Introduction
Water stress is influenced by the balance between water availability and human water demand, which vary across space and time. Although existing water-stress assessments commonly integrate multiple climatic, hydrological, land-use, and socioeconomic variables, such approaches may require extensive datasets and may be difficult to apply at the local or aquifer scale. Therefore, developing simple indices that capture both natural water availability and human-induced pressures can provide a practical approach for spatial and temporal assessment. This study develops a Water Stress Index (WSI) based on precipitation and land use to assess spatiotemporal variations in water stress in the Gorgan Aquifer, northern Iran.
Methods
Annual precipitation and land-use data from 2001 to 2022 were used. Daily precipitation data from in-situ rain-gauge stations obtained from the Golestan Regional Water Company were aggregated annually and spatially interpolated using the Inverse Distance Weighted method. Annual land-use/land-cover data were obtained from the MODIS MCD12Q1 product at 500 m spatial resolution. Cropland and urban areas were assigned weights of 1.0 and 0.5, respectively. Normalized precipitation and the weighted land-use component were integrated using a normalized difference formulation: WSI = (P − LU)/(P + LU).
Results
The results indicate an overall intensification of water stress across the aquifer during 2001–2022. Periods of reduced water stress generally coincided with relatively high precipitation and, in some years, flood occurrence, suggesting a potential contribution of increased precipitation and flood-related recharge to temporary alleviation of water stress. The spatial and temporal variations in WSI were further evaluated against annual groundwater-level changes. For validation, the pixel-wise correlation between WSI and groundwater-level change was classified using a threshold of −0.2, with values below −0.2 assigned 0 and values equal to or greater than −0.2 assigned 1. ROC analysis using the mean WSI of each pixel as the continuous predictor yielded an AUC of 0.80, indicating good discriminatory ability.
Conclusions
The developed WSI provides a simple and flexible approach for assessing water stress across spatial and temporal scales using limited climatic and land-use data.