Published on September 12th, 2012 | by Joshua S Hill0
New More Accurate Method of Predicting Hurricane Activity
Predicting hurricane activity is vitally important for the people who live in high-hurricane-risk regions, allowing businesses and governments, not to mention the populace, time to prepare and plan their way through the next hurricane season.
Researchers from North Carolina State University have recently created a new method in forecasting seasonal hurricane activity that is 15 percent more accurate than the previous technique had been.
“This approach should give policymakers more reliable information than current state-of-the-art methods,” says Dr. Nagiza Samatova, an associate professor of computer science at NC State and co-author of a paper describing the work. “This will hopefully give them more confidence in planning for the hurricane season.”
What Came Before
Previously, hurricane predicting relied on only 60 years worth of historical data and were forced to work through a horrific amount of variables to come to any decent conclusion.
In the end, the accuracy rate for previous methods only amounted to around 65 percent.
Researchers have now created a “network motif-based model” that evaluates historical data for all of the variables in all of the places at all of the times. This allows the model to determine what combinations of factors are most predictive of a hurricane season.
Not only that, however, but the new method allows the model to determine what predictors are representing a quieter hurricane season, and what predictors are representing a heavier more vicious hurricane season.
The researchers used cross validation — “plugging in partial historical data and comparing the new method’s results to subsequent historical events” — to find that their new method had an 80 percent accuracy rating of predicting the level of hurricane activity over the sample period given it, up 15 percent from the old predictive methods.
On top of it all, the new model is not only confirming previously identified predictive groups of factors, it has gone and identified some new ones.
Source: North Carolina State University