Urban thermal micro-mapping using satellite imagery and ground-truth measurements: Kyiv city area case study

• Дата: 2019
• Автори: Piestova, Iryna, Lubskyi, Mykola, Svideniuk, Mykhailo, Golubov, Stanislav, Laptiev, Oleksandr
• Формат: Стаття
• Мова: English
• Опубліковано: Scientific Centre for Aerospace Research of the Earth Institute of Geological Science National Academy of Sciences of Ukraine, Kyiv, Ukraine 2019
• Теми: city thermal regime; land surface temperature; thermal radiation emissivity; time series; городской тепловой режим; температура земной поверхности; коэффициент теплового излучения; временные ряды; тепловий режим міста; температура земної поверхні; коефіцієнт теплового випромінювання; часові ряди
• Онлайн доступ: https://ujrs.org.ua/ujrs/article/view/149
• Назва журналу: Ukrainian Journal of Remote Sensing of the Earth

The aim of this research is to enhance approaches existing for the assessment of cities thermal conditions under climate change impact by using multispectral satellite data for Kyiv city area. This paper describes the method and results of the Earth’s surface temperature (LST) and thermal emissivity calculation. Particularly, the thermal distribution was estimated based on spectral densities according to Planck’s law for “grey bodies” by using the Landsat-8 TIRS and Sentinel-2 MSI satellite imagery. Furthermore, the result was calibrated by ground data collected during the ground-truth measurements of the typical city surfaces temperature and thermal emissivity. The spatial resolution of the LST images obtained was enhanced by using the approach of subpixel processing, that is the pairs of invariant images shifted with subpixel accuracy. As a result, such an approach allowed to enhance the spatial resolution of the image up 46%, which is much higher than the potential performance of the thermal imaging sensors existing. The interrelation between the Earth’s surface type and the temperature was revealed by the results of the Sentinel-2A MSI image of 21 August 2017 supervised classification. Thus, the image was divided into the six major classes of the urban environment: building’s rooftops, roads surface, bare soil, grass, wood, and water. As a result, surfaces with vegetation much more cool next to artificial ones. The time-series analysis of 18 thermal images (Landsat TM and Landsat-8 TIRS) of Kyiv for the period from 6 Jun 1985 till 1 June 2018 was done for spatiotemporal changes investigation. Therefore, the sites of the LST thermal anomalies caused by landscape changes were developed. Among them are the sites of increased LST where thw “Olimpiyskiy” national sport center and adjacent parking was built and the site of decreased LST where the tram depot was liquidated and the territory was flooded.