A Look at the Energy-Helicity Index
by Jon Davies
Storm Track, May-June 1995
© Copyright 1995 Jon Davies
For supercells to develop, wind factors (such as helicity) and thermodynamic factors (such asCAPE, or convective available potential energy) must combine to produce an environment that isfavorable for the formation of rotating thunderstorms. Therefore, one important aspect of forecastingthe likelihood of supercells (and by association, tornadoes) is estimating and assessing the combinationof helicity and instability.
Storm enthusiasts who are focused on supercell thunderstorms in the plains in the warmseason sometimes fail to recognize that there are a variety of wind profile/instability combinations thatcan set the stage for supercell development. For example, in the southeast United States in the coolseason (late fall, winter, early spring), tornadoes often occur within stability that at first glancemay appear too weak to support significant supercells and possible tornadoes. When Bob Johns was a leadforecaster at NSSFC in Kansas City, he and I worked together on a project computing instability andhelicity for a data set of tornado cases from the 1980's. We found 35 cases when tornadoes occurredin environments where instability appeared weak (CAPE less than 1000 J/kg, using an average liftedparcel in the bottom 100 mb (or roughly 3000 ft). Nearly all these cases occurred in the eastern UnitedStates in the cool season, and were associated with large helicity values, most greater than 400 m2s-2,and some in excess of 600 m2s-2.
The opposite situation occurs in the plains in late spring and summer. Bob and I found 29 casesin our dataset where tornadoes occurred in environments where CAPE was very unstable, greater than3500 J/kg-1 (again lifting the average parcel in the bottom 100 mb of the sounding). Nearly all of thesecases were in the plains in the warm season, and more than half had relatively low helicity values, near200 m2s-2 or less.
In the weak instability cases, what appears to happen is that strong wind fields associated withlarge helicity values can (apart from their potential to induce rotation on a storm updraft) set up strongvertical pressure gradients within a storm that increase the strength of the updraft, even if the supportinginstability is relatively weak by "plains" standards. In the strong instability cases, it seems that, becausethe highly unstable environment can result directly in strong updrafts, less helicity is needed in the windenvironment to generate rotation in a storm's updraft. Many tornado cases fall into a middle groundbetween these two extremes.
So... how does one estimate and assess whether there is enough instability in an environmentwith a certain amount of helicity to potentially produce rotating storms and possible tornadoes? A usefultool is the energy helicity index (EHI), developed by John Hart and Josh Korotky, based on the dataset of tornado cases that Bob Johns and I looked at.
The EHI is a simple equation that combines helicity and instability into one number forestimating and assessing these factors in a particular environment regarding potential for supercells.It is not a magic number or a "dynamite tool" for forecasting tornadoes. It is simply another pieceof information that can be useful at times in forecasting supercells. It works well in some situations,and not so well in some others. This depends on many factors, such as the strength of winds higherup in the atmosphere and the location of sounding observation network sites.
The EHI equation is: EHI = (CAPE x H) - 160,000
In other words, multiply CAPE times helicity (H), then divide this quantity by 160,000. The EHI seemsto work best if one uses a conservative estimate of CAPE, lifting an average parcel in the bottom 50 or100 mb of a sounding. This is because relatively dryer air above the surface normally mixes with a warmmoist parcel as it ascends, making the actual CAPE somewhat less than the largest CAPE value one can"coax" out of a given sounding, often by using a surface parcel or some other method.
Let's take the OOz 4/26/94 Stephenville sounding from my preceeding article in this issue, andcompute an EHI value. Here's the temperature data from the mandatory sounding levels:
Sfc (955 mb), T = 28C (83F), Td = 18C (65F)
925 mb, T = 25C, Td = 17C
850 mb, T = 19C, Td = 15C
700 mb, T = 8C, Td = -1C
500 mb, T = -12C, Td = -32C
400 mb, T = -25C, Td = -42C
300 mb, T = -42C
200 mb, T = -58C
150 mb, T = -63C
Without going into the detail and mechanics of computing CAPE values(which Tim did last issue, see Storm Track January-February 1995),my conservative CAPE estimate is 2770 j/kg-1 for this sounding.
With this information, and a helicity value of 240 m2s-2 computed fromthe same sounding in my article earlier in this issue, the EHI equationbecomes: EHI = (2770 x 240) / 160,000 = 4.2
The EHI here is 4.2, and has no units, because m2s-2 and J kg-1 arebasically equivalent, cancelling out.
In practical experience, I've found that EHI valuesapproaching 2.5 or greater are significant andtend to be more indicative of supercells withpotential for tornadoes. In this context, the EHI from theStephenville sounding suggests potential for supercellsand possible tornadoes in this environment.Indeed, storms 50 miles to the northeast of the soundingsite produced tornadoes later that evening,including the F4 Lancaster, Texas tornado. Comparatively, otheravailable soundings (see the mapbelow) throughout the central plains that evening showed EHIvalues of much less magnitude (see diagram):
Other significant tornado activity occurred in the Talihina, Oklahomaarea that evening. Unfortunately, the closest soundings were 150 milesor more away. Still, the EHI of 2.4 at Longview, Texas suggests thatsignificant helicity and instability may have extended into southeastOklahoma.
This particular April day involved a large and dynamic weather systemwith strong wind fields coming out through the Central U.S. As one movesmore toward summer, weather systems become less intense while instabilitybecomes more widespread through the plains. When this happens, theEHI often becomes less useful as more localized factors come into play.
An example is the May 28, 1994 LP supercell and tornado inRoberts County of the Texaspanhandle (see Storm Track, November-December 1994). Factorsmeasured by the EHI may havebeen less relevant in this localized case. And, with Amarillo tothe west in dryer air, and Norman farto the east, none of the of the conventional network soundings werelocated properly to sample theenvironment of this storm. Soundings and model forecast data (notshown) didn't do much this dayto suggest supercell or tornado potential when looking at bothhelicity and instability. This emphasizesthat the EHI is only one of several complex factors that maybe relevant or useful in supercell forecastingsituations, and, like other parameters, is also at the mercy of theobserving network density.
For those familiar with the surface-based lifted index (SLI) as ameasure of instability, I'vedeveloped a version of the EHI using surface-based lifteds:EHI = ((-SLI x 322) - 208) x H) / 160,000
In other words, reverse the sign of the SLI (if it is negative, make it positive), multiply this by 322 andsubtract 208. Then multiply this quantity by the helicity (H), and divide by 160,000. Even though we'reusing a surface parcel here, I've adjusted the equation so that the instability is treated conservatively toencourage reasonable estimations. This version is not as accurate as using CAPE, but is still useful inproviding estimates when one doesn't have the temperature and dew point profile detail needed forcomputing CAPE from a sounding.
This version of the EHI allows one to use the relatively coarse "FD" forecast winds andtemperatures aloft (see my article on hodographs in the January-February 1994 issue of Storm Track)from the NGM model to make a crude estimate of helicity, instability, and EHI values in the absenceof soundings. When combined with surface data, such computations can sometimes provide usefulinformation. This depends, of course, on how localized the situation is and how well the models areforecasting winds and temperatures aloft (unfortunately, at times the FD winds can be very much inerror). Keep in mind, too, that capping inversions don't care about EHI values. A strong capping layerthat prevents storms from developing makes EHI values a moot point; if no storms fire, then there can'tbe any supercells.
To show how the EHI can be useful on days when the models are handling systems reasonablywell, the following page presents some maps showing helicity, surface-based lifted index, and EHIcomputed by combining observed surface data with FD wind and temperature forecasts aloft. Theseforecasts are from the infamous "Wedge Week" in 1993 (May 5-9, one map for each day) when on5 consecutive days numerous tornadoes, several of strong and violent intensity, occurred at variouslocations in the plains. Notice how significant helicity values may be widespread, but when combininginstability with helicity, an EHI estimate can sometimes help to narrow the area of concern regardingpotential for rotating storms and possible tornadoes. While the higher EHI values on these analysesdo a reasonable job of highlighting general areas of tornado occurence, do keep in mind that fairlydynamic weather systems were involved here. In less well-defined situations, the EHI will likely be lessuseful for some of the reasons mentioned earlier, particularly when the model forecasts are poor.
Hopefully this discussion will give interested readers, chasers, and forecasters an introductoryidea about how the potential for supercell development depends to a significant degree on theinteraction of wind factors and instability. In this respect, the energy-helicity index is a parameter thatcan be helpful.