Forschungskolloquium PG: Deep Learning for Supraglacial Lake Segmentation in the Karakoram. – Mi. 24.06.2026

Physisch-geographisches Kolloquium (PG):

Deep Learning for Supraglacial Lake Segmentation in the Karakoram.

Nicolas Munoz (International Doctorate Programme: Measuring and Modelling Mountain glaciers and ice caps in a Changing ClimAte (M³OCCA))

Moderation: Prof. Dr. Matthias Braun

Supraglacial lakes are bodies of water that form on a glacier’s surface and influence glacier mass balance through ablation as well as glacier dynamics. Within High Mountain Asia (HMA), the Karakoram Mountains have one of the highest instances of supraglacial lake area. Since glacial runoff provides a large amount of the water supply for the Upper Indus Basin (UIB), the impact of supraglacial lakes on mass balance could potentially have negative consequences on water availability for populations downstream. This could have implications in terms of water, food, and energy security. Therefore, it is essential to monitor supraglacial lakes efficiently on a regional scale using remote sensing techniques. However, traditional techniques based on manual delineation or shallow machine learning models are insufficient for the task. Deep learning provides a potential solution to this through efficient and accurate monitoring of supraglacial lakes at large scales.

Wann: Mittwoch, 24.06.2026 von 12:30 bis 14:00 Uhr
Wo: Präsenz, Seminarraum 00.210 Tennenlohe, Wetterkreuz 15, 91058 Erlangen.

Vortragsübersicht für das Institutskolloquium im Sommersemester 2026.

******