Geography and Human Relationships

Geography and Human Relationships

Investigating the impact of land use change on water stress and land surface temperature (LST) using the Support Vector Machine (SVM) model and spectral indices NDWI, NDVI, and NDMI (Study area: Astara Estil Wetland)

Document Type : Original Article

Authors
university of zanjan
10.22034/gahr.2025.530179.2510
Abstract
Land surface temperature is of great importance in climate change research. In this study, using the OLI and TIRS sensors of the Landsat 8 satellite, the spectral indices NDWI, NDMI, NDVI and LST were estimated for the Estil Wetland and their relationship with the LST index was calculated. The maximum LST values for the years 2013, 2017, 2021 and 2024 were 15, 16, 17. 21 respectively, and the maximum LST values for the same time period were 13, 22, 14 and 13. The NDMI, NDWI and NDVI indices were also used to estimate the moisture status in relation to water stress, water areas and vegetation cover. The values of these indices were obtained through bands 3, 4, 5, 6 and 10. A Pearson correlation was also established between the above indices and the LST index. Scatter plots drawn with Origin 10 software showed that the highest positive correlation of 0.23 was between LST and NDWI indices, and the highest negative correlation of -0.65 belonged to LST and NDMI indices. The resulting values showed that NDVI and NDMI indices had a negative correlation with LST index. The values obtained from land use change (LULC) were also obtained by the support vector machine (SVM) model. The results showed that vegetation cover increased from 2013 to 2024 and reached from 329 km2 to 487 km2. In contrast, the area of Estil Wetland decreased from 126 km2 to 98 km2 during this period, and the greatest decrease was observed in the eastern part of the wetland.
Keywords

Subjects



Articles in Press, Accepted Manuscript
Available Online from 27 June 2025

  • Receive Date 12 June 2025
  • Accept Date 27 June 2025