Comparative Analysis of Spectral Indices to Assess Coastal Salinization in Northeastern Nile Delta, Egypt

Soil salinization driven by seawater intrusion poses a growing threat to agriculture in the northeastern Nile Delta, where shallow groundwater tables and intensive irrigation facilitate salt accumulation at the surface. This study mapped surface salinity and its impact on vegetation along a 50 km area from New Mansoura City to Mansoura using a Landsat 8 OLI. Unsupervised K-means clustering identified five land-use/land-cover classes (water, vegetation, barren land, urban, and sabkha), allowing for the computation of eight spectral indices per class: Normalized Difference Vegetation Index (NDVI), Soil-Adjusted Vegetation Index (SAVI), Salinity Indices S1–S4, Normalized Difference Salinity Index (NDSI), and Vegetation Soil Salinity Index (VSSI). Zonal statistics quantified index distributions, and linear regression of index values against distance from the shoreline assessed spatial trends. Vegetation indices (NDVI, SAVI) effectively highlighted canopy stress but showed minimal coastal gradients (β₁ ≈ –1.8 × 10⁻⁶ m⁻¹; R² ≈ 0), reflecting the mitigating effect of local irrigation. Brightness-based indices and NDSI captured salt‐crust signatures, with mean S1–S3 values increasing from agricultural fields (13 222–14 972) to sabkha (19 122–20 080) and positive but low-magnitude coastal slopes (e.g., S1 β₁ = 0.1255 m⁻¹; R² < 0.003). The hybrid VSSI provided the strongest separation between vegetated and salt-stressed zones (mean VSSI: –79 163 vs. –125 610). These results demonstrate differential sensitivity of spectral indices to salinity and vegetation stress, which can be used to monitor soil salinity with minimum cost and apply further analysis and surveys only when necessary.