AbstractsEarth & Environmental Science

Zoonoses, diseases that usually circulate among animals and that are sometimes transmitted to humans, are complex systems that involve the pathogen, the host, (the vector) and humans. Spatial distribution models are often based on human case records as they frequently are the most readily available data. These records may be seen as the tip of the iceberg, hiding undetected zoonotic cycle. A new framework is suggested to better address the issues raised by the use of human case records for modelling zoonoses. Hantavirus and tick-borne encephalitis are examined in diverse environments and at diverse scales to illustrate these concepts. The framework is based on the concept of risk assessment that is a combination of hazard (defined as pathogen circulation in the wild) and exposure (defined as people entering into infected landscape). Results suggest that the combination of hazard and exposure is needed to improve the predictive power of models and to investigate how factors are involved in the various parts of the disease transmission system. Different modelling tools, ranging from linear regression to machine learning and from the landscape to the European scale, are investigated and compared. The multilevel approach is highly advised and three scenarios of variable response are identified, which bear diverse consequences for modelling and modelling results interpretation. (SC - Sciences)  – UCL, 2015