近30年云贵高原水蚀荒漠化遥感监测及影响因子时空分异规律

云贵高原属于典型季节性干旱生态脆弱区，其中因流水造成的水蚀荒漠化已成为限制该区域持续发展的重大生态环境问题之一。本文采用长时间序列Landsat卫星数据监测云贵高原核心区云贵两省近30年水蚀荒漠化的现状与时空变化，并分析该区域荒漠化关键影响因子的时空分异规律，提出一种多指标规则分类算法用于云贵高原水蚀荒漠化遥感监测，并选用植被面积、GDP及降水3个关键影响因子，采用地理加权回归模型，分析该区域水蚀荒漠化影响因子的空间分异规律。试验结果表明：①云贵高原荒漠化的主要分布区为滇桂黔区、川滇交界地区、滇西边境区及蒙自地区。②近30年研究区水蚀荒漠化变化趋势可分为1989—2010年、2010—2013年及2013—2016年3个阶段，1989—2010年该区域的荒漠化呈现波动增长状态；2010—2013年荒漠化面积快速扩展；2013—2016年荒漠化快速扩展的趋势得到了有效遏制。③3个关键影响因子与水蚀荒漠化关系呈现数值大小不同、正负效应和因子组合效应的空间分布差异，局部地区的经济发展和气候变化加剧了其水蚀荒漠化。本文研究成果将为云贵高原水蚀荒漠化的精细化防控和治理提供一定的理论依据和技术支撑。

Remote sensing monitoring of water erosion desertification and spatio-temporal differentiation of its influencing factors in the Yunnan-Guizhou Plateau in recent 30 years

The Yunnan-Guizhou Plateau (YGP) is a typical seasonal arid ecological fragile area, and water erosion desertification (WED) that is caused by water flow has become one of the important ecological and environmental problems in the region. In this paper, long-term Landsat datasets are used to monitor the temporal and spatial variation trend of WED for the past 30 years in the YGP, and the spatial and temporal differentiation of its influencing factors is analyzed. A multi-index rule classification algorithm is proposed for remote sensing monitoring of WED in the YGP, and three key influencing factors (such as vegetation area, GDP and precipitation) are selected to analyze the spatio-temporal differentiation of the WED influencing factors using geographic weighted regression (GWR) model in the YGP. The experimental results are shown as following:① the main area of WED in YGP can be divided into Dian-Gui-Qian area, Sichuan-Yunnan border area, Dianxi area and Mengzi region. ② the WED expansion in the YGP can be divided into three stages as follows:a slow increase from 1989 to 2010, rapid expansion from 2010 to 2013, and a slight increase from 2013 to 2016. ③ The relationship between the three key infulencing factors and water-erosion desertification presents different spatial size in different numbers, positive and negative effects and the spatial distribution difference of factor combination effects. The local economic development and climate change exacerbate water-erosion desertification in local areas. The research results will provide theoretical basis and support for the fine prevention, control and treatment of water erosion and desertification in the YGP.