A hybrid approach to study and forecast climate-sensitive norovirus infections in the USA

Abstract

Norovirus, responsible for acute gastroenteritis and foodborne diseases in the United States, is influenced significantly by environmental factors. This study employs a hybrid approach to develop a foodborne disease model that incorporates indirect incidence to examine the correlation between norovirus outbreaks and environmental conditions, specifically focusing on the impact of temperature and humidity on virus transmission. By analyzing weekly average climate data and confirmed case data from four United States regions (Southern, Northeastern, Midwestern, and Western), we assess the mortality rates and estimate transmission rates using the inverse method. Our numerical results confirm that norovirus outbreaks predominantly occur in colder months. However, higher temperatures or increased humidity during warmer months appear to mitigate the spread of the virus. Utilizing climate data, this study also forecasts transmission rates and infection cases up to eight weeks in advance using a generalized boosting machine learning model.