Extraction and Area Change Analysis of Supraglacial Lakes in Greenland Ice Sheet Using U-Net Model

Objectives As a sensitive indicator of global changes such as ice sheet mass loss, sea level rise and climate change, supraglacial lakes play an important role in the study of ice sheet stability.

Methods U-Net is a semantic segmentation model, which extracts supraglacial lakes from Landsat satellite images in different situations.

Results The extraction of supraglacial lakes by U-Net possesses the optimal performance and efficiency. And the supraglacial lakes near the Sermeq Avannarleq Glacier in Greenland ice sheet are extracted from 2000 to 2020, and the seasonal and interannual variations of the area are analyzed.

Conclusions The area of supraglacial lakes starts to increase in mid-May, reaches the peak in the mid-late-July, and almost disappears in September in a year. For the long time series change, the area of supraglacial lakes in this region shows an increasing trend from 2000 to 2020, indicating that the amount of ice sheet melting is gradually increasing.