| Summary: | This is the dataset of Little Ice Age glacier surfaces (as geotiff) created for the publication "Assessment of methods for reconstructing Little Ice Age glacier surfaces on the examples of Novaya Zemlya and the Swiss Alps" within the project PROTECT and the WP6. The sufaces were created using the upscaling approach. For more information about each dataset please see the publication ( https://doi.org/10.1016/j.geomorph.2024.109321 ) and cite the dataset using this doi. Climate change since the end of the Little Ice Age (LIA) has driven observed glacier volume loss, significantly contributing to the rise of global sea level. The related calculation of volume change requires knowledge of glacier surfaces from at least two points in time, usually represented by two digital elevation models (DEMs). These are typically derived from photogrammetric techniques using stereo images, but such images do not go back to the LIA. Accordingly, several techniques have been developed to reconstruct LIA glacier surfaces from historic outlines and modern DEMs. Here we first evaluate various surface interpolation methods by replicating modern glacier surfaces from outline elevation points and analyse elevation differences and uncertainties. Secondly, we investigate different GIS-based methods for LIA surface reconstruction including a new method that is based on up-scaling of recent glacier-specific elevation change data and works also for ice caps without lateral moraines. The methods were tested on 90 glaciers (covering 643 km 2 ) in southern Novaya Zemlya and 266 glaciers (524 km 2 ) in the Bernese Alps of Switzerland. As in previous studies, we also found that the Natural Neighbor and Topo to Raster interpolation methods in ESRI's ArcGIS performed best for glacier surface reconstruction and that all methods are challenged by replicating the variable surface curvature of glaciers. The new reconstruction method shows the smallest mean difference to the reference dataset (RMSE of 26.7 vs. 39.8 m). The often neglected small ...
|
|---|