Interpreting the GFS and ECMFW models

[Mark DeRidder]

This will be my fourth year storm chasing and I have had a lot of luck so far. However, I have to be honest about my lack of forecasting knowledge. I read a lot about interpretting the GFS and ECMFW models to help pinpoint good chase days, but I am unable to to understand what I am looking at.

Could some one post a link to the best resource for interpreting these maps? Perhaps a quick summary on what a chaser should be looking for. Direct links to the best models would be helpful too.

I apologize if I am oversimplifying things. Or if I am missing an obvious link on the forum about this. I'm sure it takes a lot of time to understand these models correctly. Thanks!

 

[Michael Riedl]

This is a good site for beginners. Under the sections Models and charts you can get a pretty good understanding of what each chart is showing. It has some great information for severe weather as well.

http://www.theweatherprediction.com/

This link is another useful one as well.

http://weather.unisys.com/model/details.html

When I look at models I mainly use the website TwisterData.com . It has a pretty easy to use interface.

I am no expert either as I am still learning as well. I hope this is what you are looking for!

 

[rdale]

Read Tim's intro book first. Going right to a model website to view the models won't help you learn. Another big resource is the model page at http://meted.ucar.edu

 

[Tim Vasquez]

Well, actually what you're doing is not really interpreting the ECMWF and GFS but interpreting the charts and seeing what kind of solution they propose. Here is an example of a 300 mb chart.

In this case there is a massive upper-level high over Texas and New Mexico, which is something we normally see in the middle of summer. This suggests warm temperatures aloft, weak instability (i.e. "warm over cold"), and weak winds (strong winds aloft are important for supercells and many types of severe weather modes). However up north in the Dakotas and Minnesota you see that important wind in the form of a jet stream.

Also the curvature of the upper level patterns gives some idea of the fronts, lows, and highs at the surface. Here we see a ridge over the northern Rockies and a trough over the Great Lakes. Typically a surface anticyclone is found under the jet where a ridge is upstream and a trough is downstream, and in this example that surface anticyclone would be somewhere in the Dakotas. This anticyclone is most likely comprised of a relatively dry, cool polar air mass.

The flip side of this is that this implies a front is somewhere to the south... Nebraska? Kansas? Oklahoma? Texas? If that front is too far south under that weak flow and under that high, then it might be too capped for frontal storms or the storms might be disorganized. If the front is further north near that upper jet then the storms might be strong to severe... in this pattern that would give the "northwest flow" type events we hear about from time to time.

So those are some things you would be thinking about looking just at this chart. Being a 360-hour forecast, though, that's FAR into the future and chances are that none of these details will pan out. But it does alert you to the fact that the jet stream might be much further north than usual, and you'd want to lay money on chasing in the northern Plains rather than the southern Plains.

Tim

 

[Bryan Stokes]

Hi Mark,

I am a complete newbie when it comes to forecasting, but I'm working hard on learning how to forecast. Here's the route I'm taking -- perhaps it might be worth looking into.

It would be difficult for anyone to give a quick summary of what to look for...mainly because forecasting is such a complex equation. There are just too many ingredients and variables. What I can suggest is picking up the following two books from Weather Graphics:

Severe Storm Forecasting
Weather Map Handbook

The first book will help you understand how and why severe weather forms. The second book will help you better understand some of the models and maps out there and how to use them in your forecasting, including the various model's strengths and weaknesses. As you begin to learn the "how" and "why" of weather, you'll start to develop a better understanding of what to look for in the GFS/ECMWF and how to interpret what you're seeing.

Both books are very in-depth, yet written where the everyday person can understand. They will be something you reference for years to come.

I hope this helps,
Bryan

 

[Mark DeRidder]

Thanks Michael for those links! I haven't seen those yet. Twisterdata.com is particularly interesting, and I'll be stopping by there often I think. Theweatherprediction.com site is a bit messy to look at, but I'll focus on the charts and models sections.

I do have Tim's book, which is fantastic. I'll comb through the forecast section (which I skimmed through originally ) to see if I can figure this stuff out.

And Tim, thanks a ton for giving me a nice summary of a current model! Your read of the map makes a lot of sense to me and I'll be looking for that 300mb model often. I think the different heights confuse me as well as upper and lower pressure systems. I wish it could be one map that would tell you everything you need to know, but its obvious I need to look at the different heights.

 

[Mikey Gribble]

I didn't read through everybodys post so if this has already been said just ignore it. I saw the part where they posted links to good site for model data. College of DuPage is a good one too. Here is a link to their model page http://weather.cod.edu/forecast/

I use COD quite a bit for forecasting. There are several other websites I use too, but I like the way the graphics are layed out on COD's charts the best. You can get a higher resolution image there than most people have.

You are basically asking how to forecast, which is a major undertaking. I worked extremely hard at it for my first few years of chasing and then have steadily tried to maintain and improve my forecasting ability since then, but I still only know enough to get by. Forecasting tornadoes is extremely complicated. You HAVE to learn from books up front. Everything is built on that fundamental knowledge. You have to have a strong foundation (like intro to meteorology textbooks and Tim's Forecasting Handbook give you) before you can start moving ahead into more difficult aspects of meteorology. I think thats where a lot of people go wrong. They try to move ahead too quickly. IMO its important to understand the physics of the atmosphere well enough to envision the fluid motions and how it works in your head. That probably doesn't make much sense, but I can't think of how else to put it. Basically TOTALLY understanding the fundamentals is the most important thing.

Learning the first half of forecasting is done in books. After that I think experience is the other half. Watching various setups come together and unfold is indespensable. You start to build an archive of past examples in your mind and you have an idea of what to expect when future setups come up that are similiar.
The other half of experience comes from following the models closely. A major part of forecasting any more is recognizing the trends from run to run and model to model. If you can recognize that, more times than not you'll have a good handle on the setup. Its more difficult than it sounds though.

I guess what I'm getting at is that if you want to learn how to more effectively interpret any models you have to put a LOT of time into it. Start off reading books and other peoples forecasts. Be extremely careful about whose forecasts you read though. Forecasts are kind of like assholes. Everybody has them and most of them stink lol. Make sure it is coming from a reputable person that you know is actually spending time watching models and not just pulling something out of their ass. I've seen lots of posts on here that state one thing or another about something a model is showing and its totally wrong. Read SPC and the NWS discussions. Those are two reliable sources. Then find a couple good chasers that post forecasts. Familiarizing yourself with those kinds of forecasts while at the same time watching the models yourself can teach you a lot.

 

[Mark DeRidder]

Wow, thanks Mikey! I really really like that COD website! The loopable animations really show you that fluid dynamics. Very interesting stuff.

I appreciate the fact that true forecast knowledge takes years if not a lifetime to fully understand. And I think it is what separates the men from the boys when it comes to storm chasing. Its that next step that you need to ask yourself if you are ready to fully jump into, because time is valuable

I guess I am ready to jump in and I'm happy to see some really nice resources suggested here. Thanks to all for your time and help! This is sending me in the right direction.

 

[Greg Blumberg]

Start looking at the 500 mb map every single day. Part of learning how to forecast and see things that the models are producing requires that you see how the atmosphere is working right now. Looking at the 500 mb map can help you on your way to acquire good pattern recognition skills. Pattern recognition is something I stress a lot when teaching beginning forecasters.

It's good to learn conceptual models too so you can compare them to what you see when reading weather maps. The website below should give you a great starting point for reading the 300 mb and 500 mb maps. You can learn A LOT about the atmosphere from looking at those maps. Judging by your website, I'm guessing you're a very visual person, so this website really ought to help you out.

http://www.aos.wisc.edu/~aalopez/aos101/wk14.html

The first pattern/conceptual model I teach for forecasting is the trough/ridge pattern in the upper atmosphere. It's very simple and basic and helps you get a good idea as to what the atmosphere is doing. Compare what you're seeing in those 500 mb and 300 mb maps to satellite images, radar data, and surface data (where the surface Highs and Lows are.) It's a good start.

Here's an example of a 500 mb chart you can practice finding troughs and ridges with:

http://www.weatherphotography.net/images/dwm500_test_20041220.gif

Something very important is you have to stick with a forecast you make. It's when you get burned is when you learn the most. Keep notes, keep track of what's happening, never stop learning. The more experience you get observing the weather, the more things you'll see when forecasting.

 

[Mark DeRidder]

Thanks Greg, those visuals do help me! It is nice to see things in 3 dimensions and preferably in motion. I hope one day weather forecasting models will be displayed as 3D real-time imagery so one could spin around and look at it from all angles. If this already exists please let me know!

Btw, I like your storm chasing tips! I play lots of John Williams music on my chases

 

[rdale]

If this already exists please let me know!

It does - and it's free - but there are still plenty of issues to work out. Get IDV from Unidata...

 

[Michael Wurzburger]

This is a good site for beginners. Under the sections Models and charts you can get a pretty good understanding of what each chart is showing. It has some great information for severe weather as well.

http://www.theweatherprediction.com/

This link is another useful one as well.

http://weather.unisys.com/model/details.html

When I look at models I mainly use the website TwisterData.com . It has a pretty easy to use interface.

I am no expert either as I am still learning as well. I hope this is what you are looking for!

Pretty awesome links Michael!