the field.

 

[Jeff Snyder]

What's the advantage of CAPE over LI? Lifted index is a measure of temperature difference at one fixed level (500mb usually). CAPE, on the other hand, is in integration of positive buoyancy (Tv parcel > Tv environment) through the entire layer... LI can be quite deceptive sometimes, such as for those cases with a warm layer near 500mb. Or when you are dealing with low-topped supercells in which the EQ level may only be 500mb (in which case LI could be 0), despite a fat CAPE profile between the LCL and the EQ level. Yes, you can't easily 'eye-ball' CAPE, so for that regard, LI is EASIER. However, I think the strength of CAPE in terms of proving a more accurature picture of potential energy available to an updraft makes it considerably better than a fixed-level index such as LI or Showalter's-Index.

 

[Mike Smith]

More accurate for what?

I re-ask my original question: What do you forecast differently with a CAPE of 3,250 versus 3,000? Especially (in the real world) if you were to launch a rawinsonde in the same airmass at the same time 20 miles away, the CAPE might vary by 1000 j?

I have seen times when computer-generated CAPE (I did not look at the sounding myself) was zero when the SI was negative and snow thunderstorms developed (granted, it may have been a problem with calculating CAPE in winter). Often, when we have to rely on a computer to do calculations, this type of error can occur.

My point is we spend far too much time worrying about CAPE, MUCAPE, etc., etc.

 

[Glen Romine]

I would agree with Mike that CAPE does seem to be overused, particularly model forecast CAPE. I'm not sure I'd sell out on LI and SI though - but really what seems most important is whether there is a thermodynamic environment supportive of deep convection. LI gives surface instability, and Showalter gives elevated instability in typical cases. However, because of the use of fixed layers in the calculations - you will likely miss a few events (e.g., when the moisture isn't at 850 mb), but for basic forecasting I agree it is probably the quick and dirty approach. CAPE and MUCAPE can serve as substitute for similar information, the former having similar problems with defining the surface parcel characteristics as in LI, and MUCAPE capturing elevated storm potential similar to SI.

I prefer CAPE and MUCAPE for the reasons Jeff mentioned - but really these parameters are low on my list relative to other forecast problems. Model forecasts of instability are often poor, and finding regions with otherwise favorable conditions and at least the 'potential' for parcels to become buoyant (even if the forecast model doesn't show it) is likely to work better, at least it does for me. Otherwise, small changes in moisture distribution, cloud cover, temperature, precipitation coverage, etc.... can lead to wholesale changes in instability profiles, and ruin a model forecast of them.

Glen

 

[Jeff Snyder]

More accurate for what?

I re-ask my original question: What do you forecast differently with a CAPE of 3,250 versus 3,000? Especially (in the real world) if you were to launch a rawinsonde in the same airmass at the same time 20 miles away, the CAPE might vary by 1000 j?

Mike, what do you forecast differently with an LI of -6.5 vs. -6? Both your question and my own are dealing with instability 'increases' of ~8%... But again, are you going to forecast differently for -6.5 than you would for -6?

I will agree that, because of the larger scale of CAPE (in the thousands) versus LI (in the single digits and occassionally double-digits), people tend to overanalyze model output (e.g. decide a target area based on ETA-forecast 2700 SBCAPE vs. 2500 SBCAPE), but from an objective standpoint, I certainly think that CAPE provides a 'better' view of potential instability. Yes, there's no way to eyeball it with much precision, so in that matter, LI is likely better if you are sitting out in a field looking at an 18z sounding. Then again, if you're out in a field looking at an 18z sounding, you obviously have access to a computer systme of some sort, and likely have access to CAPE calculations, as most sites put CAPE values on the sounding image.

Is it overused? Maybe... But instability is one of the key "ingredients" for deep convection... I tend to keep CAPE forecasts in mind when making a target area, since model forecast instability parameters are really all we have... I mean, you could just randomly pick a date and location sometime and hope for some action, but I'd rather take what we have (observations, model output, etc) and go from there... Yes, model forecasts of CAPE are sketchy, but again, that's the best we have. Until we have a surface observation network with very high spatial and temporal resolution (say, every 5-10 miles with obs every 5 minutes) nationwide, we have to make with what we have. Heck, one could argue that many convective forecast parameters (CAPE, helicity, bulk shear, etc) are meaningless given the high variability of said parameters, and given the paucity of the upper-air sounding network in the U.S. (and across the world for that matter), though that would make target-area forecasting success align more with luck than skill...