1982—2015年黄土高原植被变化特征及归因

Characteristics and driving forces of changes in vegetation coverage on the Loess Plateau,1982—2015

黄土高原植被对于黄河中下游泥沙量减少以及区域生态安全维持发挥着重要作用。气候变化和人类活动是驱动黄土高原植被覆盖度变化的两大因素,然而先前研究多基于统计方法建立植被覆盖度与气候变化的线性关系,忽略了植被覆盖度与气候变化之间复杂的非线性关系,限制了对黄土高原植被变化驱动机制的理解。基于1982—2015年黄土高原植被覆盖度(GIMMS NDVI)数据,以1999年退耕还林工程开始实施为界限,采用随机森林和残差阈值法等手段量化并分析了1982—2015年植被变化特征及气候变化和人类活动对植被变化的影响。结果表明:(1)1982—2015年黄土高原植被覆盖度整体上呈现增长趋势,1999年之后植被覆盖度增加的面积大于1999年之前,1999年以前显著增长的区域只占总面积的54.68%,1999年以后显著增长的区域面积占比增长至85.11%;(2)在考虑时滞效应的基础上,1982—1998年黄土高原超过80%的区域的逐月温度、降水、日照时数与植被覆盖度存在显著的正相关关系,表明1999年以前植被覆盖度变化主要与气候变化有关。随机森林算法能较好地模拟植被覆盖度与气候因子之间的关系,99.71%的区域拟合的决定系数R2达到0.5以上;(3)2000—2015年,气候因素和人类活动对植被覆盖度的影响具有空间异质性。人类活动对植被覆盖度产生积极影响的面积增加了55.94%,气候消极影响的面积减少了42.58%,气候积极影响的面积减少了12.86%。研究提出的量化气候变化与人类活动对植被覆盖度影响的方法能够更好的区分黄土高原植被覆盖度变化的驱动因素,为黄土高原生态建设以及黄河流域的生态安全和高质量发展提供数据支撑。

Vegetation coverage on the Loess Plateau plays an important role in reducing soil sediment in the middle and lower reaches of the Yellow River and in sustaining regional ecological functions.Climate change and human activity are two important factors that drive changes in vegetation coverage on the Loess plateau.Previous researchers used statistical methods to describe the linear relationships between vegetation coverage and climate change on the plateau.However,these studies typically ignored the complicated nonlinear relationships between vegetation coverage and climate change;thus,their understanding of the mechanisms behind changes in vegetation coverage was limited.Therefore,in the present paper,the impacts of climate change and human activities on vegetation changes on the Loess Plateau from 1982 to 2015 were quantified and analyzed using a random forest algorithm and the residual threshold method with GIMMS NDVI data.The results showed the following:(1)Vegetation coverage on the Loess Plateau presented an overall increasing trend from 1982 to 2015.Only 54.68%of the total area showed a significant increase in vegetation before 1999,while this area reached 85.11%after 1999;(2)From 1982 to 1998,the monthly temperature,precipitation,and sunshine hours had a significantly positive correlation with vegetation coverage in more than 80%of the total area of the Loess plateau,indicating that vegetation coverage changes before 1999 were mainly caused by climate change.Using a random forest algorithm,the accuracy of a simulation of 99.71%of the area was above 0.5;(3)From 2000 to 2015,the impacts of climate factors and human activities on vegetation coverage were significantly spatially heterogeneous.The positive driving effect of human activities on vegetation coverage increased by 55.94%,while the negative and positive driving effects of climate-related factors decreased by 42.58%and 12.86%,respectively.The driving factors of changes in vegetation cover on the Loess Plateau can be best described by the method proposed in this study,which quantifies the impact of climate change and human activities on vegetation coverage and provides data support for ecological construction on the Loess Plateau.