Thaw-Season InSAR Surface Displacements and Frost Susceptibility Mapping to Support Community-Scale Planning in Ilulissat, West Greenland

• Published in: Remote sensing (Basel, Switzerland) Vol. 15; no. 13; p. 3310
• Main Authors: Scheer, Johanna, Caduff, Rafael, How, Penelope, Marcer, Marco, Strozzi, Tazio, Bartsch, Annett, Ingeman-Nielsen, Thomas
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 01-07-2023
• Subjects: active layer; Analysis; Arctic infrastructure; Artificial satellites in remote sensing; Built environment; Climate change; Community planning; Composition; Construction; Construction planning; Deformation; Freeze-thaw; Freezing; Frost; Geospatial data; Ground ice; ice content; Image classification; Information management; Infrastructure; InSAR; Interferometric synthetic aperture radar; Land cover; Mapping; Melting; Methods; Permafrost; Remote sensing; Risk management; Seasons; Sedimentary basins; Sediments; Soil degradation; Spatial discrimination; Spatial resolution; Statistical models; Surface properties; Synthetic aperture radar; Thawing; Thickness measurement; Trends; Urban environments; Greenland; Arctic region
• ISSN: 2072-4292; 2072-4292
• DOI: 10.3390/rs15133310

In permafrost regions, ground surface deformations induced by freezing and thawing threaten the integrity of the built environment. Mapping the frost susceptibility of the ground at a high spatial resolution is of practical importance for the construction and planning sectors. We processed Sentinel-1 Interferometric Synthetic Aperture Radar (InSAR) data from thawing seasons 2015 to 2019, acquired over the area of Ilulissat, West Greenland. We used a least-squares inversion scheme to retrieve the average seasonal displacement (S) and long-term deformation rate (R). We secondly investigated two different methods to extrapolate active layer thickness (ALT) measurements, based on their statistical relationship with remotely sensed surface characteristics. A generalized linear model (GLM) was first implemented, but the model was not able to fit the data and represent the ALT spatial variability over the entire study domain. ALT were alternatively averaged per vegetation class, using a land cover map derived by supervised classification of Sentinel-2 images. We finally estimated the active layer ice content and used it as a proxy to map the frost susceptibility of the ground at the community scale. Fine-grained sedimentary basins in Ilulissat were typically frost susceptible and subject to average seasonal downward displacements of 3 to 8 cm. Areas following a subsiding trend of up to 2.6 cm/yr were likely affected by permafrost degradation and melting of ground ice below the permafrost table. Our approach enabled us to identify frost-susceptible areas subject to severe seasonal deformations, to long-term subsidence induced by degrading permafrost, or to both. Used in combination with traditional site investigations, InSAR maps provide valuable information for risk management and community planning in the Arctic.