The PhD will aim at modelling using multi-source Earth Observation data the use of landscape by domestic and wildlife animal populations to better characterize the contacts and their determinants and to estimate the risk of pathogen transmission. We propose to take as a model the contacts between domestic and wild herbivores on the periphery of protected areas in southern Africa, a region of the world with many national parks and where animal movements between natural and anthropic habitats are frequently observed in both directions. Two diseases will be studied: foot-and-mouth disease and Rift Valley fever. 

The thesis work will be based on telemetry data, surveys and epidemiology data already available, and will focus on the development of innovative spatial modeling methods for the simulation of animal mobility. The main methodological challenges that will be addressed are the assimilation of Earth Observation multi-sensor data in the models, and the modeling of mobility at different spatial scales. The thesis will be structured in three parts:

Use of remote sensing to characterize and monitor the environmental determinants of space occupation, movements of and contacts between wildlife and domestic animals. The potential of Earth Observation data at different spatial and temporal resolutions to describe the main known determinants (land use and boundaries, vegetation activity, water resource dynamics, savannah fires, villages and associated crop and hunting activities) of habitat and mobility of domestic and wild herds will be evaluated. Processing chains will be developed for the processing of multi-spectral images at medium (MODIS) and high (Sentinel-2, SPOT-6) spatial resolutions.

Modeling the dynamics of land use taking into account these environmental determinants.This second step involves the development of a model that will integrate the spatial and functional relationships that underlie the use of space by herds of wild buffaloes and livestock, as well as the dynamics of the landscapes in which they evolve. This cartographic simulation tool will (1) synthesize biological and ecological knowledge (population dynamics, behavioral) on mobility and herd dynamics; (2) provide a better understanding of the impacts of landscape structure and its intra-annual variations on animal mobility and on wildlife/livestock contacts. For the development of the models, the Ocelet domain specific language developed by TETIS research unit will be used (http://www.ocelet.fr).

Modeling the transmission of pathogens between wild and domestic animal populations.The aim of this third part is to apply the animal mobility models developed in Part 2 to the study of transmission dynamics of two pathogens associated with important diseases, a direct transmission animal disease, foot-and-mouth disease, and a vector-borne zoonotic disease, Rift Valley fever.  Infectious contact definition will vary between diseases (i.e. based on pathogen-specific modes of transmission) and will impact the risk of pathogen spread at the wildlife/livestock interface with different implications for disease prevention and control.