Well put on the "check back often" ... couldn't agree with you more. I am using the GFS which goes out 16 days, the ECM, 10 days, and the Candian (Global Model) which goes out 8 days. And of course I am analyzing other charts as well. From short to long range, I always analyze the 850MB, 700MB, 500MB, 300MB charts. You are correct that the models definitely lose their accuracy the further out you go. I've been doing this for over 20 years so I do look for consistencies when I am forecasting long range. If you look at the patterns of the different models out 11 to 14 days from today, which are the Stomp dates, the upper level patterns and storm track favor temps and precipitation close to climatology. I am not just picking the norms for those days in my current forecast. The main troughing and frontal patterns do not favor a huge cold outbreak at this time. That said, the GFS, for example, can yoyo with the best of them, and I am very much aware of that. This is exactly why I said in yesterday's initial outlook, this can very well change so please check back every day. What I am putting out right now is a preparation forecast, and each day, I will update this forecast ... sometimes more, depending on if the models change in such a manner I see it fit to put out a new forecast so folks can see what may be happening during the Stomp dates. I'm sure you have been doing this as long as I have, and if so, one becomes aware of the different nuances when forecasting short and long range. I don't shy away from sticking my neck out, and yes it has been chopped off plenty of times, but I like my track record for being correct.
I very much appreciate your input.
Take care,
Weather
DrCloud, on 06 February 2011 - 12:04 PM, said:
So...what you're doing now (more than ten days out) is pretty much forecasting to climatology, right? After all, it's well known that numerical weather prediction (computer modeling) loses skill 'round about two weeks.
Given the recent weeks of cold weather in Estes Park, I'd say it's critical for everyone involved to keep up with changes in this forecast -- meaning check back often -- as more timely information becomes available. Those 40-degree days look nice just now, but be prepared for the possibility of a a cold shock in the days to come.
(This "forecast" of mine, which says that it's been cold for the past few weeks and so may continue to be, is what we meteorologists call a "persistence" forecast. Like, if it's sunny right now and you don't see clouds, it'll stay sunny for a while. Persistence forecasts work quite well on very short time scales. Numerical weather prediction works surprisingly well on time scales from about twelve hours to ten days. After that, carefully interpreted climatology tends to be better. Because there's a large-scale event going on -- La Niña -- pure climatology won't work as well as one corrected for La Niña's effects, so my using long-term persistence in this case isn't completely off-planet.)
[/lecture mode]

hph
Edited by weatherbe, 06 February 2011 - 02:43 PM.