Analyzing the Drivers of Spatio‐Temporal Riparian Vegetation Variability in a Dam‐Reservoir‐River System From Remotely Sensed Data

ABSTRACT Dams have a considerable impact on river systems and lead to changes in riparian vegetation. The Normalized Difference Vegetation Index (NDVI) is widely used for researching ground surface processes and analyzing and monitoring vegetation dynamics. To date, few studies have comprehensively quantified the impact of cascade dams on riparian vegetation NDVI in dam‐reservoir‐river systems. We analyzed the spatial distribution of the mean annual maximum NDVI value (NDVImax) of the riparian areas of the middle‐lower reaches of Hanjiang River (MLHan) from 1986 to 2018 using 30‐m resolution Landsat series data. We explored the spatio‐temporal trends of NDVImax using the Mann‐Kendall test and Sen's slope estimator. Additionally, we quantified the individual and combined influences of natural factors and human activities on NDVImax changes using Geodetector. The main findings and conclusions are as follows: (1) the NDVImax in the riparian zone of the MLHan showed an increasing trend, with a rate of 0.005–0.01 per year from 1986 to 2018; (2) water occurrence was the main driving factor of NDVI change in the study area, contributing over 60% to NDVImax variability, followed by land use/land cover with a 49% contribution. The interaction between land use/land cover and water occurrence accounted for an 80% contribution rate; (3) the section from Cihe Town to the Cuijiaying Water Conservancy Project and the Yuekou‐Wuhan Reach showed a decreasing trend in NDVImax from 1986 to 2018 that likely resulted from the expansion of construction land. Water occurrence had the greatest influence on riparian NDVI change in the study area. This study provides a practical reference for investigating the effects of water occurrence on riparian vegetation.