| Summary: | Valleys are rarely studied as specific relief elements. We propose a typology based on morphological data processing from a digital elevation model (DEM). Valleys were extracted from the hydrographic network and STRAHLER stream ordering, as well as a width strip, which varied depending on the ordering. Elevation data were processed using the "Topographic Position Index" (TPI). Changes in the slope at the bottom and on the sides of the valley as well as the width of the valley floor were the main criteria used for segmentation of the valley into homogeneous sections. For each section, twelve variables were statistically analysed to build a typology. The method was applied to the Maine watershed (22,300 km2), which has a high density of valleys in the sedimentary strata in the western part of the Paris Basin as well as in the hard rocks of the Armorican Massif. Data analys identified 548 sections divided into twelve types and subtypes. Results reveal a wide range of forms whose distribution is not explained by the stream ordering, but is clearly linked to the nature of the substratum rocks. These results can be used in different approaches for the identification of landscape units and to study floodplain areas.
|
|---|
