What is the best place for weather predictions, 10 days to several months?

[Buppies]

Please note these guys are two week one month out predictors not tomorrow's weather and the ones I know concentrate on the eastern part of our country

 

[s219]

There is a forecaster in Richmond who is pretty good long range man but a lot of his info is on his paid site. The facebook side is still free. I have listed him on the good morning thread several times there are others as well but will requie you to read and have an understanding of the terms they use. This winter so far has confused a lot of them because the signals showed one thing and then changed about three weeks ago. If you are interested you can PM me and I'll share what I have/know. I have followed them for 13 years and while wrong sometimes at least you know in advance what's possibly coming.

If it's the WxRisk guy, he is excellent (though a bit arrogant and lacking manners sometimes). I find reading his FaceBook site a bit tedious due to all the morons following him and asking stupid questions (actually, I find all of FaceBook tedious for about the same reason). He will usually give a brief score of how his predictions panned out, but nothing in detail on the Facebook page.

I have been writing weather software since about 2006, and rarely see retrospective data analysis or validation of models against results, at least not in public. There is some of that done at the research level, in a way that is not really useful to people on the street (i.e., they will focus on vorticity or some other high-level aspect of the weather).

To be honest, I think anything more than about 15-20 days out is throwing darts. Weather is too chaotic, too sensitive to changing conditions. The equations and models are highly sensitive to initial conditions, and when you start a model run on day 1 it can be wildly out of the scope of reality by day 7, never mind day 15 or 20. I follow some of the weather model discussions on the americanwx forum, and you see that all the time. Weather predictions have to continually restart models based on the latest weather observations to get the near range 1-3 day and 3-7 day forecasts, since previous longer-range runs have diverged from reality.

 

[GPintheMitten]

I just checked accuweather for my area. It says Feb 8 is going to be 24 but Feb 9 is going to be warmer 32.

I wonder who is looking at the magic 8 ball to come up with this stuff. It's ridiculous to say anything other than what's normal over 30 days in advance.

 

[oosik]

I can understand wanting the most accurate wx forecasts. But, why not just do something to protect the exposed pipes also.

 

[Buppies]

If it's the WxRisk guy, he is excellent (though a bit arrogant and lacking manners sometimes). I find reading his FaceBook site a bit tedious due to all the morons following him and asking stupid questions (actually, I find all of FaceBook tedious for about the same reason). He will usually give a brief score of how his predictions panned out, but nothing in detail on the Facebook page.

I have been writing weather software since about 2006, and rarely see retrospective data analysis or validation of models against results, at least not in public. There is some of that done at the research level, in a way that is not really useful to people on the street (i.e., they will focus on vorticity or some other high-level aspect of the weather).

To be honest, I think anything more than about 15-20 days out is throwing darts. Weather is too chaotic, too sensitive to changing conditions. The equations and models are highly sensitive to initial conditions, and when you start a model run on day 1 it can be wildly out of the scope of reality by day 7, never mind day 15 or 20. I follow some of the weather model discussions on the americanwx forum, and you see that all the time. Weather predictions have to continually restart models based on the latest weather observations to get the near range 1-3 day and 3-7 day forecasts, since previous longer-range runs have diverged from reality.

I would agree with your assement. Personally I'm looking for trends as my work is completely weather dependent, too cold or too wet no work. This year the models are in chaos and not reading things well at all. So I try to look at MJO phases of the EPO etc to come up with my own thoughts. DT is a bit rough but a good moderate to long range man. Have you followed WXSouth?

 

[Buppies]

I just checked accuweather for my area. It says Feb 8 is going to be 24 but Feb 9 is going to be warmer 32.

I wonder who is looking at the magic 8 ball to come up with this stuff. It's ridiculous to say anything other than what's normal over 30 days in advance.

In my opinion I think they are worthless but again just my opinion. Trends can be somewhat predicted actual temperatures no way

 

[s219]

I would agree with your assement. Personally I'm looking for trends as my work is completely weather dependent, too cold or too wet no work. This year the models are in chaos and not reading things well at all. So I try to look at MJO phases of the EPO etc to come up with my own thoughts. DT is a bit rough but a good moderate to long range man. Have you followed WXSouth?

I have not followed WxSouth, but will check it out!

 

[blin]

All my local forecasters called for 100% rain today for several hours. I put off a major project based on their forcast. We only got a couple of drops.
Sometimes all you can do is flip a coin. Even then you still have a 50% chance of being wrong.

 

[newbury]

I can understand wanting the most accurate wx forecasts. But, why not just do something to protect the exposed pipes also.

The pipes are not all exposed, they run in the walls. The normal min reported temp is 32 degrees F, last year it hit 8 degrees F. The heat source for the 40 x 60 workshop was one overhead propane heater, until someone stole my 250 gallon propane tank.
So if I can plan on being down there I just put small electric heat sources near the pipes and walls and warm it up a few degrees. If I'm not down there I turn off the water at the street. I prefer to try and plan on being down there.

 

[motownbrowne]

Weather underground for me. During syrup season, I pay a LOT of attention to several forecasts. Weather underground is the best in my opinion. It has a slick little graph on the page that displays temperature, wind speed and direction, and percent chance of precipitation. I can get a detailed forecast at a glace.