The Bahr Al-Najaf depression is a unique geological and ecological feature in Iraq that has undergone significant environmental transformations over the past two decades. This study quantifies the spatiotemporal dynamics of Land Use and Land Cover (LULC) changes in the depression from 2002 to 2025 using a multi-sensor remote sensing approach based on Landsat and High-Resolution Google Earth imagery. The Random Forest (RF) machine learning algorithm was implemented to classify the region into five distinct classes: Deep Water, Shallow Water, Dry Soil, Wet Soil, and Built-up Areas. The classification model, optimized with 500 decision trees and validated using a stratified random sampling of 250 ground-truth points, achieved a high overall accuracy of 92.1%. The results reveal a dramatic and irreversible urban expansion, with built-up areas increasing by 192%, growing from 29.26 km 2 (6.71%) in 2002 to 85.46 km 2 (19.61%) in 2025, predominantly encroaching on the stable eastern plateau. In contrast, hydrological features exhibited severe instability; total water surface area plummeted from 155.70 km 2 (35.73%) in 2002 to a critical low of 47.40 km 2 (10.87%) during the 2015 drought peak. While a partial recovery was observed in subsequent years – aligning with NASA’s regional observations – the water extent remained at 74.12 km 2 by 2025, failing to reach its initial baseline. Analysis of the Digital Terrain Map (DTM) confirms that topographic constraints play a deterministic role in directing urban growth, while the depression’s low-lying center remains vulnerable to hydrological shocks. These findings highlight a critical shift toward an anthropogenically modified landscape, demonstrating a robust remote sensing framework that is directly transferable to estuarine, lagoonal, and deltaic wetlands. Ultimately, this study provides critical insights for the integrated management of fragile hydrological systems facing similar anthropogenic and climate-induced stresses globally.
Helaly et al. (2026) studied this question.