Spatiotemporal analysis of surface temperature dynamics in the Supii River basin using regression methods

• Дата: 2025
• Автори: Marhes, Serhii, Filipovych, Volodymyr, Lubskyi, Mykola
• Формат: Стаття
• Мова: English
• Опубліковано: S. Subbotin Institute of Geophysics of the NAS of Ukraine 2025
• Теми: linear regression; time series; land surface temperature; annual increment; global warming
• Онлайн доступ: https://journals.uran.ua/geofizicheskiy/article/view/333289
• Назва журналу: Geofizicheskiy Zhurnal

The research presents an approach for detecting spatial and temporal changes, specifically the annual increase in land surface temperature. The study area encompasses the Supii River basin, which spans the Chernihiv, Kyiv, and Cherkasy oblasts and flows into the Dnipro. This region is characterized by intensive agriculture, poor aquifer recharge, low-quality groundwater, and prolonged droughts. Temperature data for July and August were obtained from publicly available Landsat mission archives for the period 1984—2024. It is recommended to recalculate Landsat thermal images based on emissivity, since their Level 2 product may contain pixels with missing information. To ensure the highest accuracy, pixels affected by clouds and their shadows were masked; the next stepinvolved time series analysis of the filtered images. The time series analysis aimed to identify key patterns in the evolution of temperature dynamics, functionally dependent on various influencing factors. Accurate spatial alignment of the imagery enabled consistent undistorted calculation of the ground’s physical characteristics over the entire study area and for each year of observation. A simple linear regression was applied to each pixel in each raster image. To visualize the spatial distribution of long-term temperature dynamics, the regression gradient (or slope coefficient) was used, representing the average annual increase in temperature. The results are presented as spatial indices of annual surface temperature growth within the Supii River basin, highlighting settlements with dominant positive trends, and the distribution by land use cover. This provides insight into where and to what extent climate conditions may become critical in the coming years, assuming the temperature continues to rise. As a mitigation measure, it is proposed to vegetate the urban and rural areas (particularly in larger local communities) to help reduce the impact of global warming.