1990—2020年黑龙江省植被覆盖度的时空变化趋势及驱动力

Spatial and temporal trends and drivers of fractional vegetation cover in Heilongjiang Province,China during 1990-2020

植被覆盖度(FVC)作为表征植被生长状况与生态系统变化的定量指标,其时空变化趋势及驱动力是全球及区域生态环境的重要研究内容。本研究基于GEE云计算平台,应用像元二分模型对黑龙江省1990—2020年FVC进行估测,并通过Mann-Kendall突变检验、Sen趋势分析、Mann-Kendall显著性检验、相关性分析及结构方程模型,分析FVC的时空变化趋势及驱动力。结果表明:基于像元二分模型法计算研究区FVC具有较高精度(R^(2)>0.7、均方根误差<0.1、相对均方根误差<14%)。1990—2020年,黑龙江省年均FVC为0.79,呈波动上升趋势(0.72~0.85),年均增长0.4%,市级行政区年均FVC也呈不同程度的增长。黑龙江省以极高FVC类型为主,其面积占比逐渐增加。FVC呈增加趋势的面积占比为67.4%,呈减小趋势的面积占比为26.2%,其余保持不变。人为活动因子对年均FVC的相关性高于生长季月平均气象因子对年均FVC的相关性。人为活动因子是黑龙江省FVC变化的主要驱动因子,其次为土地利用类型,而生长季月平均气象因子对FVC变化的总影响为负。本研究结果可为黑龙江省长时间植被覆盖监测与驱动力分析提供技术支撑,也为生态环境的恢复与保护、相关土地利用政策的制定提供了依据。

Fractional vegetation cover(FVC)is a quantitative indicator for vegetation growth conditions and eco⁃system change.Clarifying the spatial and temporal trends and driving factors of FVC is an important research content of global and regional ecological environment.Based on Google Earth Engine(GEE)cloud computing platform,we estimated FVC in Heilongjiang Province from 1990 to 2020 using the pixel dichotomous model.We analyzed the temporal and spatial trends and drivers of FVC using Mann⁃Kendall mutation test,Sen’s slope analysis with Mann⁃Kendall significance test,correlation analysis,and structural equation model.The results showed that the estimated FVC based on the pixel dichotomous model had high accuracy(R^(2)>0.7,root mean square error<0.1,relative root mean square error<14%).From 1990 to 2020,the annual average FVC in Heilongjiang was 0.79,with a fluctua⁃ting upward trend(0.72-0.85)and an average annual growth rate of 0.4%.The annual average FVC at the munici⁃pal administrative districts level also showed different levels of increase of FVC.The area with extremely high FVC dominated the Heilongjiang Province with a gradual increase proportion.The area with increasing trend of FVC accounted for 67.4%of the total area,whereas the area with decreasing trend only accounted for 26.2%,and the rest remained unchanged.The correlation of human activity factor on annual average FVC was higher than that of growing season monthly average meteorological factor.The human activity factor was the main driver for FVC change in Heilongjiang Province,followed by land use type.The total effect of monthly average meteorological factor during the growing season on FVC change was negative.The results would serve as technical support for long⁃term FVC monitoring and driving force analysis in Heilongjiang Province,and provide a reference for ecological environment restoration and protection,as well as the formulation of related land use policy.