Ten days ago I issued my annual fall-winter forecast for Oregon (with a nod to the rest of the Northwest and the continental U.S.). Immediately I began to see misinterpretations of the forecast, especially in headlines: things like "Climatologist says to prepare for severe winter," or "wild winter with lots of snow." I thought I had made things pretty clear, but apparently I didn't. So here I am, trying to set the record straight.

First, I'll explain how I make my forecasts. The primary influence on our winter climate is the tropical Pacific. When we have an El Niño event, we're likely to get warmer and drier winters than average, with reduced mountain snow. La Niña events, on the other hand, tend to bring cooler and wetter than average conditions to the Northwest.

The southern tier of states (Southern California, Arizona, New Mexico, Texas, and so on) tend to get just the opposite, at least in terms of precipitation: wetter with El Niño and drier with La Niña.

Note that I said "tend to." By that I mean "more often than not, the following happens…" Using El Niño/La Niña for forecasts guarantees nothing, it merely changes the probabilities. For example, in an "average" year there's a 50-50 chance of above- or below-normal conditions. In an El Niño there might be a

70 percent chance of above-average temperatures and a 65 percent chance of below-average precipitation - nothing assured but pretty good guesses.

Right now there is a moderately strong La Niña developing off the South American coast. That means that the normally cool water off Peru and Chile gets even cooler than average, and the warm water in the western Pacific (near Indonesia) gets even warmer. The existence of the La Niña increases the likelihood of cooler and wetter conditions this winter.

In addition, I select "analog" years, years with general conditions (ocean and atmosphere) similar to the current year. If I find some good matches, I assume that "history repeats itself," that the winters in those analog years will repeat themselves, at least to some degree. On the one hand, no two weather years are exactly alike, but many similarities exist most of the time. The analog years that most closely resembled 2007 were 1952, 1953, 1970 and 1988.

I put it all together, with the help of my colleague, Cadee Hale, and posted the forecast on the Oregon Climate Service website. Here's the synopsis:

1. Winter conditions will be a bit late in arriving (typical of La Niña events) but will be in full swing by November. Storms will become more frequent after the first of the year.

2. Coldest temperatures will occur in the second half of the winter. At this point, February looks especially cold.

3. In any given year, there's about a 50-50 chance of low-elevation snow. This year I think the odds go up - let's just say to 65 percent in favor, though that's just a guess.

4. Overall mountain snow pack will be above average.

5. Seasonal precipitation will be somewhat above average.

For more details, including month-by-month predictions and lots of highly technical graphs and tables, visit our website.

One more thing. That "misquote" that said "Climatologist says to prepare for severe winter." That's always a good thing to do, by making sure our houses are well insulated, safe from big wind storms, and that we can cope with possible power failures and water stoppages. Those things always make sense.

A good motto: plan for the worst and hope for the best.

George Taylor is the state climatologist for Oregon and manages the Oregon Climate Service at OSU. E-mail:

taylor@coas.oregonstate.edu