Assessing the drivers of land use change: Drought, conflict, and governance in the Syrian Upper Orontes watershed (1991–2024)

This study analyzes land use and land cover (LULC) dynamics within the Al-Qusayr watershed, located in the Syrian component of the Upper Orontes Basin. Five Landsat datasets for 1991, 2000, 2009, 2017, and 2024 were analyzed to capture key pre-conflict, conflict, and recovery phases, thereby distinguishing irrigated from rainfed agriculture in this vulnerable semi-arid karst watershed. We applied supervised Support Vector Machine (SVM) classification and a multi-step post-classification correction to refine the results. The accuracy assessment for each period yielded overall accuracies between 91.4% and 94.2%. Five main LULC classes were identified: Agricultural Land (including Farmland and Rainfed/Winter Season), Water (including Water bodies and Wetland), Urban, Unused Land/Natural Vegetation, and Dry Stream. Over the study period, significant shifts occurred. Total agricultural land expanded by 24.6% and urban areas expanded by 74.6%. This growth corresponded with a 53.4% reduction in unused land. A significant conversion from irrigated to rainfed agriculture was observed during the conflict followed by a reversal of this trend after 2017. To provide quantitative context for driver attribution, we integrated CHIRPS precipitation to derive a standardized precipitation anomaly (z-score; 1991–2024) and VIIRS nighttime-light radiance (2014–2024) as an indirect proxy for disruption and recovery conditions, and related these indicators to irrigated and rainfed agricultural changes. Irrigated farmland contraction during the conflict period coincided with negative precipitation z-scores (drought) and reduced nighttime lights, whereas the recovery of irrigated agriculture after 2017 was accompanied by rising nighttime lights despite continued dry conditions, suggesting that improved stability may have contributed to agricultural reactivation. Overall, the Al-Qusayr watershed experienced substantial LULC fluctuations associated with the combined pressures of drought, policy shifts, and conflict. However, the landscape demonstrated notable adaptive capacity, particularly through agricultural resilience. These findings provide a useful baseline for decision-makers targeting sustainable land rehabilitation and resource management.