Recent reports of permafrost degradation are widespread and models indicate that much of the near-surface permafrost in the northern hemisphere will likely disappear by the end of the current century. Warming of near-surface permafrost may lead to widespread terrain instability of ice-rich permafrost in the Arctic and the Subarctic, ultimately resulting in thermokarst development. There is increasing interest in the spatial and temporal dynamics of thermokarst and other thaw-related features from diverse disciplines including landscape ecology, hydrology, engineering, and biogeochemistry. Therefore, techniques and tools need to be developed to observe and quantify changes to near-surface permafrost terrain.
The Cold Regions Research Group at the Alaska Science Center is utilizing remote sensing datasets across a range of sensors and spatiotemporal scales to quantify the impact of disturbances on permafrost-influenced terrain. The team recently published a paper demonstrating the utility of multi-temporal airborne lidar data for quantifying thermokarst development in Arctic coastal lowlands and work continues on the use of multi-temporal airborne lidar data to study the role of tundra fires in inducing widespread thermokarst development. Scaling of the lidar-based change detection results with a Landsat-derived burn severity index explained 83% of the landscape-scale, postfire permafrost thaw subsidence. These findings will be used to scale up the impact of past tundra fires on near-surface permafrost using the Landsat Climate Data Record archive and enable future vulnerability assessments of ice-rich permafrost terrain to shifting disturbance regimes.
(a) A QuickBird image from July, 5 2008, the year following a fire, showing a portion of the northern extent of the burn area and the distinction between burned (dark gray) and unburned (light gray) tundra. Hillshade images of the (b) 2009 and (c) 2014 1-m resolution lidar digital terrain models (DTMs) showing ice wedge degradation in the burn area. (d) The raw difference DTM (dDTM) created by subtracting the 2009 DTM from the 2014 DTM. Panels (e) and (f) show detectable change determined using a Fuzzy Inference System (FIS) propagation of errors threshold of (>~0.2 m) and a FIS 95% probability threshold of (>~0.5 m), respectively.