Weather Blog

And you thought forecasting a week ahead was tough!

And you thought forecasting a week ahead was tough!

It's an area that has high interest, but low accuracy.  Trying to forecast the weather a few days out is challenging enough, but imagine trying to forecast the weather 30-90 days out, or even next year?

But brave climatologists at NOAA's Climate Prediction Center do jus that, using a whole mess of weather observations, ocean temperatures, and computer model data to try and at least gauge the overall trend of where the weather patterns are heading.

You won't find, say, a forecast of rain or sun for July 4th this early, but instead just whether odds are higher or lower than 50/50 of being above or below normal temperature and rainfall for given 30 and 90 day periods, up to 18 months in advance.

Here's an example of their current 90 day forecast for the February, March and April time frame. First up: temperature:

The light blue area over the Northwest means there is a 33 percent chance over 50/50 that temperatures will average below normal through this three month period, and a 40 percent chance over 50/50 over Washington.  In other words, don't put away the heavy coats yet.

On the flip side, the southern half of the U.S. is looking like a warm rest-of-winter. 'EC' means equal chance either way. (Or, break out the loose change and start flipping.)

Here is the rain fall one, with similar legend, only above normal rain is green and below normal rain is brown:

This spells bad news for Kentucky, who really doesn't need anything more falling from the sky after their big ice storm, but then again, combine the two maps and it looks like it'd be warmer and wetter.

Anyway, this forecast generally coincides with the fact that we are drifting a bit back toward La Nina ocean pattern, which is where we were last winter and spring that kept us cool and wet. Maybe Punxsutawney Phil was right?

So, how accurate are these things? Hard to say. Like I said, long range forecasting is very difficult, even in general terms, and I've found these charts not too spectacular.

Research meteorologist Mark Albright found this chart from October 16th, predicting the long range forecast for November through January. Note the expected very high chance for warm temperatures in the upper Midwest and southern Alaska:

(Note, in my original blog posting on Monday, I had the wrong map listed here. This is the correct map)

How'd that work out? Several areas in Wisconsin set monthly snow records and in Chicago they just had their 18th coldest December and 10th coldest January on record. And Alaska? Anchorage was 3-4 degrees colder than average in December and January.

But if you want to soldier on, here is the link to find these charts.

Or, there's always the Farmer's Almanac, which, incidentally, did do pretty well in nailing the cold snap of late December, but bonked November's warmth pretty bad, plus it missed the big floods of early January.