Models

[steelhead]

Hi John,

When formulating your forecasts how many computer models do you typically use?
Are these models updated on a regular schedule during the day, say every 6 hours or do they typically update when changes occur?
I have watched some hurricane coverage where they show the "cone" and projections of where a particular hurricane may come ashore. Sometimes they will show the different models and there is usually one outlier. This may be similar to your "wild card".
If it wouldn't require a lot of work it would be fun to see the different models the next time a snow event is projected for the northwoods.

 

[jd]

It will vary, depending on what is lying in wait for us. Right now, the next 4 days looks pretty quiet, so I will stick to around 3 different models for guidance. When a big storm is in the offing within the next 2-3 days, then I may use as many as 5-6 models for guidance. When Lake Snows are the main issue, 1 or 2. There are only a handful of models that go out beyond 3 days, so for forecast features out beyond 3 days, I am more limited in my resources.

As far as looking at the different models, most of the output is not in a form where you can just look at a map and be looking at snowfall potential. Some maps show precip, but then you have to determine by other maps what form it will fall in, so it would be kind of difficult for me to display or link to maps that show the different models ideas on snowfall with a pending storm.

-John

 

[steelhead]

Thank you.

 

[eagle1]

John-
Question about these "models". What actually is a model? Just a program? And do all forecasters use or have access to the same ones?

Thanks, Guy

 

[mattus]

John-

Can anybody access these models or are they just in the meteorologist community?

Thanks

 

[jd]

The models are computer programs that use mathematical equations to try and simulate or "model" the atmosphere and predict what it will look like in the future based on how it is right now.

The models are not exclusive to meteorologists, although most of the ones I look at would probably not be of much use to anyone that does not have some formal training in atmospheric science (and maybe some who do!).

Here is one of the sites I use from time to time:
http://mag.ncep.noaa.gov/NCOMAGWEB/appcontroller

-John

 

[lenny]

gee's, I thought we were talking about a 70 Dodge charger scale model

 

[josh_4184]

Regarding the model guidance, John is correct, there are many different models out there including the North American Model (GFS), European model (ECMWF), which are the most popular especially in the long-range which in many cases still ends up being useless. You also then have the NAM which is a higher resolution model compared to the GFS which "generally" is better in the shorter ranges but not always the case. One other popular model which is used for short term usually less than 12hours is the RUC or rapid update cycle model which basically is a "Now cast" model used primarily during the onslaught of a storm all the way through the completion. Finally you have other models including the Canadian model (GEM), and United Kingdom (UKMET), plus not even counting the ensemble models which are separate from the operational models. Probably getting everyone confused even more now.

Keep in mind that these computer models are just one tool that meteorologists use when making forecasts, it’s up to them to extrapolate the data with other tools/techniques/experience to make the best forecast. I look at the models as a hobby, I am nowhere near educated enough to make informed decisions based of model guidance which as many meteorologists would say need to be seriously updated anyway. I leave the forecasting and opinions to the experts like John, but there is no reason why people like me can't enjoy it as a hobby, but in reality a lot of it goes over my head because of the extreme complexity in atmospheric modeling.


God I must of been pretty bored to right all that up lol.