Remote sensing-based estimation of surface soil salinity around Uyuz Lake using UAV and Sentinel-2 data

Soil salinity is one of the most critical land degradation processes, leading to reduced soil productivity and, in severe cases, complete land abandonment due to salinization. Therefore, early detection and accurate assessment of soil salinity are essential to enable timely remediation using fast, cost-effective, and labor-efficient approaches.In this study, soil salinity in and around Uyuz Lake, located in the Sulusaray district of Tokat Province (Türkiye), was spatially modeled using Unmanned Aerial Vehicle (UAV) multispectral imagery and Sentinel-2B satellite data, both of which are widely used remote sensing systems.The dataset used in this study was based on electrical conductivity (EC) measurements obtained from 39 soil samples collected around Uyuz Lake in 2024. UAV flights equipped with a multispectral camera were conducted to acquire imagery in red, green, blue, near-infrared (NIR), and red-edge bands. In addition, Sentinel-2B imagery with a spatial resolution of 10–20 m, acquired in the same period, was atmospherically and geometrically corrected prior to analysis. Various spectral indices, including NDVI, NDSI, VSSI, and SI, were derived from both datasets and statistically evaluated against EC values using correlation and regression analyses. Subsequently, linear regression models were developed to predict the spatial distribution of EC. According to the analysis results, the Sentinel-2B green band showed the highest correlation with logEC (r = 0.786, p < 0.01), while the UAV red-edge band exhibited the strongest correlation (r = 0.618) among UAV bands. The linear regression model based on the Sentinel-2B green band yielded an R² value of 0.62, indicating that the green band can significantly represent soil salinity variations. Spatial analysis revealed that high EC values (≥ 4 dS m⁻¹) were concentrated mainly within the lake basin and along the shoreline, whereas low EC values dominated the western and northern parts of the study area. This spatial pattern is attributed to high evaporation rates, shallow groundwater levels, and the closed-basin morphology of Uyuz Lake. Overall, the findings demonstrate that integrating UAV and Sentinel-2B data provides an effective approach for monitoring soil salinity dynamics and improving modeling accuracy in small-scale lake basins.