Methodology:

Data were derived from a time-series of point clouds of ice-margin topography. The point clouds were generated at 24 hour to 28 day intervals by using Structure-from-Motion techniques to process discrete image sets acquired from between 7-15 time-lapse cameras. Successive point clouds were co-registered and differenced to quantify calving properties. Classification of calving mechanisms was based on geometric changes at the ice-margin, interpretation of camera imagery pre- and post-calving, and existing calving categorisations. Full details of camera installation and point cloud processing are included in the following:

Mallalieu et al. 2017 An integrated Structure-from-Motion and time-lapse technique for quantifying ice-margin dynamics. Journal of Glaciology, 63(242), 937-949. doi:10.1017/jog.2017.48

Mallalieu, J., Carrivick, J. L., Quincey, D. J., & Smith, M. W.. (2020). Calving seasonality associated with melt-undercutting and lake ice cover. Geophysical Research Letters, 47, e2019GL086561. https://doi.org/10.1029/2019GL086561

Data collection:

Time-lapse camera imagery was acquired by LtL Acorn 5210A trail cameras. Ground Control Points were surveyed using a Leica System 500 dGPS. Camera imagery was processed in Agisoft Photoscan Professional v.1.2.5 to generate point clouds. Point clouds were co-registered, differenced and analysed in CloudCompare v.2.7.

Resolution: Point cloud densities range from 18-37 pts m2

Data quality:

Point cloud quality varies throughout the survey due to the nature of the time-lapse camera setup. Please see the dataset for parameters of individual point clouds, including: georeferencing RMSE; point cloud size; point cloud density; number of ground control points; and number of aligned cameras. Variability in the sampling interval between successive clouds was due to inclement weather conditions.