| Summary: | Evidence-based geomorphological research (analyses of glacial landform record) is one of the commonly used types of glacial reconstructions applied for studying the past ice sheets. The work covered by this thesis contains a geomorphological data set of more than 17 000 glacial and glacifluvial landforms in Gausdal Vestfjell, south-central Norway. The study area is located in an inner region of the former Fennoscandian Ice Sheet (FIS) and contains a palaeogeological record on the flow pattern evolution and retreat during the Late Weichselian. The geomorphological mapping was carried out by exploiting the accessibility of high resolution LiDAR (Light Detection And Ranging) data and the capabilities provided by the latest advances in Geographical Information System (GIS) technologies such as terrain visualisations in an interactive 3D environment. A consequence of the ‘age of LiDAR’ is the increase in the resolution of the findings with more and smaller-sized landforms being mapped, that earlier were likely to be left out as unrecognised. These small-sized landforms may provide information on relatively short-lived events representing changes of glacial dynamics. This is particularly the case when such landforms are found overlapping other larger landforms. The mapping results have been validated by field observations. Since the manual mapping of landforms is a time consuming and potentially subjective process, part of the thesis explores the semi-automated mapping (SAM) techniques that may be a reliable and effective alternative for data extraction. Several SAM methods are proposed for glacial streamlined landform extraction, yet none was considered optimal for the extraction of glacial ice flow directions in a complex terrain, such as the study area in the Scandinavian Mountains. Therefore, the potential of the grayscale thinning (skeletonisation) application for the extraction of directional trends from the terrains is explored. The meltwater landform domain is a valuable source of information on the dynamics ...
|
|---|