Always offered between semesters, 12.310 reveals the principles of fluid dynamics that govern the atmosphere’s movement. In lectures, real forecasting exercises, and a trip to WBZ TV’s on-air weather center, students learn where newsrooms get their daily weather predictions — and how they can make their own.
“Our aim is that students come to understand that weather forecasting is not guessing and is based on real science,” says Lodovica Illari, senior lecturer in synoptic meteorology in MIT’s Program for Atmospheres, Oceans, and Climate (PAOC), who has taught the course for 20 years. “We hope the students get interested in fluid dynamics and come back and do some more.”
To reveal the laws governing the atmosphere’s motion “in action,” Illari used demonstrations with MIT’s Weather in the Tank materials. In one class, Illari rotated a cylindrical tank of water on a turntable to emulate the dynamics of Earth’s atmosphere. A bucket of ice at the tank’s center created a temperature gradient. As students peered in, Illari dropped ink into the water to make visible the small currents forming in the water, demonstrating the conditions that create weather systems.
12.310 stands as a yearly reminder of MIT’s distinguished history in weather forecasting. In fact, Carl-Gustaf Rossby, who transformed weather forecasting into an atmospheric science, founded the country’s first meteorology program at MIT in 1928. A line of prominent MIT faculty meteorologists followed, including Jule Charney, Norman Phillips, Victor Starr, and Edward Lorenz. The fundamental contributions to fluid dynamics made by these past professors resonate in MIT’s modern-day research into much longer-time-scale atmospheric phenomena such as climate.
As the students of 12.310 learned the basics of weather forecasting, they came to appreciate an insight of Professor Lorenz’s that changed meteorology forever. At MIT in 1961, Lorenz found that rounding a few digits off one decimal number in a computer weather simulation changed its projected long-term pattern. He had discovered the big implications of the chaos theory principle of “sensitive dependence on initial conditions” for weather prediction. Slightly imprecise measurements, even single-digit differences in dew point, can skew long-term forecasts. The chaotic nature of short-term weather places a limit of less than 10 days on accurate forecasts.
To make their own forecasts, the students used raw data spit out of weather prediction models run at the National Weather Service, which are read out in maps of the temperature, air pressure, dew point, wind, and moisture across the country. With that information, the students anticipated the likely passage of warm or cold weather fronts by searching for regions of strong gradients in temperature or moisture and shifts in wind direction. The class ended with a TV-style presentation of the day’s weather and a Boston weather forecasting competition. This year, Fiona Paine ’17 took home the prize of a digital indoor/outdoor thermometer.
The course was organized and taught by Illari and co-instructor Jeff Scott, a research scientist at PAOC and the Center for Global Change Science, along with teaching assistants Casey Hilgenbrink ’15 and PAOC graduate student Vince Agard ’11, who are both members of the MIT Weather Forecasting Team.
Many of the undergraduates signed up for 12.310 to get a taste of the kind of science they could pursue in PAOC. Others said that knowing how to predict the weather is just plain useful. And everyone found the class eye-opening. “It was surprising to me just how difficult it can be to predict tomorrow's weather, even when using the newest technology,” says Oren Katzen ’16. “I will be certain to be more forgiving to the weatherman in the future.”