Statistics seminar

Abstract: Estimating population growth at a local scale is crucial for understanding ecological dynamics and informing conservation efforts. In this work, we propose a novel methodology that uses marked point processes to model bird populations as a spatial process of unknown intensity influenced by environmental factors. The population is represented as a point process P, while observations are obtained through a thinning mechanism using a homogeneous Poisson birth-and-death process Ot. This approach accounts for both spatial dependence, such as Cox or Gibbs processes, and temporal variations by focusing on intersections of observed areas over consecutive time points.

By applying stabilization theory, we demonstrate convergence properties and ensure robust local estimations of population variation. The ability to estimate population dynamics at fine spatial scales distinguishes this approach, making it particularly suited to ecological applications.

An implementation using French breeding bird survey data highlights its potential to capture localized trends and advance biodiversity monitoring.