Spatio-temporal sensitivity of MODIS land surface temperature anomalies indicates high potential for large-scale land cover change detection in Arctic permafrost landscapes

摘要: Abstract The accelerated warming of the Arctic climate may alter local and regional surface energy balances, for which changing land temperatures (LSTs) are a key indicator. Modeling current anticipated changes in balance requires an understanding spatio-temporal interactions between LSTs cover, both can be monitored globally by measurements from space. This paper investigates accuracy MODIS LST/Emissivity Daily L3 Global 1 km V005 product its sensitivity to properties Canadian High permafrost landscape. cover ranged fully vegetated wet sedge tundra barren rock. were compared with situ radiometer areas collected over 2-year period July 2008 2010 including summer winter conditions. was − 1.1 °C root mean square error 3.9 °C entire observation period. Agreement lowest during freeze-back periods where LST showed cold bias likely due overrepresentation clear-sky A multi-year analysis spatial anomalies, i.e., difference mean, revealed robust pattern. Highest variability anomalies found freeze-up thaw as well open water early presence or absence snow ice. anomaly pattern similar all three years despite strong differences precipitation, air temperature net radiation. Summer above 5.0 °C greatest diversity four distinct 2.0 °C classes. − 4.5 °C 2.6 °C average standard deviation 1.8 °C. Dry ridge heated up most, while wetland dry sparsely bedrock high albedo remained coolest. observed used baseline against evaluate past future that relate balance. classes mainly reflected combination wetness. potential use this tool monitor drying wetting should therefore further explored. multi-sensor approach combining thermal satellite optical radar imagery promises effective dynamic, process-based ecosystem monitoring scheme.