Abstract Precipitation is scarce in arid regions, making water redistribution a key driver of vegetation patterns, especially for non‐zonal vegetation in oases. Nevertheless, the coupling and mutual feedback mechanism between water and vegetation distribution is complex, and it has not been fully addressed. Taking an arid region of Northwest China as a case study, this study extracted vegetation distribution using multiple methods including field surveys, drones, remote sensing, and other multi‐source data fusion supervised classification. Following that the study quantitatively analyzed the impact of water conditions on non‐zonal vegetation distribution, and identified spatial patterns characterized by water conditions in arid areas with strip vegetation, riparian, and patchy terminal desert zones, and established quantitative formulas relating the NDVI of each pattern to key water and morphology variables. It is found that across arid dune, riparian, and terminal zones, heterogenous vegetation patterns are driven primarily by various environmental controls on water availability, shaped by precipitation, topography, river discharge, and groundwater at each of the typical landscapes of an arid area. In particular, for the riparian zones of the case study area, a threshold of 2,000 m from the river channel provides a critical inflection of riparian vegetation; in terminal areas depth to groundwater (DTG) is the dominant factor—shallow DTG (<5 m) produces clustered vegetation; whereas deep DTG (≥15 m) yields largely random distributions. These findings clarify the mechanisms driving vegetation heterogeneity in arid environments and provide scientific support for ecological restoration and greening in arid regions.
Building similarity graph...
Analyzing shared references across papers
Loading...
Xing Li
Yong Wang
Yong Zhao
Water Resources Research
China Institute of Water Resources and Hydropower Research
Changjiang Water Resources Commission
PowerChina (China)
Building similarity graph...
Analyzing shared references across papers
Loading...
Li et al. (Fri,) studied this question.
www.synapsesocial.com/papers/69fd7e90bfa21ec5bbf06c0e — DOI: https://doi.org/10.1029/2025wr042090