Remote-sensing based assessment of post-fire changes in land surface temperature in Arctic-Boreal permafrost regions
In recent years, wildfires became more predominant in northern high latitude permafrost regions. Arctic warming, as a consequence of climate change, causes drying of vegetation being more flammable and promotes lightning incidents. Hence, the Arctic wildfire season extents accompanied by an increase in wildfire frequency as well as burn severity (BS). By now, boreal forests are known as carbon sink but will become a carbon source, further enhancing climate change. Within loss in surface organic layer due to wildfires, the thermal conductivity of the soils changes, impacting the underlying permafrost. Thawing permafrost again releases greenhouse gasses, resulting in a positive feedback, further accelerating climate warming. Regarding these global consequences of wildfires, a better understanding of small regional scale processes is necessary for reliable future predictions. Therefore, the aim of this study is to assess post-fire impacts on permafrost in the north-eastern Siberian Yana river catchment using remote sensing data. As previous studies announced a future spread of wildfires northward from Taiga to Tundra ecosystems, both will be considered in the study analysis to distinguish between their influence quantity. In order to answer the research question, the effects on permafrost after wildfire were investigated using 9 Siberian fire sites including their accompanied control sites, along the Yana river. The yearly mean land surface temperature (LST), calculated from Landsat images over a time period from 2006- 2020 for the summer months (June, July, August) serves therefore as data basis. Based on that, the Permafrost_CCI products including the yearly mean ground surface temperature (GST) and active layer thickness (ALT) between 1997-2018, were consulted for comparison purposes. Created time series of LST, GST and ALT were individually analyzed by visual interpretation, descriptive statistics and trend analysis. Finally, GST and ALT time series were correlated against LST time series. Additionally, the normalized burn ratio (NBR) was calculated from Landsat images to get supportive information about the BS and vegetation recovery, as these factors play a very important role in influencing the magnitude of permafrost variations due to wildfires. The main findings show a correlation between LST and ALT resulting in a decrease of permafrost as the ALT increases within increasing LST after a wildfire. The coherence between LST and GST does not show unique results though, but result in increasingly warmer temperatures in the soil as well. Regarding differences between Taiga and Tundra ecosystems, impacts are causing a greater threat for permafrost in Tundra regions, especially in context with future predicted increase of wildfire frequency and BS. Nevertheless, studying permafrost remains still challenging due to the remoteness of the study area, resulting in a lack of in-situ data, as well as remote sensing data.