A Novel Method for Automated Supraglacial Lake Mapping in Antarctica Using Sentinel-1 SAR Imagery and Deep Learning

Abstract

Supraglacial meltwater accumulation on ice sheets can be a main driver for accelerated ice discharge, mass loss, and global sea-level-rise. With further increasing surface air temperatures, meltwater-induced hydrofracturing, basal sliding, or surface thinning will cumulate and most likely trigger unprecedented ice mass loss on the Greenland and Antarctic ice sheets. While the Greenland surface hydrological network as well as its impacts on ice dynamics and mass balance has been studied in much detail, Antarctic supraglacial lakes remain understudied with a circum-Antarctic record of their spatio-temporal development entirely lacking. This study provides the first automated supraglacial lake extent mapping method using Sentinel-1 synthetic aperture radar (SAR) imagery over Antarctica and complements the developed optical Sentinel-2 supraglacial lake detection algorithm presented in our companion paper. In detail, we propose the use of a modified U-Net for semantic segmentation of supraglacial lakes in single-polarized Sentinel-1 imagery. The convolutional neural network (CNN) is implemented with residual connections for optimized performance as well as an Atrous Spatial Pyramid Pooling (ASPP) module for multiscale feature extraction. The algorithm is trained on 21,200 Sentinel-1 image patches and evaluated in ten spatially or temporally independent test acquisitions. In addition, George VI Ice Shelf is analyzed for intra-annual lake dynamics throughout austral summer 2019/2020 and a decision-level fused Sentinel-1 and Sentinel-2 maximum lake extent mapping product is presented for January 2020 revealing a more complete supraglacial lake coverage (~770 km2) than the individual single-sensor products. Classification results confirm the reliability of the proposed workflow with an average Kappa coefficient of 0.925 and a F1-score of 93.0% for the supraglacial water class across all test regions. Furthermore, the algorithm is applied in an additional test region covering supraglacial lakes on the Greenland ice sheet which further highlights the potential for spatio-temporal transferability. Future work involves the integration of more training data as well as intra-annual analyses of supraglacial lake occurrence across the whole continent and with focus on supraglacial lake development throughout a summer melt season and into Antarctic winter.

Supervised classification of slush and ponded water on Antarctic ice shelves using Landsat 8 imagery

“…Banwell and others, 2014; Dell and others, 2020; Moussavi and others, 2020) and machine learning (ML) (e.g. Dirscherl and others, 2020, 2021; Halberstadt and others, 2020). The former identifies water bodies where pixels exceed a reflectance threshold in specific bands or band combinations.…”

Recent Changes of Glacial Lakes in the High Mountain Asia and Its Potential Controlling Factors Analysis

“…More recently, there have been efforts at producing glacial lake dataset in the whole HMA region [20,34]. Apart from the optical images, high resolution satellite Synthetic Aperture Radar (SAR) data such as TerraSAR-X, Radarsat-2 [35,36], ALOS PALSAR imagery [26,37], and Sentinel-1 [38,39] can also be employed for reliable detection of glacial lakes.…”

Comparing Methods for Segmenting Supra-Glacial Lakes and Surface Features in the Mount Everest Region of the Himalayas Using Chinese GaoFen-3 SAR Images

“…Small and elongated lakes (Figure 10) were most prominent during the ablation season, and the contribution of these lakes to the total lake area was significant. However, because of the small shape and narrow width of lakes, the identification and mapping of these supra-glacial lakes have been found to be problematic in SAR data [76]. If lake width is similar to the spatial resolution of the SAR images, this may result in mixed pixels along the lake edge, thus complicating the mapping.…”