CMStatistics 2018: Start Registration
View Submission - CMStatistics
B0543
Title: Spatio-temporal modelling of the spread of chalara (illness of the ash tree) in France Authors:  Anne Gegout-Petit - INRIA Universite de Lorraine-IECL BIGS (France) [presenting]
Coralie Fritsch - Inria (France)
Benoit Marcais - INRA (France)
Marie Grosdidier - INRA (France)
Abstract: Chalara is an illness of the tree ash that appears in Europe by Poland in 1992 and was observed for the first time in France in 2008. From 2008 to now, around 500 places of forest are visited every year in the purpose to notice the proportion of infected ashes. The mechanism of the illness is due to a fungus that grows during summer on ash rachis in the previous year's fallen leaves. The spores are produced and are transmitted by wind. According to this knowledge about chalara and in the purpose to a better understanding of the effect of covariates on the spread of the illness, we have built a spatio-temporal mechanistic model of propagation of the illness. It is based on a latent parametric model of spores production accounting on the effect of humidity and temperature and a reaction-diffusion model for the spores diffusion. For the inference, only based on the proportion of infected trees, we have used a Bayesian framework and MCMC simulations.