Quantification Of Permafrost Melting Risk Using Fractal Analysis Of Remote Sensed Images Using Insights From The Anaktavuk River Fire Of 2007

An estimated 1.5 trillion tons of methane (CH4) and CO2 are trapped underneath permafrost. Effective large-scale quantification of permafrost degradation is critical to characterizing permafrost melt risk. Thawing permafrost results in thermokarsts that contribute to surface roughness. This study investigated the potential of using fractal analysis of landsat images to quantify surface roughness of the image, to rapidly characterize and identify permafrost regions undergoing critical melting. Data and remote sensed images from the Anaktavuk river tundra fire of 2007 was used to establish baselines The study evaluated fractal dimensions of annual Landsat pictures of permafrost regions (n=62), and performed linear regression against the region's average annual thaw temperature. The results showed that fractal dimension correlates inversely with average thaw temperature (p-value = 0.006). The results reveal that surface roughness derived from remote sensed image analysis is a potential indicator of permafrost degradation. This approach presents the prospect to perform large scale permafrost monitoring.