That's according to a new study published this week in the Proceedings of the National Academy of Sciences.
Researchers looked back at the winters of 2003 to 2008 in New York City and examined real-time data from Google Flu trends, a website which aggregates online search queries. They were able to develop a model that could forecast the timing of a flu outbreak more than seven weeks ahead of the actual peak.
Dr. Jeffrey Shaman of Columbia University is the study’s lead author.
“Not only was the model, at times, predicting when the peak of the influenza outbreak would occur well in advance, it was also giving us indictors of how clear that prediction was," he says.
Flu epidemics tend to happen after very dry weather, and the model takes humidity into account, but Shaman those real-world observations were also critical to making it work.
"[Dry weather] seems to be a key factor in determining that we get influenza in the winter and not summer," he says, "But it's not going to tell you that we're going to have an outbreak and it's going to peak on December 15 this year, and next year it's going to peak on February 25th. That's where the observations and the data assimilation come in."
Shaman adds that a flu forecast could eventually be broadcast along with the local weather forecast to help inform the public and health officials about the potential risk.
The study was co-authored by Alicia Karspeck of the National Center for Atmospheric Research; it was funded by the National Institutes of Health and the Department of Homeland Security.
It’s estimated that influenza kills about 35,000 people in the U.S. each year.