Article: The hype around Models

[terstorm1012]

good find, blp. I really like that they do forecast verifications.

I like the model threads myself. They are good at honing my amateur forecast skills. I personally find it's better to look at the atmosphere first, then the models, then try to suss out a solution.

in my field (not weather) we use models and we always tell ourselves the old adage: "all models are wrong, but they are often useful." They're a pretty cool tool.

[blp]

Repost in case anyone missed the NHC's response below to the first article. I think the NHC did a good job explaining their side. Kudos to Dr. Franklin.

https://noaanhc.wordpress.com/2015/09/1 ... orm-erika/

[wxman57]

I don't pay any attention to them at all. I never open the Model threads here and automatically skip over any and all posts about them within other threads. They serve no purpose to me.

Models are an excellent and valuable tool. They should not be ignored when making a forecast. However, they must be interpreted correctly. I don't know how anyone could make a forecast if he ignored all model guidance. We'd be forecasting in the blind, using guesses and intuition. Not smart.

[WPBWeather]

I don't pay any attention to them at all. I never open the Model threads here and automatically skip over any and all posts about them within other threads. They serve no purpose to me.

Models are an excellent and valuable tool. They should not be ignored when making a forecast. However, they must be interpreted correctly. I don't know how anyone could make a forecast if he ignored all model guidance. We'd be forecasting in the blind, using guesses and intuition. Not smart.

I believe the point of the article is an over reliance on models and over confidence generally in forecasting...

[tolakram]

Excellent response from the NHC, as others have said.

[Alyono]

I don't pay any attention to them at all. I never open the Model threads here and automatically skip over any and all posts about them within other threads. They serve no purpose to me.

Models are an excellent and valuable tool. They should not be ignored when making a forecast. However, they must be interpreted correctly. I don't know how anyone could make a forecast if he ignored all model guidance. We'd be forecasting in the blind, using guesses and intuition. Not smart.

I believe the point of the article is an over reliance on models and over confidence generally in forecasting...

the point I got was the media is improperly relaying the model data, and not telling the whole story. They are doing this to the extent that it is clouding the message OF the actual forecast.

If they are going to show models, they need to show them in the proper context and stop using them to cry wolf

Also, we really need to get away from deterministic forecasts and go toward probabilistic forecasts, that properly convey the level of risk based upon atmospheric uncertainty

[WeatherGuesser]

And yet there is still heavy reference to them in most of the storm threads here, in context or not. It becomes annoying and tedious.

It may be time to close them to the public. Make them accessible to Mets only with usage terms and conditions barring most public discussion and disclosure. Use them for developing in-house forecasts only.

[Yellow Evan]

And yet there is still heavy reference to them in most of the storm threads here, in context or not. It becomes annoying and tedious.

It may be time to close them to the public. Make them accessible to Mets only with usage terms and conditions barring most public discussion and disclosure. Use them for developing in-house forecasts only.

I don't see any reason for them to be hiding from the public. That would kill off the private industry.

I also don't see what's so annoying about model references. They're gotten better, but it's a 2 step forward 1.5 step back kind of thing.

[blp]

And yet there is still heavy reference to them in most of the storm threads here, in context or not. It becomes annoying and tedious.

It may be time to close them to the public. Make them accessible to Mets only with usage terms and conditions barring most public discussion and disclosure. Use them for developing in-house forecasts only.

The average Joe is not getting their weather forecasts from Storm2k. Nothing wrong with talking about the models here. If you take the models out of here you might as well close this and other sites. I don't know what you are talking about this site is very well organized and has one thread for Global Models. The models are in seperate threads. If it bothers you that much then don't go to those threads

[WeatherGuesser]

I don't know what you are talking about this site is very well organized and has one thread for Global Models. The models are in seperate threads.

It would be one thing if that were true. But they're also discussed/referenced several times on most pages of most storm threads.

[Weatherwatcher98]

Where I see an issue is with the media cherry picking model runs and putting it on the news which can lead people to make assumptions. Oh the majority of the runs are off the coast so we should be Ok for example. It is Ok for people like us on weather boards to show the runs since we all know not to take them verbatum and we understand the good models from the bad but I just think there is no value in news outlets posting model runs. That is not something we saw years ago with the media. The media used to report everything from the Hurricane Center and now I see a little more forecasting versus reporting. Now I don't know how the NHC can control what the media reports that part seems like it would be difficult to do. Just my two cents.

Actually the euro was close on track. Intensity has always been the wild card even years before now. Computer models are made by people and subject to errors. Seems like a moot point

[Weatherwatcher98]

Where I see an issue is with the media cherry picking model runs and putting it on the news which can lead people to make assumptions. Oh the majority of the runs are off the coast so we should be Ok for example. It is Ok for people like us on weather boards to show the runs since we all know not to take them verbatum and we understand the good models from the bad but I just think there is no value in news outlets posting model runs. That is not something we saw years ago with the media. The media used to report everything from the Hurricane Center and now I see a little more forecasting versus reporting. Now I don't know how the NHC can control what the media reports that part seems like it would be difficult to do. Just my two cents.

Actually the euro was close on track. Intensity has always been the wild card even years before now. Computer models are made by people and subject to errors. Seems like a moot point

sorry wrong quote

[Weatherwatcher98]

Totally ridiculous article here.

I will completely disagree, and show a single graphic that should be enough evidence that something needs fixed.

Actually the euro was close on track. Intensity has always been the wild card even years before now. Computer models are made by people and subject to errors. Seems like a moot point

[Sanibel]

It's fairly easy to determine when the models are not detecting negative conditions and adjust.


The problem is probably more due to humans remembering favorable seasons where the models were automatic.

[Hammy]

Totally ridiculous article here.

I will completely disagree, and show a single graphic that should be enough evidence that something needs fixed.

Actually the euro was close on track. Intensity has always been the wild card even years before now. Computer models are made by people and subject to errors. Seems like a moot point

I believe someone else (not sure who) pointed out that those maps are automated and the program just calculated current position to forecast track.

I did a study of the Euro this year, and interestingly with already formed storms it's had a low bias with everything but Erika, and I think part of the problem is not that the conditions remained unfavorable, but that it never reached the favorable conditions and continued west in the Caribbean and then into land; I think the failure here was more that there was not a well defined circulation, and by inputting a center point into the models, they assumed there was.

[blp]

I believe someone else (not sure who) pointed out that those maps are automated and the program just calculated current position to forecast track.

I did a study of the Euro this year, and interestingly with already formed storms it's had a low bias with everything but Erika, and I think part of the problem is not that the conditions remained unfavorable, but that it never reached the favorable conditions and continued west in the Caribbean and then into land; I think the failure here was more that there was not a well defined circulation, and by inputting a center point into the models, they assumed there was.

You could be right. I had always heard though from pro mets that you did not need a defined center for a model to forecast. The models were capable of resolving the evolution of the center.

[LarryWx]

I did a study of the Euro this year, and interestingly with already formed storms it's had a low bias with everything but Erika, and I think part of the problem is not that the conditions remained unfavorable, but that it never reached the favorable conditions and continued west in the Caribbean and then into land; I think the failure here was more that there was not a well defined circulation, and by inputting a center point into the models, they assumed there was.

Hammy,
In addition to Erika, don't forget that the Euro had 94L in the W Gulf as a H (including major at least once if I recall correctly). As we know, that 94L never even made it to TC status.

Do you think that the Euro is still the King vs. the other models or at least least of the evils? I would have to say that I still have lots of interest in its solutions during critical times even to the point of keeping me up late sometimes. For that reason, I sometimes wish they'd get rid of DST. The heck with late evening sunlight if it makes it harder to stay up for the Euro lol.