Longitudinal subglacial bedform semi-automated mapping and measurement

This thesis addresses methodological issues in the morphometric inventorying of relict drumlins and mega-scale glacial lineations (longitudinal subglacial bedforms, LSBs) which pose limits to a robust description of LSB morphometry and thus to testing hypotheses of LSB genesis, with implications for postdicting past, and predicting future, ice sheet behavior. Focus is on a) the adequacy of previously used morphometric measurement methods (MMM) (GIS) and b) the development of LSB semi-automated mapping (SAM) methods. Dimensions derived from an ellipse fitted to the LSB footprint based on Euler’s approximation are inaccurate and both these and orientation based on the longest straight line enclosed by the footprint are imprecise. A newly tested MMM based on the standard deviational ellipse performs best. A new SAM method outperforms previous methods. It is based on the analysis of normalized local relief closed contours and on a supervised ruleset encapsulating expert knowledge, published morphometric data and study area LSB morphometry.