An integrated Structure-from-Motion and time-lapse technique for quantifying ice-margin dynamics

Fine resolution topographic data derived from methods such as Structure from Motion (SfM) and Multi-View Stereo (MVS) have the potential to provide detailed observations of geomorphological change, but have thus far been limited by the logistical constraints of conducting repeat surveys in the field. Here, we present the results from an automated time-lapse camera array, deployed around an ice-marginal lake on the western margin of the Greenland ice sheet. Fifteen cameras acquired imagery three-times per day over a 426 day period, yielding a dataset of ~19 000 images. From these data we derived 18 point clouds of the ice-margin across a range of seasons and successfully identified calving events (ranging from 234 to 1475 m2 in area and 815–8725 m3 in volume) induced by ice cliff undercutting at the waterline and the collapse of spalling flakes. Low ambient light levels, locally reflective surfaces and the large survey range hindered analysis of smaller scale ice-margin dynamics. Nevertheless, this study demonstrates that an integrated SfM-MVS and time-lapse approach can be employed to generate long-term 3-D topographic datasets and thus quantify ice-margin dynamics at a fine spatio-temporal scale. This approach provides a template for future studies of geomorphological change.