Abstract: In order to reveal the dynamic changes of vegetation and its response to natural and anthropogenic factors in Yunnan Province，this study wasbased on the dataset of normalized difference vegetation index (NDVI),natural factors and anthropogenic factors from 2000 to 2020. We analyzed the spatiotemporal variation characteristics of NDVI in Yunnan Province by using Theil-Sen trend analysis and Mann-Kendall test,and explored the driving forces of NDVI spatial heterogeneity in different periods with Geographical detectors,and to explore the influence of the dominant factors on the changes of NDVI. The results showed that vegetation cover in Yunnan Province had a spatial distribution pattern as higher in the southwest and lower in the northeast. In terms of temporal variation,the proportion of regions with increasing NDVI from 2000 to 2020 was 83.15%,while the proportion of regions with decreasing NDVI 16.85%. The main influencing factors of spatial heterogeneity in NDVI were the land use type,precipitation and relative humidity,among which the land use type had the highest explanatory power,significantly higher than other factors. The transformation of land use type had a significant effect on the NDVI changes. The implementation of forestry ecological projects increased the area of forest lands and improved the regional vegetation cover,which played a positive role in the increase of NDVI;while the expansion of settlement land destroyed the regional vegetation,resulting in the decrease of NDVI. This study would provide a scientific basis for the development of future ecological planning strategies in Yunnan Province.

Key words: Remote sensing, Vegetation, Normalized difference vegetation index, Geographical detectors, Land use type, Anthropogenic activities

摘要： 为揭示云南省植被动态变化及其对自然因素和人类活动因素的响应,本研究基于2000—2020年归一化差异植被指数(Normalized difference vegetation index,NDVI)、自然因子和人类活动因子数据集，采用Theil-Sen趋势分析法和Mann-Kendall检验分析了云南省NDVI时空变化特征，结合地理探测器探究了不同时期NDVI空间分异的驱动力，并探讨了主导因子对NDVI变化的影响。结果表明：空间分布上，云南省植被覆盖呈现西南高，东北低的空间分布格局。时间变化上，2000—2020年NDVI上升的区域面积占比为83.15%，而NDVI下降的区域面积占比为16.85%。NDVI空间分异的主要影响因子为土地利用类型、降雨量和相对湿度，其中土地利用类型解释力最高，远高于其他因子。土地利用类型的转变对NDVI变化具有显著影响，林业生态工程的实施增加了林地面积，改善了区域植被，对NDVI上升起到了积极作用；而建筑用地扩张，破坏了区域植被，导致NDVI降低。本研究为云南省未来生态规划策略的制定提供了科学依据。

关键词: 遥感, 植被, 归一化差异植被指数, 地理探测器, 土地利用类型, 人类活动