Weather Forecasting Is Really Chaos

            Everyone complains about the inaccuracy of weather forecasts, but consider the plight of American Airlines and American Eagle.  On April 29, 1995, an unexpected storm dropped hail nearly the size of footballs on their fleet of jets at the Dallas/Forth Worth International Airport, damaging the planes and forcing the airlines to cancel 450 flights over the next week.

          It’s a sad tale, and one echoed at one time or another by everyone who can recall the day—or perhaps the hundreds of days—that the forecasters bungled things enormously, necessitating a change in the weekend plans, or, in the case of American and American Eagle, causing damage and lost profits in excess of  $26 million. 

          First announced in 1989, the modernization program has been sidetracked by delays, cost overruns, and technical problems.  But it’s not all bad news.  Just a few months ago, for example, the weather service announced the successful positioning of a new satellite that for the first time, will offer pictures of developing weather patterns over the Pacific Ocean.

          The modernization program also includes several other items of interest:

         -- A network of Doppler radars that track the speed and position of moisture in the air.  Computer programs that analyze data from these radars can tell the amount of rain or snow falling to earth, the power and path of thunderstorms, and whether tornadoes are beginning to form.

        -- More than 900 new “ASOS” weather stations that automatically report data on  temperature, dewpoint, wind direction, wind speed, and barometric pressure.  Eventually, these technological wonders may replace human observers.

         -- Better computers and software.  Already, the machines that run the forecaster’s weather models are 20 times more powerful than a decade ago.  The models themselves, which extrapolate future weather events from known conditions, are also much-improved, affecting the speed and reliability of the forecasting process.

          The phenomenal amount of money being spent on these improvements is probably justified, given the way the nation’s economy depends on reliable forecasts.  Each day, tens of thousands of decisions are made on the basis of the weather by farmers, construction workers, commodities traders and utility companies.

         So important are reliable forecasts, in fact, that a $150-million a year industry has grown up to provide supplemental, private weather data to businesses and individuals.   Paying the money to get it right makes sense.

         The trouble is, despite the money, meteorologists aren’t getting it right.  Even with the weather service’s ongoing “modernization,” attempts to forecast the weather past just a few days are still basically unreliable. What’s the problem?

            It’s the meteorologist’s dirty little secret: though forecasts have improved over the last two decades, and will continue to improve slightly, unpredictability is inherent in the weather.  Though it’s not their fault, the meteorologists will never be able to get it right.  The golden age of weather forecasting?  Forget it.

         There’s a scientific term for this depressing truth: chaos.  The earth’s atmosphere is a “chaotic” system of the sort first described in a 1963 paper by meteorologist Edward Lorenz.  In simple terms, this means that unlike “deterministic” systems such as the planetary orbits around the sun, in which tiny errors in the starting conditions don’t affect long-term forecasts, small differences in the weather do just the opposite—they multiply, quickly throwing even the simplest predictions into disarray. 

          The implications of chaos theory did not fully dawn on the rest of the scientific community until the mid-1970s.  Twenty years later, for reasons both complex and profound, chaos theory is now considered meteorology’s great contribution to 20th century science, with applications for biologists, physicists, astronomers, geologists, and physicians.

           But don’t expect to hear about chaos theory during the weather segment of the evening news.  After all, why should the forecasters reveal their meteorological fine print to the public—that they can predict temperatures easily enough, but rain and snow are constant headaches?  That seven-day forecasts are a pipe dream, and two-week forecasts a hallucination?  That despite scientific advances and “modernization,” they’re still not too far removed from soothsayers and fortune-tellers?  That chaos makes their job impossible? 

         No, this would be sheer stupidity.  The fact is, the public is fascinated with weather, and a good dose of the truth would only spoil the mood.   As this fascination is demonstrated once again by the public’s interest in spring storms and the success of the  movie “Twister,” the informed weather-watcher should understand that it’s simply not fair to blame the meteorologists.

       To paraphrase Edward Lorenz, what’s amazing isn’t that the meteorologists botch so many forecasts, but that they occasionally get some right.  Rather than jeer, you should await the results of the weather service’s probably-modest improvements with an open mind.  And while you’re waiting, give a kind word to your local forecasters, and compliment them on their usually-good performance of one of the toughest jobs on earth—unless, that is, you’re an executive with American Airlines.  

Back