Why Weather is So Hard to Predict ...And Why I Love Predicting It

Meteorologists rely heavily on computer models and atmospheric simulations to predict future weather conditions. However, these predictions are only as good as the models' understanding of the underlying physics/thermodynamics (and to a lesser extent, chemistry). And since atmospheric processes are so complicated, and depend on an overwhelming number of separate factors/influences, these models have definite limitations.

While the computer models meteorologists use have improved vastly over the past 20 years, as a 5-day forecast is now as about accurate as a 3-day forecast was in the 1970s, there is still a good deal of error in these forecasts.

Case in point: one of the 2 major US computer models (there are many other national and global models) makes forecasts out 384 hours every 6 hours. This evening's "run," 5 hours ago, predicted that the temperature at 11pm would be 37 degrees. Instead, we're exactly 20 degrees warmer than that: currently it's 57.

Moral of the story? If a five hour forecast is off a whopping 20 degrees, which is an enormous bust, how do we expect these models to accurate portray what's going to happen in 24 or 48 hours, let alone 16 days from now -- which is how far out this model forecasts? That's why a not insignificant basis of my weather forecasts is from the gut/intuition as much it is from computer models.

I would be remiss if I didn't add that forecasting weather in the Boulder/Denver region is especially tricky. In fact, it's probably one of the most difficult areas in the country to forecast for because of (among other regional factors) the huge effect the Rockies have on weather patterns. "Mountain waves," "Chinook winds," and"lee troughs," and many other terms describe some of the many complexities associated with mountain weather prediction.

But all of these complexities and challenges are what make me so enamored with weather and meteorology: I, personally, see weather forecasting a bit like assembling a jigsaw puzzle -- putting lots of individual pieces of information together to form one coherent whole; but one missing piece can seriously alter the picture.

Luckily, as I've heard several times before: what other job position allows you to be wrong 50% of the time and keep your job? Now, I take exception to that "50%," and believe that my forecasts are far better than tossing a coin; that being said, I'll be the first to admit that they are far from perfect. And they never will be. Thank you, chaos theory.