Use of open-source GIS for the pre-processing of distributed hydrological models
Utilisation de logiciels SIG open-source pour le pré-traitement de modèles hydrologiques spatialisés
Résumé
In distributed hydrological models, elementary modelling units are often derived from raster and vector geographical maps such as digital elevation models (DEM), land use and soil properties maps. Model pre-processing can be defined as the operations transforming raw GIS data into the adequate modelling units. This paper describes our experience with open-source GIS for these pre-processing operations through a set of examples. The softwares tested were PostGIS, OpenJump, OrbisGIS, GRASS, SAGA, MapWindow and QuantumGIS. The pre-processing operations dealt with DEM processing, topology corrections, and advanced editing of vector layers. None of the tested software provided a ready to use solution. However GRASS GIS appeared as the most promising candidate for performing all the preprocessing operations and automating them. Although sometimes tricky to handle by GIS non-expert users, open-source GIS allow easy tests and comparisons and therefore are well suited to research purposes.