Identification automatique des vallées glaciaires à partir d'un modèle numérique de terrain

Authors: Bakari, Bilel
Advisor: Guilbert, EricMoulin, Bernard
Abstract: The glacial valley is a landform associated with mountains that have been affected by regional glaciation. The study of this kind of valley helps researchers to better understand the development of the regional glacial system that reflects global climate change. The task of identifying these forms on a digital elevation modelis essential for all geomorphological or environmental studies. With the development of GIS and IT tools, the automation of this task becomes a solution for reducing the time and cost required by the manual method.However, current automatic methods like morphometric classification or Object Based Image Analysis are limited when trying to identify an entire landform such as the glacial valley. These methods are generally based on land classification approaches of terrain elements without taking into consideration the geomorphological context. In this respect, we aim to develop an automatic method for extracting glacial valleys. Our method is based on the idea of salience resulting from human cognition of relief. We admit that the glacial valley is defined by the spatial assembly of the bottom and slopes in each geomorphological context. The identification of these forms is based on the identification of streams and foothills as salient elements of its components and their assembly around a global salience, the thalweg. We applied our automatic method on digital terrain models of different glacial zones, including the Jacques-Cartier Valley in Quebec. We obtained spatial entities that characterize the geographic extent of the glacial valley and its component elements accompanied with form indicators.
Document Type: Mémoire de maîtrise
Issue Date: 2019
Open Access Date: 29 November 2019
Permalink: http://hdl.handle.net/20.500.11794/37413
Grantor: Université Laval
Collection:Thèses et mémoires

Files in this item:
Description SizeFormat 
35803.pdf5.18 MBAdobe PDFThumbnail
View/Open
All documents in CorpusUL are protected by Copyright Act of Canada.