Background and Objectives: Drought is one of the most significant climatic hazards affecting water resources, vegetation cover, and agricultural production, particularly in arid and semi-arid regions. The Karkheh River Basin, as one of Iran’s most important agricultural basins, plays a crucial role in food security and regional socio-economic stability. Due to its large area and pronounced climatic variability, this basin is highly vulnerable to drought events. Conventional drought monitoring methods based on ground observations are often limited by sparse station networks and insufficient spatial coverage. Therefore, remote sensing-based approaches and composite vegetation indices provide an effective alternative for spatiotemporal drought monitoring. The main objective of this research is to monitor spatial and temporal drought in the Karkheh basin using the NDDI composite index and evaluate its capability as a diagnostic tool for agricultural drought monitoring and providing early warnings. This approach, by providing a quantitative framework for analyzing drought processes, can contribute to improved water resources management and development of climate change adaptation strategies in the country.
Materials and Methods: MODIS sensor images with a spatial resolution of 500 m acquired in May (peak vegetation growth period) were used to calculate NDVI, NDWI, and subsequently the NDDI. Drought conditions were classified into five categories: no drought, mild, moderate, severe, and extreme drought. Ground-based precipitation data from 11 selected meteorological stations were quality-controlled and homogenized prior to analysis. SPI values were calculated at different time scales to validate the satellite-derived NDDI. The accuracy of NDDI was assessed using statistical error metrics, including Root Mean Square Error (RMSE) and Mean Absolute Error (MAE).
Results: The results demonstrated that NDDI effectively captured the spatial and temporal patterns of agricultural drought in the Karkheh River Basin. Approximately 85.3%, 89.9%, and 74.8% of the basin area experienced varying degrees of drought in 2007, 2008, and 2014, respectively. The most severe drought conditions were observed in 2008, when more than 78% of the basin was classified as moderate to extreme drought. Validation results indicated a strong agreement between NDDI and SPI, particularly at the one-month time scale, with minimum RMSE and MAE values of 0.11 and 0.12, respectively.

Conclusion: Given the importance of Karkheh watershed in terms of climatic, social, and economic conditions, accurate monitoring and assessment of drought in this basin is very important. In this regard, the present study used the NDDI composite index to monitor the spatial and temporal extent of agricultural drought during the 2001-2021 statistical period. The results showed that about 85.3%, 89.9% and 74.8% of the basin area were affected by drought in 2007, 2008 and 2014, respectively. Also, the results of the accuracy assessment criteria including RMSE and MAE indicate the high efficiency of these indicators in the region. Beyond drought mapping, this study emphasizes the role of NDDI index as a diagnostic and early warning tool to investigate the interaction of vegetation and water and its impact on agricultural resilience. The combined use of NDVI and NDWI indices and creation of NDDI allow tracking ecosystem responses to moisture deficiency at different temporal and spatial scales. However, limitations such as land use changes, vegetation density, and seasonal conditions can affect the accuracy of the index. Therefore, combining NDDI with ground data and climate indicators can increase the accuracy of drought monitoring and help in better decision-making