• 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].

Better forecasting

Still, I am a complete beginner and even this seems too complicated. I will check out the MetEd-pages thoroughly hoping this might be the key to start my learning. I am starting to recognize and slightly understand definitions like CAPE, trough, lift etc it is very difficult to look at a weather map and start "seeing things". Right now it merely feels like I am picking up puzzle pieces here and there in no organized learning.

It's a gradual process and every little bit helps. I initially forecasted just based off of pattern recognition and ingredient overlap. Not the best way considering a diagnostic approach is much more helpful and accurate, but it's a decent way to get a feel for everything you're seeing.

Personally, the easiest way for me to forecast on a chase day is to go down in scale (synoptic features down to mesoscale features).
1.)I start out with the 500 mb upper air map from 12Z
2.) Print off the HPC sfc analysis map
3.) Check current observations across my target area
4.) Check closest sounding site for 12Z sounding unless site is north of front or behind dryline
5.) Check radar
6.) Check satellite imagery
7.) Hand-draw a sfc chart to pick up on smaller features like moisture convergence on a boundary, mesolows, outflow boundaries, cold pools, etc.( I'll usually have satellite imagery up as well to pick up smaller features that may slip through the cracks of the surface observation network)
8.) Keep all these out and re-analyze when I reach my target area based off of current data
 
For a thread encapsulated within the Introductory Weather & Chasing area such as this, I actually think that at some point somebody should lock the thread and say "Enough!" The point being, obviously, that for relative beginners it is easy to get overloaded with suggestions. Other than that, I do want to say that one of the most important lessons I have learned in my decade-and-a-half of chasing is that there is a lot more that can go wrong with a setup than the near-perfection that it takes to have a photogenic tornado. Therefore, my preparation mostly involves process-of-elimination; i.e. looking for which of the dozens of potential setup problems are going to exist. Honing that process saves a lot of gas, LOL, but yeah, sometimes I miss a good tornado due to being too pessimistic, but that's okay with me.
 
Karen & Mike: Thanks for the advice! I am really glad to get help here. I have seen the SPC Mesoanalysis map referenced plenty but haven't really found where it was so this was quite interesting.

Still, I am a complete beginner and even this seems too complicated. I will check out the MetEd-pages thoroughly hoping this might be the key to start my learning. I am starting to recognize and slightly understand definitions like CAPE, trough, lift etc it is very difficult to look at a weather map and start "seeing things". Right now it merely feels like I am picking up puzzle pieces here and there in no organized learning.

I totally agree that understanding comes way ahead of forecasting. I will go through all the educational links that has been provided in this thread and see what I can come up with!

How about my questions about alerts, which one do you prefer?

My suggestion is www.theweatherprediction.com. He has a link specific to severe weather forecasting. Bob mentioned all the things that can go wrong. To each his own. I'm limited on the days I can chase at the moment, so if there's a decent chance for something when I can chase, I'm gonna chase it. In all but a couple cases I've at least seen storms, so hey, that's better than nothing. The days I saw tornadoes were pretty obvious outbreak days, with the exception of one that was 30 minutes from home.
 
One thing that would really help me out in learning is to have a step-by-step tutorial on which resources (including links) you use on any given day, like (and this is my guess which is very likely to be wrong, but just to give you an example)

I'm actually not sure this would be helpful, as it can be overwhelming, and not everything will mean something to you at first. Figure out a routine that works for you, and expand as you find new resources and learn more!

For example, everyone should pretty much start with the 4 ingredients of severe storms as a basis. Let's start with:

Is there ample moisture?

Well there isn't a predefined amount. Many people go by 60deg dewpoints at the surface as an indicator. But then you notice with experience that storms can drop tornadoes in Colorado or Texas with 50deg dewpoints? How does that work? Well, orographic features to be blunt, but it's yet another thing you'll need to investigate.

Then you notice that forecast models are often wrong on dewpoints. You planned a 2 day chasecation when you saw 65deg dewpoints forecast, but in reality they ended up being 55deg and the storms lacked oomph. So now the question is why did the models over-forecast moisture, and can they under-forecast?

Then you notice 65deg dewpoints but storms still had problems, and you notice there was no moisture at 850mb. So now the question is how deep does moisture need to be and how does 850mb cyclogenesis work?

You start asking enough of these questions and experiencing enough and you build up a framework of what you check for a forecast. Friend people on Facebook and learn about new model sites when they link to them. Ask people here how to find certain information. Google search, and archive search here. Checking something is only useful if you find it useful, understand it, and incorporate it into your decision making process.

Go through this process for the other ingredients. Re-evaluate. Iterate. Improve. Adjust.

For an example, here is what I do on chase mornings:
*I've already been forecasting at least a day in advance for everything but the most surprise local setups, so I have a general idea of how well the four ingredients are being met, and how things will likely play out. This is based on 5 years of doing this for almost every decent setup.
1) I quick check the SPC outlook image (only) to make sure nothing has drastically changed. If I expected the risk to be in OK, and all of a sudden the new outlook is a slight risk in NE, I need to wake everyone up and we need to get into panic mode and figure out what changed or if the SPC is on drugs.
2) Then I check the HRRR to see when it breaks out precipitation and what type of precipitation. I treat this very generally. I worry if I see no precip breaking out, but otherwise I give supercells a 2-3 hour window around the time the HRRR shows it. Either way, I make note of it for addressing the cap later.
3) I check current observations and satellite to identify surface synoptic features like fronts and lows
4) I go back to the HRRR to see how these synoptic features are likely to move throughout the day
5) I check the HRRR to assess all the ingredients - usually MLCAPE for instability, surface dews for moisture, and SRH/500mb winds/850mb winds/sfc winds for shear. Lift is kind of something that is hard to assess and there's a lot of implied and derived outcomes based on what I see throughout this whole process.
6) I check vis sat/wv loops to identify short waves, outflow boundaries, and the like
7) Now I check SPC text to verify whether I missed anything or might want to reconsider something. I don't want their thoughts tainting my forecast or adding bias before I've even done it.

*get into position*

8) Sit and watch vis sat for signs of the atmosphere being ready to produce
 
Last edited by a moderator:
Unfortunately there's no quick and easy step-by-step process to chase forecasting. There are lots of good resources in this post to get started. Skill, however, will only come with experience. 'Armchair chasing' during any severe weather event is a great way to practice. The 'chase cases' on this forum are great too. Forecast discussions and MDs by the NWS and SPC are helpful to a point, but most days they won't help you pinpoint the best chase target in time.

There is too much on each of these to go into detail, but for starters, here is a list of model charts to look at leading up to a chase day:

- 500mb winds (severe weather is more likely with 500mb flow above 25-30 knots).
- 850mb winds and surface winds (look for clockwise turning with height compared to 500mb). IE, southeasterly and surface, southerly at 850, westerly/southwesterly at 500mb
- Surface dewpoints (the higher the better)
- CAPE: surface-based or low-level cape is needed for tornadoes
- 700mb temps: provides a good picture of the strength of the cap
- 3 to 6 hour total precip and simulated radar: shows whether the model thinks storms will develop and where

On the day of the event:

- Visible satellite: the most important day-of data source. Use these to find boundaries and watch the progress of cumulus fields. Loops (animations) are the most helpful.
- Surface (metar) data: Can be used to locate boundaries (dryline, fronts, outflow boundaries), backing winds and better moisture.
- SPC mesoanalysis page: Good for watching trends of many different parameters.

Again, you could write a book on each of those points, and there are many more - but those are at least starters. Spend some time Googling and reading on each of those bullet points and practice even on days you don't chase.
 
Thank you so much, everyone! I have tons of things to look into now!

Since I am based in Sweden armchair chasing is my best option. Today seems like a good day to try things out due to the slight risk in Eastern Texas / Louisiana. Is anyone gonna chase this? Preferrably with a web cam? It would be really cool to "shadow chase" with you.

My last (yeah, right :)) question is: Is it possible to follow radar close up over the web. Such as it can be seen in the GR Level-software?
 
My one suggestion to new forecasters is to read the discussions from the SPC(Storm Prediction Center), NHC(National Hurricane Center), and the various local National Weather Service offices. These are full of information about how veteran forecasters make their decisions in real-time, ongoing events. You cannot get a better teaching resource. And its free.
 
My last (yeah, right :)) question is: Is it possible to follow radar close up over the web. Such as it can be seen in the GR Level-software?

SimuAWIPS is probably the closest you'll get. A real desktop app can handle that task better than anything currently available that is web-based.
 
4. This could be a chase day! But let me just check out the upper level winds and moisture, I do this here: http://www.whichsitethiswouldbe . com
5. Great, now I need to know the CAPE I want to see if it is above 3000 (?).
...and so on.

You can check parameters such as upper level winds, CAPE and much more at various sites containing numerical model data. There are many resources for data, I have literally hundreds of sites bookmarked but for the broadest information I use sites such as these that include output from multiple models:

http://weather.cod.edu/forecast/

http://www.twisterdata.com/index.php?prog=home&page=about

http://weather.rap.ucar.edu/model/

http://www.weather.unisys.com/index.php

Do your research on the various models and you’ll learn which are more useful or relevant based on the situation. The above barely scratch the surface and you’ll likely use a multitude of resources for additional model data, surface analysis, precipitation forecast, radar, satellite, soundings and more.

My last (yeah, right :)) question is: Is it possible to follow radar close up over the web. Such as it can be seen in the GR Level-software?

COD is an excellent source for NEXRAD radar that you may find useful:

http://weather.cod.edu/satrad/index.php
 
Thanks everyone for all the additional info. I was trying to shadow chase the storms of TX yesterday but struggled with not having a good radar tool, this (http://weather.cod.edu/satrad/index.php) will come in handy. I have a Mac so I can't use GR LevelX and RadarScope is not possible to purchase in AppleStore outside of the US for some peculiar reason.

I have tons of material to look through now, it will keep me busy for a while!
 
I have a Mac so I can't use GR LevelX

I used GR on my Mac all the time. Use Bootcamp or whatever its called now (free), or buy Parallels/Fusion/etc. I think the NOAA WCT will run on a Mac as well, since it's Java. Basically, there isn't an easy out-of-the-box solution, but if you really want radar you'll find a way :)
 
You're absolutely right, I will go through an friend with an american credit card and purchase RadarScope that way. Since I will use it for educational purposes only I want to be able to switch between my other programs (on the Mac-end) as well.

After tips from Peter Wharton I found this: http://training.spotternetwork.org/ , which was a big chunk of what I was looking for. I hadn't really figured the basics out and this was perfect! Even aced the test.

Now the next step is to figure out weather analysis on a synoptic scale and try to find the sources for that.
 
My one suggestion to new forecasters is to read the discussions from the SPC(Storm Prediction Center), NHC(National Hurricane Center), and the various local National Weather Service offices. These are full of information about how veteran forecasters make their decisions in real-time, ongoing events. You cannot get a better teaching resource. And its free.

This was a big part of my learning. When I read something that I didn't know what it was, I then googled it and learned about it.
 
Now the next step is to figure out weather analysis on a synoptic scale and try to find the sources for that.

Many chasers including myself have found this to be an excellent educational resource:

http://theweatherprediction.com/

It's a very thorough introductory level course-type site that should help you in learning how to analyze the weather and will likely prove beneficial in providing a foundation for advancing your knowledge.
 
There is too much on each of these to go into detail, but for starters, here is a list of model charts to look at leading up to a chase day:

- 500mb winds (severe weather is more likely with 500mb flow above 25-30 knots).
- 850mb winds and surface winds (look for clockwise turning with height compared to 500mb). IE, southeasterly and surface, southerly at 850, westerly/southwesterly at 500mb
- Surface dewpoints (the higher the better)
- CAPE: surface-based or low-level cape is needed for tornadoes
- 700mb temps: provides a good picture of the strength of the cap
- 3 to 6 hour total precip and simulated radar: shows whether the model thinks storms will develop and where

On the day of the event:

- Visible satellite: the most important day-of data source. Use these to find boundaries and watch the progress of cumulus fields. Loops (animations) are the most helpful.
- Surface (metar) data: Can be used to locate boundaries (dryline, fronts, outflow boundaries), backing winds and better moisture.
- SPC mesoanalysis page: Good for watching trends of many different parameters.


I get back to this thread a bit now and then while trying to learn about forecasting. I think this answer was the closest to what I was looking for. This really gives me something hands on what to look for and why!

Many of you provided good answers but I think many of you overestimate a bit what a beginner knows about all the details and definitions of forecasting and weather. I mean, I have learned a lot lately but there are still so many things I don't have a clue about. I really believe the SPC text is a great source of knowledge but it is not Forecasting 101. Let me give you an example of how it is reading this for me (at my current skill level).

This is Day3@March24th and in citation marks are my thoughts while reading (the questions are not really questions to this forum but rather questions in my head):

THE MID/UPPER FLOW APPEARS LIKELY TO CONTINUE TO TREND MORE ZONAL
AND PROGRESSIVE ACROSS THE U.S.


"'Mid upper flow' - that's like 500 mb winds, right? 'More zonal and progressive', I have no idea what that means. Whatever, let's go on..."

DURING THIS FORECAST PERIOD. WHILE
A DEEP LOWER/MID TROPOSPHERIC CYCLONE GRADUALLY MIGRATES
NORTHEASTWARD AWAY FROM THE NEW ENGLAND COAST..


"'Tropospheric cyclone' - Hmmm, doesn't mean anything to me. Is that like a low pressure? Well, I guess I should look for something that looks round and cyclonic off the New England coast and I may find out what they're talking about. What map could I see that in? Let me go through my bookmarks"


.A SHORT WAVE
IMPULSE...


"'Short wave impulse' - I recognize this. It feels important in a way, I have no idea what this really means in terms of weather maps but I'm guessing a short low pressure that quickly comes and goes"

EMBEDDED WITHIN THE MAIN BELT OF WESTERLIES EMERGING FROM
THE MID LATITUDE PACIFIC...IS FORECAST RAPIDLY MOVE INLAND ACROSS
THE PACIFIC COAST...TOWARD THE CENTRAL PLAINS.


"I think I get this. There are westerly winds coming of the central (or what does 'mid latitude' really mean?) Pacific. 'Rapidly move inland' - I like the sound of that! I guess this means moist and warm air from the Pacific is moving in towards the central plains. This I may be able to find in the Mesoscale analysis map? I wonder which level of mb I should be looking for these winds in?".

CONSIDERABLE SPREAD EXISTS AMONG THE VARIOUS MODELS CONCERNING THE
MANNER/SPEED WITH WHICH THIS OCCURS...


"Ok, so the data is not clear about this. Alright, go on..."

AND ITS IMPACT ON SURFACE
CYCLOGENESIS WITHIN/ALONG A LEE SURFACE TROUGH/DEVELOPING COLD FRONT
ACROSS THE CENTRAL PLAINS INTO UPPER MISSISSIPPI VALLEY REGION.


"'Surface cyclogenesis' - A low pressure being built up at the surface? Even if that is what that means, I have no idea of what it means". 'Lee surface trough' - that's an isolated low pressure, right? Or, what did trough mean again?! Darn, I remember learning this!"

HOWEVER...GUIDANCE DOES APPEAR IN GENERAL AGREEMENT THAT SURFACE
RIDGING ALONG THE WESTERN AND CENTRAL GULF COAST WILL BE SLOW TO
RECEDE AND WEAKEN.


"'Guidance does appear in general agreement' - Does that kinda mean 'We don't know but it seems like...' in SPC-speak? 'Surface ridging' - don't really understand what this relates to..."

AND INLAND MOISTURE RETURN WITHIN A DEVELOPING
SOUTHERLY RETURN FLOW OFF THE GULF OF MEXICO...AHEAD OF THE LEE
SURFACE TROUGH /VIA THE LOWER RIO GRANDE VALLEY/...AS WELL AS ABOVE
THE MODIFYING BOUNDARY LAYER ALONG THE GULF COAST...STILL APPEARS
LIKELY TO BE SOMEWHAT MODEST.


"'Inland moisture return' - Now, where could I see this best again? The mesoscale analysis map? I should look for winds coming off the Mexican Gulf, rigjt? In genereal I actually think I understood this part quite well despite what 'weight' they put in 'somewhat modest', I assume 'somewhat modest' doesn't mean it's very much and that is something negative".

----

I have read about 20-30 SPC texts and usually I understand a bit more than this example (especially on archive moderate/high risk days) but this is still an example of how I perceive these texts. I will follow the advice many of you have given and read through WeatherPredictions.com completely before I give this another shot. I will continue reading the SPC texts every day because I still pick up a thing or two. My main problem is typically to know what map to find the things, they're writing about, in...and also to understand what to look for.
 
Back
Top