SEN12MS – A CURATED DATASET OF GEOREFERENCED MULTI-SPECTRAL SENTINEL-1/2 IMAGERY FOR DEEP LEARNING AND DATA FUSION

Comprehensive data sets of calving front change are essential to study and model outlet glaciers. Current records are limited in temporal resolution as they rely on manual delineation. We apply deep learning to automatically delineate calving fronts of 23 Greenland glaciers. Resulting time series resolve long-term, seasonal and sub-seasonal patterns. We discuss the implications of our results and provide the cryosphere community with a data product and an implementation of our processing system.

Multimodal data fusion is an intuitive strategy to break the limitation of individual data in Earth observation. Here, we present a multimodal data set, named MDAS, consisting of synthetic aperture radar (SAR), multispectral, hyperspectral, digital surface model (DSM), and geographic information system (GIS) data for the city of Augsburg, Germany, along with baseline models for resolution enhancement, spectral unmixing, and land cover classification, three typical remote sensing applications.