My "Forecasting Technique"

[Sam Kennedy]

Here is what I do when I attempt a forecast:

1) Check some soundings for deep moisture along the gulf
2) Look at surface chart for convergence, boundaries: wind direction, dew points, temperature etc...
3) Look at the 500mb chart for negatively tilted low pressure troughs.
4) Look at the 300mb chart for divergence.
5) Check the soundings for CAPE,CIN,LI and shear
6) Compare my prediction to the models.
7) Check the visual satellite imagery for cumulus towers going up, because they show up on there before they do on radar

Is this good enough for a basic start? What can I do to improve it and make it more accurate?

 

[Chris Nuttall]

First, each person's technique for forecasting will vary not just on the person, but based on the situation. Everyone has their own methods and their own "favorite products" that they like to use. Since you're just starting, I highly recommend focusing on using the "Forecast Funnel."

As an example, I almost always start off every forecast by looking at a N. America WV loop. Next, I'll check out CONUS sat images and some constant height charts. At the end, I usually take what I've gleaned from models and my general predictions and look for the mesoscale boundaries to modify my forecast based on things the model can't resolve, like outflow boundaries.

As far as the basic components, you've got it pretty well-covered. One thing I like in particular in #6. There are way too many people that rely solely on models. Instead of forming a general idea first, they jump straight to their favorite model and treat it as absolute without doing in-depth comparisons on how they initialized compared to the current conditions at the time, and how they compare to each other.

 

[Greg McLaughlin]

3) Look at the 500mb chart for negatively tilted low pressure troughs.

The rest of your list looks fine. You will find many great chase days (even some of the best tornado days) occur when you aren't even dealing with a negative tilt trough. Obviously negative tilt troughs are often responsible for widespread, organized outbreaks, however just the same you can end up with non-tornado severe weather events under negative tilt systems.

I really like the forecast funnel Chris put up there. On the day of the chase, especially as you head into the afternoon, the forecast/nowcast almost exclusively becomes focused on mesoscale features. I believe one of the things that makes a great forecaster is the ability to determine when the models aren't handling a setup very accurrately and then being able to recognize a pattern based on real-time data.

 

[Paul Knightley]

The ingredients-based forecasting approach should always be used with severe storms. I.e. are the following 3 ingredients there? If so, then there's a chance of thunderstorms:

1) Instability
2) Moisture
3) Lift (trigger)

For severe storms, stick wind shear in too.

For tornadoes, looking for some kind of boundary is very important. When chasing on May 26th this year near Pratt, KS, we noted an outflow boundary moving from the SE towards our location. A supercell was developing to the WSW, and so we knew we had to be in position for when the storm intercepted the boundary - we were, and saw 2 tornadoes.

 

[Sam Kennedy]

I did include shear in my forecast, check number 5
I don't rely on the ingredients based technique, because when OK had that storm outbreak on the 5th, there was very little CAPE, strong cap, and I can't remember the LI, but there were still storms, so I want to rely on something more reliable

 

[Paul Knightley]

I would suggest that there would have been some CAPE otherwise storms wouldn't have formed! I remember looking at this evening and seeing several hundred J/Kg of CAPE, and LIs of about -3C.

 

[Sam Kennedy]

But a couple of weeks ago (When Obama Was Elected/Bonfire Night/Wednesday) The soundings looked rubbish, very low cape (<100) an a positive LI, but storms developed anyway, so I don't fully trust the soundings.

 

[Brandon Sullivan]

But a couple of weeks ago (When Obama Was Elected/Bonfire Night/Wednesday) The soundings looked rubbish, very low cape (<100) an a positive LI, but storms developed anyway, so I don't fully trust the soundings.

Im no expert, but it doesn't take a vast amount of CAPE to get storms to form. Especially in a cold core setup. I dont know what setup we are talking about. Soundings at least observed soundings are usually pretty accurate.

Download Bufkit. Its a nice program that sorts a lot of info out for you. Graphical viewing of the indices. Storm type maps. Just lots of features!

 

[Aaron Wesson]

The ingredients-based forecasting approach should always be used with severe storms. I.e. are the following 3 ingredients there? If so, then there's a chance of thunderstorms:

1) Instability
2) Moisture
3) Lift (trigger)

For severe storms, stick wind shear in too.

I use

M - Moisture
I - Instability
L - Lift
E - Exhaust

 

[Sam Kennedy]

I understand MIL, but what is exhaust?

 

[Skip Talbot]

I always look for moisture, instability, lift, and shear. I never heard exhaust before and had to look it up quickly. Apparently it does mean shear. I'm assuming exhaust means separation of downdraft and updraft through ample wind shear so the storm doesn't crap on itself?

 

[Sam Kennedy]

That is what I thought, but shear has already mentioned so I thought it was something different. I will probably get better at this after I practise with Forecast Laboratory and see what actually happens during the spring and summer.

 

[Gabe Garfield]

When it comes to forecasting where tornadoes are going to occur, I look at a few things:

• Is the deep-layer shear sufficient for supercells? Without a supercell, you're not going to see a tornado (unless it's a landspout, of course). I like to see at least 35 kts of deep-layer shear in high CAPE situations and probably 45 kts with low-moderate CAPE situations.
• What will be the preferred storm mode? All the CAPE and shear in the world won't help you if you get a squall line. Will discrete cells occur anywhere? If the deep-layer shear vector is parallel to the initiating boundary, you're not likely to see tornadoes (unless storms initiate well ahead of the boundary, in which case you have a very weak cap or strong large-scale ascent).
• Where will low-level shear be maximized? Deep-layer shear may be sufficient for supercells, but if there isn't any low-level shear, cold pool propagation will be significant (which could ruin all your fun). Outflow boundaries are often great places to find enhanced low-level shear in the absence of a strong, large-scale, low-level jet.
• Finally, is the cap breakable? As with storm mode, the greatest juxtaposition of parameters in the world will do you no good if you don't have a storm. A quick look at 700 mb temps can give you a good clue where storms are likely to initiate.

 

[Sam Kennedy]

Got a few questions:
1) What is deep-layer shear?
2) What is low-level shear?
3) Isn't a Juxtaposition from Back to the Future?
4) Why do you use 700mb to see if the cap will break, and how can you tell if it will?

 

[rdale]

1) Through a good part of the atmosphere
2) Near the low levels vs aloft
3) Intersection is another word to use
4) Mid-level temps can stop convection from forming

Many of these questions are answered in the METED links I've posted here, have you checked those out?