Intégration de plusieurs formes de représentations spatiales dans un modèle de simulation
Abstract
In this paper, we focus on modeling expert knowledge for simulating complex landscape
spatial dynamics. One modeling tool to do that is the Ocelet modeling language that uses
interaction graphs to describe spatial dynamics. Most present approaches impose an a priori
choice of spatial format between: (i) a vector format representing the shapes of the entities, or
(ii) a gridding of space into regular elements (raster). In this paper we show how Ocelet was
extended to support the interaction semantics between these two spatial formats (vector and
raster). As case study, we present a runoff model in a tropical insular environment.