Journal of Brilliant Engineering (BEN)

Md. Osman Gani Rasel Md. Mahabub Rahman Md. Mahfuj Ahmed Sizan

The current study aimed to quantifying the spatiotemporal variation of land surface temperature (LST) and investigate its association with vegetation greenness, built-up areas, and salinity in Khulna District of Bangladesh between 2002 and 2020 by processing MODIS derived products in Google Earth Engine (GEE) platform. The Mann–Kendall (MK) test and Sen's slope estimator were used to quantify temporal trends while the Pearson correlation analysis was employed to examine relationships among the variables. Results did not show any long term statistically significant monotonic trend at the district scale of the LST (τ= 0.0435, p=0.3304) with an estimated trend of 0.002°C/yr. The monthly mean LST values varied between 20.9°C and 32.3°C and were higher in peri-urban and urban areas in general. The normalized difference vegetation index (NDVI) had a significant positive trend (0.0002 yr-1, with τ=0.1365, p=0.0023), but had a very weak and non-significant relationship with LST (r=0.0455, p=0.4964). The normalized difference built-up index (NDBI) showed a slight decreasing trend (−0.0002 yr⁻¹; τ=−0.1276, p=0.0043), whereas LST was weakly positively correlated with built-up intensity (r=0.16, p=0.016), probably due to spectral mixing and surface reflectance properties rather than actual decreases in urbanization. No significant temporal trend was observed in the spectrally derived salinity proxy but there was a weak positive correlation with LST (r=0.18, p=0.0054) especially near coastlines. However, all these statistically significant relations explained less than 4% of the LST variance (R² < 0.04) and suggested that other climatic and environmental factors have the main influence on the variability in LST.

https://doi.org/10.36937/ben.2026.41138