Estimating active layer thickness at the high Arctic study site Zackenberg from remotely sensed ground subsidence

Abstract: The active layer thickness (ALT) is an indicator of permafrost thaw, which potentially leads to the release of large amounts of greenhouse gases under global warming, and thus could further amplify climate change. The thaw depth of the active layer also governs seasonal surface deformation, caused by the volume change between ice to water, which poses risks for slope stability and infrastructure foundations. Currently, the monitoring of the ALT is accomplished through modelling and field measurements. This thesis aims at testing a recently introduced remote sensing method for ALT monitoring by comparing the estimated ALT to ALT field measurements and environmental parameters. The method is based on remotely sensed seasonal ground subsidence, derived through the differential interferometric synthetic aperture radar technique (DInSAR). By estimating the ice content in the soil, the ground subsidence can be related to ALT development, which was used to determine the ALT throughout the thawing season 2017 at Zackenberg valley, Greenland. The ice content was derived by combining saturation fraction with porosity. The saturation fraction was determined based on a linear regression between the normalized soil moisture index and soil moisture measurements. The porosity was obtained through organic matter field measurements and a mineral soil database. The resulting ALT estimate was statistically similar to the ALT field measurements, both considering ALT maximum (p-value = 0.55) and temporal development (agreement to ideal fit: p-value = 0.33). The ALT distribution is plausible considering environmental factors (aspect, topography, vegetation). Yet, the method was found to perform best in water-saturated areas and showed more variable results in drier areas. For future applications it is recommended to incorporate modelling of the subsurface ice content, as the ALT result in drier soils is especially dependent on a correct estimate of the ice content. If this is accomplished, and if the subsidence retrieval considers 3D-ground displacement, the tested method is promising for the ALT monitoring at high spatial resolution during the snow-free season.