a different type of computer model.
Bennett began his career as a meteorologist, working as a forecaster at The Weather Channel, WeatherData, and other media outlets. He later obtained a law degree, and eventually landed at Enron, where he helped develop extended forecasts used by the energy giant in natural gas commodities trading.
After Enron imploded in scandal, Bennett says their seven-person team joined the Chicago-based Citadel Investment Group, which expanded their group to 150 people in a global energy trading business focused on oil and natural gas. “The gist of the job was to use weather forecasts and related data to support our commodities pricing forecasts in futures markets,” he says, explaining that trading in a variety of commodities futures can be highly influenced by extreme weather events that affect supply and demand. During his last two years at Citadel, Bennett said he focused on insurance and reinsurance investment strategies for large swaths of U.S. coastal areas that are subjected to hurricanes.
Because extreme weather can disrupt commodities markets, Bennett says he became increasingly interested in the possibility of using predictive analytics to extend the window of conventional weather forecasts beyond 10 to 14 days. Developing such capabilities, however, required the kind of expertise that usually resides in research universities.
Bennett says that Tony Haymet, director of Scripps Institution of Oceanography, viewed the research as a valid scientific issue for the school, which has gained recognition for its research into global climate change. With guidance from Bennett, who joined Scripps on a part-time basis, climate scientist Alexander Gershunov and postdoctoral researcher Kristen Guirguis developed a proprietary methodology to validate whether certain weather patterns were statistically consistent. For example, when a certain jet stream pattern forms over Alaska, the odds that severe cold will descend on the Midwest increase significantly.
The research succeeded in identifying a
