Thermal remote sensing for mapping the sub-Arctic permafrost and refining its southern limits

Permafrost soils store more than one fourth of global soil organic carbon. This important carbon pool is threatened by carbon release from permafrost thawing. Especially in sub-Arctic transition zones, accurate mapping of permafrost is crucial for timely documentation of permafrost retreat and the consequent carbon uptake in its fragile ecosystems. Utilizing the 8-day Aqua/MODIS land surface temperature (LST) products in 2003-2020, this study extracts the southern limits of a sub-Arctic boreal mountain permafrost in the Amur River Basin bordering China, Russa, and Mongolia. A temporally dynamic MODIS Look-up-table is built to scale the MODIS LST to daily ground surface temperature. Using a thermal equilibrium model, the mean annual ground temperature (MAGT) on top of permafrost ground is estimated. The 18-year average MAGT = 0 degrees C defines the southern limits of the permafrost. Integrating the spatial distributions and temporal frequencies of annual MAGT < 0 <degrees>C, a probability index of permafrost continuity is extracted to classify permafrost types in a km-scale. The extracted permafrost map shows a similar pattern of permafrost distributions as the currently adopted global permafrost map from the International Permafrost Association (IPA). More importantly, it reveals an apparent northward shift of the southern limits. This study showcases the advantages of thermal remote sensing in exploring permafrost retreat, which provides the base information for modeling carbon release in the vulnerable transitional zone. The spatially and temporally continuous satellite observations could improve our carbon monitoring in cold lands under the pressure of global warming.