Application of Humidity Data for Predictions of Influenza Outbreaks.

Seasonal influenza outbreaks infect millions of people, cause hundreds of thousands of deaths worldwide, and leave an immense economic footprint. Potential forecasting of the timing and intensity of these outbreaks can help mitigation and response efforts (e.g., the management and organization of vaccines, drugs and other resources). Absolute (or specific) humidity has been identified as an important driver of the seasonal behavior of influenza outbreaks in temperate regions. Building upon this result, we incorporate humidity data from both NASA's AIRS (Atmospheric Infra-Red Sounder) instrument and ERA-Interim re-analysis into a SIRS (Susceptible-Infectious-Recovered-Susceptible) type numerical epidemiological model, comprising a prediction system for influenza outbreaks. Data for influenza activity is obtained from sources such as Google Flu Trends and the CDC (Center for Disease Control) and used for comparison and assimilation. The accuracy and limitations of the prediction system are tested with hindcasts of outbreaks in the United States for the years 2005-2015. Our results give support to the hypothesis that local weather conditions drive the seasonality of influenza in temperate regions. The implementation of influenza forecasts that make use of NCEP humidity forecasts is also discussed.