• While Stormtrack has discontinued its hosting of SpotterNetwork support on the forums, keep in mind that support for SpotterNetwork issues is available by emailing [email protected].

Strategy for forecasting storm initiation?

Rob H

EF5
Joined
Mar 11, 2009
Messages
825
Location
Twin Cities, MN
Would anyone mind sharing their strategy for forecasting the time of convective initiation? In preparation for the upcoming season I'm realizing this is definitely one of my weak areas.

Right now, I look at the 00z model output and try to paint a best mix/bulls-eye of moisture, lift, instability, and shear. That only helps me identify a target at 00z and doesn't help me plan my several hundred mile drive at all.

My initial thoughts are that the cap is the key, and that I either need to estimate the erosion of the cap or the presence of significant lift. Any pushes in the right direction would be much appreciated.
 
Robert, I will try to give you my two cents on how I do it....

First if I know I have a potential chase day coming up I will start dissecting the model runs. Especially 00z and 12z. Whichever model(s) have the best handle on the system are the ones I will tend to slant my guess towards. As the event gets closer the models will give you 6, 3, or 1 hr intervals depending on the model. From there I watch when they have precip breaking out and if I think capping will be a problem I will start looking into the soundings of surrounding areas.

On the day of the chase I tend to look at the RUC and really start watching the satellite, surface charts, soundings etc to determine where the best parameters are at and if I can find a good boundary to get close to.

I'm far from perfect in my chase success, but I feel like this strategy has worked ok for me. I too have to make a several hundred mile drive to chase the good stuff in God's country. Forecasting capping is hard. When driving this far it is also hard to predict when everything will be a good match for severe weather/tornadoes. Its frustrating to drive all day and know within an hour or two that the CAPE won't be sufficient enough to break the cap or keep updrafts sustained in high shear environments.
 
Robert, its important to have a decent grasp of the meteorology involved, but there are several plots I use to get a quick idea of when initiation might occur:

From Earl Barker's NAM severe parameters:
http://www.wxcaster.com/smallfiles_central_svr.htm
http://68.226.77.253/models/eta/central/CENTRAL_ETA212_ATMOS_LSI_00HR.gif
The cap index is showing you the width of the capping inversion as plotted on the sounding in degrees Celsius. Its the opposite of a lifted index, measuring stability instead of instability. Assuming the upper level support has arrived, I look for values to drop into the greens, yellows, and reds for initiation (near 0 C). If the plot is fairly accurate this can be a good indicator of when initiation will occur. If the cap index opens up for only an or two and then fills back in, this often signals a bust chase, as the window of opportunity is too narrow, and if storms do get up they often become elevated once the cap fills back in. Unless there is a very sharp boundary with strong shear, anything above 3 C is probably going to be a cap bust as well. If the values are in the negatives all day and there is already a jet overhead, expect ongoing storms throughout the day.

You can also cheat and just look at the 4km WRF simulated reflectivity:
http://www.emc.ncep.noaa.gov/mmb/mpyle/cent4km/conus/00/
They call it the God model as it verifies with spooky accuracy sometimes. I've seen it fire off storms on cap bust days though, but when its on its usually within an hour or so of initiation. Its nice to understand the lifting mechanisms and dynamics behind the initiation though.

I'll also look at vertical velocity plots. When there are spikes you usually have storms underway. COD has some nice plots of those: http://weather.cod.edu/forecast/
The precip plots or qpf are based off these vertical velocities, which Kevin mentioned, and I'll look at those too. If there is doubt though, I just make sure I'm out there several hours early and follow the progression of the surface features.
 
Last edited by a moderator:
I have learned the hard way over and over, that u need to be in the general area a few hours before hand. Plan on being early to give yourself enough time to adjust. With that being said, convective initiation is one of the hardest things in forecasting. We still dont have any solid ways to forecast this and its an area of active research.

The things mentioned above are useful and can help. You will also learn to keep a good eye on the satellite. Pay attention to Cu fields and boundrys. If theres a strong cap you want to make sure there will be forcing around otherwise it may not be worth going out at all.
 
The links that Skip have provided are very useful. Forecasting a few days in advance I like to look at the lid strength index. This gives you a rough guide of what the cap may do. You want the cap to be fairly strong through most the day, but breaking down by afternoon. Also, like Skip mentioned you don't want to see a breaking cap by evening time for only a narrow 2-3 hour window and then filling in after this. This definitely has cap bust all over and storms will not be able to root and become surface based to enhance the tornado potential. Also, you don't want negative reading all day as there is basically no cap and storms will be ongoing all day which will also decrease moisture and instability in the forecast area.

What I do that seems to work is get into a general target area in the afternoon to allow for adjusting. Look at visible sat. to see where areas of cumulus clouds are developing. I also look to see where the best surface heating is going on free of clouds over the area or even precip. Keep an eye on the lifting mechanism you are focusing on whether it's a warm front, cold front, dryline or triple point. You don't want to be too far from these features as this is close to or right where initiation will occur.

The 4km WRF base reflectivity is scary reliable like Skip mentioned. Keep up with the RUC the day of the chase and see when it breaks out precip. Also, you may want to look at the SPC mesoanalysis page. Bottom line keep and eye on visible sat, radar for echo returns and a focus on your lifting mechanism. Sometimes will cap bust and others not. It's very hard and discouraging to predict.
 
Predicting initiation time is probably the singularly most difficult part of storm forecasting, as it's hard enough figuring out whether the cap will break or not. Predicting initiation time demands that we refine this question further, so my approach has always been only to forecast whether storms will break early, midday, late, or not at all. There is no reliable technique to use as it differs from one situation to the next. In essence, the time that storms go up is dependent on: (1) how soon enough heating occurs to get a parcel to its LFC, (2) how soon enough "dynamic lift" occurs to cool portions of the troposphere where a parcel can arrive at its LFC, and (3) how soon enough low-level forcing occurs to bring the parcel to its LFC. Most situations have a mix of all three and this makes it no small challenge.

In my own forecasting, I assess #1 and #2 independently. I will look at soundings across an area along with forecast maximum temperatures to see whether the parcels have any shot of making it to their LFC, given no real change in the environmental sounding except for heating. This in turn is not very cut and dry, since there can be local areas of enhanced lift (like the Caprock or Raton Mesa) that can make this happen earlier than expected, and there can be things that play havoc with parcel ascent like mixing with dry layers beneath the cap. But in most respects this involves a basic convective temperature forecast -- but actually working it out on the sounding rather than going to a website that spits out Tc numbers. This gives me some idea whether convection will occur early, late, or not at all.

Then I will look at the strength of lift coming in using the model forecasts... this includes the vorticity advection and div q charts (if available... it is annoying these days trying to find good UA diagnostic charts, though I think DuPage carries them) as well as conceptual models (e.g. jet streak quadrants and adjusting for these in anticyclonic/cyclonic flow). This gives me some idea how the soundings might further destabilize. You can crosscheck this by looking at the model's forecast temperature changes with time at key levels like 500 and 700 mb. Do not look for mere temperature advection at upper levels; I won't get into why here.

With low-level forcing you're looking to see if areas of strong low-level convergence will be a factor, as mass continuity dictates that low-level convergence will produce rising motion. Obviously with a strong cold front rolling in from Tucumcari and Roswell you're going to get concentrated low-level lift and a "wall of forcing" that will bias you towards a much sooner estimate of initiation. This is more subtle with dryline situations, old outflow boundaries, and mesolows where you have slower ascent that slowly changes the environmental sounding from the bottom upward, and perhaps the best thing to do here is visually assess the convergence shown in the forecast low-level wind plots and bias the estimates based on how much or how little is shown. There are also smaller-scale thermally-driven circulations on a boundary that can produce initiation, being more pronounced with baroclinic boundaries showing a temperature gradient, in which case you might consider biasing things toward a sooner estimate. Here is where you can also take into account orographic effects, depending on terrain and the type of wind flow regime impinging on the terrain as the day goes on.

Then of course you can cross-check this against the estimates of model-derived reflectivity. There is no silver bullet technique here, as the model reflectivity is not infallible, but if you're looking at your own diagnosis and looking for consistency between that and the model predictions then you can rest assured you've done your homework here and will have a handle on things if an unexpected development occurs later in the day.

Tim
 
This is exactly what I was think about this morning. Everything said here already is great! Thanks.. I do have another set of questions to add to this thread. I have very limited knowledge on these things. THe reason I ask these questions is last yeah I found myself in 2 different situation about 60 to 120 miles east of where I should have been at Ignition, and it was frustrating.

1) How do you know exactly where the dryline is?

2) How do you know if the storms will form "On" the dryline or just to the east?

3) What's the best way to find Out Flow Boundaries? Visible satellite?

4) On a triple point setup, why choose one over the other. (Warm Front, DryLine)

Thanks in Advance!!!

Brian
 
Brian, on locating the dryline, I often start by looking at the Oklahoma mesonet dewpoint map. Many times, the dryline is very obvious with a dewpoint gradient of even 25 to 45 degrees within the space of a few miles. If the area of interest is north or south of Oklahoma, you'll have to take raw hourly surface observations from stations further apart, but with a little bit of interpolation you can usually identify the location of the dryline.

With regard to locating outflow boundaries, one of my favorite techniques is to look at radar in areas where a boundary is suspected (based on last night's convection, for example). This works best when the outflow boundary is fairly close to the radar site. The reflectivity of the boundary is often from flying insects caught up in the movement of the airmass rather than actual precip. Putting these radar images in motion, you can often get a good handle on the speed and direction the outflow boundary is moving.
 
Thanks Mike. Yeah, I have actually seen those OFB's on radar before. That's crazy! Insects? WOW! So I guess you just drive with the OFB hu?

So the Dryline, you look more at Dews then winds? That's been my problem I think. So mesonet, that's like a fine tuned analyses? Like short term, very specific, current info?

Thanks again!

Looks like I will be using some of my buddy's buddy passes to get out there Sunday Night. Flying into ICT! Ready for Monday and Hopefully Tuesday too!
 
Thanks Mike. Yeah, I have actually seen those OFB's on radar before. That's crazy! Insects? WOW! So I guess you just drive with the OFB hu?

So the Dryline, you look more at Dews then winds? That's been my problem I think. So mesonet, that's like a fine tuned analyses? Like short term, very specific, current info?

Thanks again!

Looks like I will be using some of my buddy's buddy passes to get out there Sunday Night. Flying into ICT! Ready for Monday and Hopefully Tuesday too!

Well, you can think of the outflow boundary as a mini-cold front. As with a synoptic cold front, the opposing air masses can create covergence. At the surface, when air masses converge, the air has nowhere to go but up. The rising motion is often a source of lift which can enhance initiation of convection.

Yes, the mesonet has a higher density of stations and takes more frequent observations (every 5 minutes in the Oklahoma Mesonet), as opposed to the hourly observations at regular NWS reporting stations. Given that the SPC Mesoanalysis usually lags about 40 minutes behind the last hourly observations, taking a peek at the mesonet can certainly give you an advance on current trends.
 
There are sooooooooo many different factors that go into this. I usually use color coded weather maps available on www.coolwx.com and I try to find sharp boundaries between temps or dewpoints or whatever. Usually if there's a moist flow from a body of water like the gulf of mexico moisture. A bunch of other factors go into this subject but that should start you off well especially if you live within 500 miles of the gulf or Atlantic.

Nick,
 
Thanks Everyone.. I have one other question about the mesonet thing. Do you guys know of any mesonet sites like the OKC one? I can't find one for Kansas, CO, TX etc?

I agree this is a great tool in finding the boundaries! Something I have NEVER been good at.

Thanks in advance!

Bp
 
Back to the initiation subject, isn't there some product that's available on some WFO's on the SPC site that is supposed to guide forecasting of initiation? I think it's called "Short Fuse."

I took a look at this last year; I was pretty lost on what some of the parameters meant. Nonetheless, if you study up on the topics this tool may be good at forecasting inish in the short-run, factoring in both time and location. For long term time/location, what people have said so far seems like a pretty good idea.

On the subject of OFB's, I've seen a few marked by slight wind shifts and slight changes in temp. Won't these guys usually travel in the direction of the surface wind, or are they high up enough to get pushed around more by 850's? I'm also having difficulty figuring out what size of storm/timing of storm/what kind of environment is conducive to supporting outflow boundaries significant enough to possibly affect initiation.
 
Last edited by a moderator:
Back
Top