UGA Research Magazine

Forecasting: Weather Behavior Depends on Demographics

by Carole VanSickle



As a twelve-year old in North Carolina, Alan Stewart wanted to predict the weather to find out when snowfalls would cancel school. Years later, the professor of counseling and human development at the University of Georgia still likes the idea.

But instead of hoping for a lazy morning at home, Stewart now analyzes weather forecasts in terms of how weather descriptions affect the decisions and morale of the people hearing them. And he’s developed a new banner under which to study this unique hybrid of meteorology and psychology: behavioral climatology.

“Weather affects all aspects of human behavior,” Stewart said. “The things people do, the ways they think, what they think about and how they view themselves as a function of that.”

To analyze how people view themselves in meteorological lights, Stewart compiled a list of 153 weather-related words and asked residents of the southeast to rate the words on a 1-5 scale with 1 being the least descriptive of local climate and 5 being the most.

He then used the results in tandem with respondents’ demographic origins and their favorite types of weather to determine what words most clearly conveyed classic meteorological conditions like “hot,” “cold and snowy” or “severe” weather.

Stewart, an associate member of the American Meteorological Society, says it’s important to understand culturally determined climatological distinctions in order to deliver forecasts more effectively. For example, “hot” weather will likely be interpreted very differently by a Seattle native and a lifelong Georgian.

His current research focuses on hurricane categories, which he believes may not sufficiently communicate danger levels. “Storm damage is a function of the maximum wind speed of the hurricane, cubed,” Stewart said. However, the categories number 1 through 5, so people visualize the threat linearly.

“People tend to think if they made it through a 2, for example, then the next step up won’t be that much worse. They’re wrong—it’s an exponential increase in danger.” Stewart conducted an extensive survey of Gulf Coast residents during the 2006 hurricane season to confirm how individuals interpret the categories.

“The results are quite eye-opening,” he said. “Although people state familiarity with the Saffir-Simpson hurricane scale, they do not know just how damaging major hurricanes (category 3 and above) can be. “It’s a strong argument for revising the current warning systems,” he said.

For more information contact Alan Stewart at


Research Communications, Office of the VP for Research, UGA
For comments or for information please e-mail:
To contact the webmaster please email: