Impacts of Flooding on Vegetation: A Case Study of the 2025 Xinglong Mountain Flood

This dataset was developed to investigate the topographic–hydrodynamic controls on vegetation responses to mountain flood disturbances in arid and semi-arid environments. The underlying research hypothesis is that terrain conditions modulate hydrological processes during flood events, which in turn drive spatial heterogeneity in vegetation dynamics. Specifically, terrain-derived hydrological indices (e.g., flow accumulation, topographic wetness) are expected to influence vegetation resistance and recovery patterns following flood disturbances.
The dataset integrates multi-source geospatial data, including satellite-derived vegetation indices, digital elevation model (DEM)-derived topographic variables, and land cover information. Vegetation conditions were primarily characterized using the Normalized Difference Vegetation Index (NDVI), derived from Sentinel-2 imagery. Topographic and hydrological variables (e.g., elevation, slope, aspect, topographic wetness index) were extracted from DEM data. Land cover data were used to classify surface types and assist in stratified analysis.
All data were preprocessed following standard procedures, including atmospheric correction, cloud masking, geometric correction, and spatial resampling to a consistent resolution. Terrain variables were calculated using GIS-based spatial analysis methods. The study area was further divided into terrain zones using a classification approach (e.g., Jenks natural breaks), and all raster pixels were assigned corresponding zone labels to facilitate statistical comparison.
The dataset reveals clear spatial differentiation in vegetation response patterns under varying terrain conditions. Areas characterized by higher moisture accumulation potential (e.g., valley bottoms and concave slopes) tend to exhibit stronger vegetation recovery, whereas steep slopes and well-drained areas show weaker or delayed responses. These findings support the hypothesis that terrain-driven hydrological processes play a critical role in regulating vegetation dynamics in flood-affected arid mountain regions.
Users of this dataset should interpret the variables in a spatially explicit context. Each raster layer represents a specific environmental factor, and pixel values correspond to measured or derived quantities at a given spatial resolution. The terrain zone classification layer can be used as a categorical variable for comparative or statistical analysis. The dataset is suitable for applications such as ecological modeling, hazard assessment, vegetation resilience analysis, and machine learning-based environmental prediction.
To ensure reproducibility and correct usage, users are advised to consider the spatial resolution, temporal coverage, and preprocessing steps applied. The dataset can be directly used in GIS or remote sensing software and supports further analysis such as regression modeling, classification, or spatial statistics.

本数据集旨在探究干旱半干旱环境中，山地洪水扰动下植被响应的地形-水动力控制机制。本研究的核心假说为：地形条件可调控洪水事件期间的水文过程，进而驱动植被动态的空间异质性。具体而言，地形衍生的水文指数（如流量累积、地形湿度）预计会影响洪水扰动后的植被抗性与恢复格局。
本数据集整合了多源地理空间数据，包括卫星衍生植被指数、数字高程模型（Digital Elevation Model，DEM）衍生的地形变量以及土地覆盖信息。植被状况主要通过归一化植被指数（Normalized Difference Vegetation Index，NDVI）表征，该指数由Sentinel-2遥感影像提取得到。地形与水文变量（如高程、坡度、坡向、地形湿度指数）均从DEM数据中提取。土地覆盖数据用于地表类型分类，并辅助开展分层分析。
所有数据均按照标准流程进行预处理，包括大气校正、云掩膜、几何校正以及空间重采样以达成统一分辨率。地形变量通过基于地理信息系统（Geographic Information System，GIS）的空间分析方法计算得到。研究区域进一步通过分类方法（如詹克斯自然间断点分级法）划分为多个地形分区，并为所有栅格像元分配对应的分区标签，以方便开展统计对比分析。
本数据集清晰揭示了不同地形条件下植被响应格局的空间分异特征。水分累积潜力较高的区域（如谷底与凹坡）通常表现出更强的植被恢复能力，而陡坡与排水良好的区域则呈现出较弱或延迟的植被响应。上述研究结果验证了本研究假说：在受洪水影响的干旱山地环境中，地形驱动的水文过程在调控植被动态方面发挥着关键作用。
本数据集的使用者应在空间显式的研究框架下对各变量进行解读。每一个栅格图层均代表一个特定的环境因子，像元值对应特定空间分辨率下的实测或衍生参数。地形分区分类图层可作为分类变量用于对比分析或统计研究。本数据集适用于生态建模、灾害评估、植被恢复力分析以及基于机器学习的环境预测等应用场景。
为确保研究可重复性与正确使用本数据集，使用者应留意其空间分辨率、时间覆盖范围以及所采用的预处理流程。本数据集可直接在GIS或遥感软件中使用，并支持回归建模、分类或空间统计等后续分析。