A dynamically forced cholera model for operational forecasting in Haiti

Waterborne diseases are among the leading causes of deaths in developing countries. The prompt intervention during an outbreak is crucial to limit the spread of an epidemic and reduce the number of casualties. The epidemiological models developed in the past years at the ECHO lab to study cholera and schistosomiasis epidemics, produce reliable predictions of the infected population. Human mobility fluxes together with the hydrological network constitute the main spatial drivers of the model, as they delineate the possible network along which the epidemic might spread. Due to the lack of available data, human mobility is typically described through a simple gravity-model, constant in time. However, mobility fluxes might highly change, especially during the beginning of the outbreak or after a hurricane The main objective of this project is to perform a sensitivity analysis of the spatially-explicit cholera model developed at the ECHO lab to the possible estimations of human mobility. To achieve this goal, different formulations of human mobility will be considered, including both mathematical estimations via a gravity model and direct estimations derived from mobile-phone data. Depending on the availability of the mobile phone data, the analysis will be performed on the 2010 Haitian outbreak or on the 2016 epidemic dynamics in Haiti after hurricane Matthew.