Periglacial vegetation dynamics in Arctic Russia: decadal analysis of tundra regeneration on landslides with time series satellite imagery

Abstract:

Changes in vegetation productivity based on Normalized Difference Vegetation Index (NDVI) have been reported from Arctic regions. Most studies use very coarse spatial resolution remote sensing data that cannot isolate landscape level factors. For example, on Yamal Peninsula in West Siberia enhanced willow growth has been linked to widespread landslide activity, but the effect of landslides on regional NDVI dynamics is unknown. Here we apply a novel satellite-based NDVI analysis to investigate the revegetation patterns of active-layer detachments following a major landslide event in 1989. We analyzed time series data of Landsat and very high-resolution (VHR) imagery from QuickBird-2 and WorldView-2 and 3 characterizing a study area of ca. 35 km². Landsat revealed that natural regeneration of low Arctic tundra progressed rapidly during the first two decades after the landslide event. However, during the past decade, the difference between landslide shear surfaces and surrounding areas remained relatively unchanged despite the advance of vegetation succession. Time series also revealed that NDVI generally declined since 2011 within the study area. The VHR imagery allowed detection of NDVI change “hot-spots” that included temporary degradation of vegetation cover, as well as new and expanding thaw slumps, which were too small to be detected from Landsat satellite data. The processes analyzed here are extremely widespread in NW Siberia, and in other parts of the arctic, which is why this local study has broad geographic implications for tundra vegetation dynamics.