Low-Cost Terrestrial Photogrammetry System for Monitoring Calving from Lake-Terminating Glaciers

In-situ monitoring of calving process from lake-terminating glaciers in the Himalayas remain scarce, despite being a significant component controlling glacier retreat. Acquiring in-situ measurements are challenging due to harsh terrain and extreme weather conditions, which has typically resulted in the use of expensive, state-of-the-art terrestrial observational systems. Such limitation in data acquiring leads to an incomplete understanding of calving-induced glacial mass loss and retreat. In this study, we evaluate the effectiveness of a low-cost Raspberry-Pi based terrestrial photogrammetry system for annual monitoring of calving from lake-terminating glaciers. We tested the proposed system for monitoring calving at lake-terminating Panchinala-B Glacier, Indian Western Himalayas. The photogrammetry system consists of an array of Raspberry-Pi powered time-lapse cameras, which took multi-view stereo images of the glacier front face over a 12-month period between August 2023 and August 2024. The images were processed in a Structure-from-Motion (SfM) workflow to generate two, annually separated, point clouds. The Multiscale Model-to-Model Cloud Comparison (M3C2) distance of the annually separated point clouds yielded a mean terminus position change (retreat) of 0.88 m, with a mean absolute error of (+/-) 5.8 m. Using the above obtained parameters a calving rate and calving mass flux of  16.3 (+/-) 4.29 m/annum and 0.00017 (+/-) 0.000125 Gt/annum respectively can be quantified. Further, numerous environmental and system design-based challenges were encountered, which affected the quality of the obtained calving estimates. These challenges were carefully understood, and we provide further recommendations for the future use of similar low-cost systems for long-term glacier monitoring, which demonstrate good  potential for characterising the magnitude and frequency of calving processes at lake-terminating glaciers.