Comparison between In Situ and Satellite Multiscale Temperature Data for Russian Arctic Cities for Winter Conditions

This study compares three approaches to microclimate research by the example of Russian Arctic cities in winter conditions: (1) using high-resolution thermal images by the Landsat 8 satellite, (2) using low-resolution images of the MODIS imaging system, and (3) using direct measurements of the surface air temperature. The latter involves observations of automatic weather stations and temperature sensors of the Urban Heat Island Arctic Research Campaign (UHIARC) network and Roshydromet weather stations. Two methods for calculating the land surface temperature (LST) from Landsat 8 satellite images have been considered: the first method is based on the atmospheric correction of images using the MODTRAN radiation transfer model and tabulated emissivity values for different land cover types, and the second method uses no atmospheric correction. The study was performed for the cities of Apatity, Vorkuta, Salekhard, Nadym, and Novy Urengoy. The land surface temperatures calculated from Landsat 8 images without atmospheric correction have been shown to agree with MODIS data and observations better than the results obtained with atmospheric correction. This indicates an inaccuracy in the value of surface emissivity. For a number of cases, the spatial variation patterns of the land surface and air temperatures are closely related; here, both types of data are indicative of the effect of urban heat island with urban-rural temperature differences up to 4 degrees C in the daytime. These results are fundamentally different from those obtained previously for lower latitudes, which indicates the prospects of using high-resolution satellite temperature data for mapping and further studies of the microclimate of Arctic cities in winter conditions.