| This thesis has two primary objectives. The first objective is to investigate effects of land use and land cover on spatial patterns and temporal trends of the land surface temperature in the vicinity of the Norilsk Industrial Region in the Russian Arctic. A combination of weather station data, MODIS land surface temperature product, land cover classification and DEM was used to analyze the land cover and land use specific surface temperatures over the 2002-2011 period. The results indicate a strong positive relation between the extent of build-up area and the increase in land surface temperature suggesting a strong heat island effect. The spatial average magnitude of the heat island, estimated as an average difference between built-up and non built-up LST was found to be 3.1°C in winter and 1.0 °C in summer. Among natural land cover categories, dominance of barren and tundra contribute to land surface warming while presence of dense closed forest has a cooling effect to LST. The second objective was to reconstruct LST back to 1980 and extend to the whole Yenisei River Region using combination of spatial datasets. The reconstruction combines MODIS LST products, Modern Era Retrospective-analysis for Research and Applications (MERRA) near surface air temperature, multi-decadal land cover maps, and Digital Elevation Model (DEM). Validation of the reconstructed LST (rLST) indicates that the model is able to explain over 60% of total variance. Comparison of the reconstructed LST with the air temperature from five meteorological stations within the study area demonstrates a high correlation (R>0.93) between inter-annual variability of modeled rLST and observed air temperature. |
