Spatiotemporal analysis of drought in Iran’s Hyrcanian Forests using remote sensing-based vegetation and thermal indices

Introduction and Objective: Drought is one of the most significant global climatic hazards, exerting wide-ranging impacts on the water cycle, agriculture, and ecosystems. The Hyrcanian forests of northern Iran, characterized by high biodiversity and critical roles in carbon storage and water regulation, are particularly vulnerable to drought and climate change. Due to the complexity and high costs of field monitoring, the use of remote sensing data and vegetation and thermal indices provides an effective means to assess drought. This study employs MODIS and Landsat data, integrating multiple vegetation and thermal indices, to analyze the spatiotemporal trends of drought in the Hyrcanian forests during 2000 and 2024, aiming to identify drought-sensitive areas and changes in ecosystem stability.
Material and Methods: Satellite data from Landsat and MODIS, including NDVI (Normalized Difference Vegetation Index), VCI (Vegetation Condition Index), EVI (Enhanced Vegetation Index), SAVI (Soil Adjusted Vegetation Index), NDWI (Normalized Difference Water Index), TVDI (Temperature Vegetation Dryness Index), and Land Surface Temperature (LST), were utilized to assess vegetation health, surface moisture, and the severity of water stress. Data were preprocessed in Google Earth Engine, and 1,678 randomly sampled points with a minimum distance of one kilometer were selected for index extraction. Relationships between LST and vegetation indices were analyzed using Pearson correlation and collinearity analyses to accurately evaluate the impact of environmental stressors on vegetation.
Results: Analysis of vegetation and drought indices in the Hyrcanian forests between 2000 and 2024 indicated a marked decline in vegetation density and health. Annual mean NDVI decreased from 0.642 to 0.498, VCI from 68.42 to 54.37, SAVI from 0.527 to 0.412, and NDWI from 0.214 to -0.067. Concurrently, TVDI increased from 0.412 to 0.536, and maximum LST rose from 47.60°C to 53.08°C, indicating intensified drought and heat stress. Due to strong collinearity among NDVI, SAVI, and NDWI, NDVI and SAVI were removed from the model while NDWI was retained due to its relevance for moisture assessment. Correlation analysis showed that LST had a very strong positive correlation with TVDI in 2000 (r = 0.948), which decreased to a weak correlation in 2024 (r = 0.248). Additionally, LST exhibited significant negative correlations with NDWI (r = -0.076 and r = -0.127) and EVI (r = -0.002 and r = -0.244), reflecting the substantial impact of drought and heat stress on vegetation health and moisture in the Hyrcanian forests.
Conclusion: This study highlights the importance of vegetation and thermal indices for assessing environmental stress and drought in the Hyrcanian forests. Rising LST and intensifying drought stress have significantly affected vegetation health and functionality, emphasizing the need for continuous and comprehensive monitoring. The use of a multi-index approach, incorporating NDWI and TVDI as moisture- and drought-sensitive indicators, allows for more precise drought detection and management in humid forest ecosystems. Special attention to more vulnerable areas, such as the eastern (particularly Golestan Province) and southern sections of the Hyrcanian forests, is essential for conservation and management planning. These approaches can play a key role in developing effective policies to address climate challenges and sustainably protect sensitive forests.