甘肃省高寒草甸植被覆盖度反演及其时空变化研究

Study on Vegetation Coverage Inverse and Spatial-temporal Variation of Alpine Meadow in Gansu Province

本研究以甘肃省高寒草甸为研究区,基于2000—2019年遥感数据和2014年实测数据,采用经验回归模型法构建植被覆盖度(fractional vegetation cover,FVC)估算模型,并研究了过去20年高寒草甸FVC时空变化规律、稳定性及变化原因,以期为FVC动态监测提供科学依据。结果表明:在6种植被指数(vegetation index,VI)中除比值植被指数(ratio vegetation index,RVI)外,其余VI与FVC相关系数均大于0.65(P<0.01);模型精度检验发现高寒草甸FVC最佳反演模型为归一化植被指数(normalized difference vegetation index,NDVI)二项式模型:y=-0.65 x 2+1.97 x—0.23(R 2=0.81,RMSE=7.33);甘肃省高寒草甸FVC呈现南高北低的格局,20年FVC均值处于较高水平,年均FVC呈现波动上升趋势,平均增速0.15%·a-1,其中FVC增加的面积占52.76%,稳定的面积占27.58%,下降的面积占19.66%,FVC变异系数小于0.15的面积占86.57%。综上所述,2000—2019年甘肃省高寒草甸FVC整体上呈现增加趋势,并且稳定性较高。

In this study,the alpine meadow in Gansu Province was used as the study area.Based on remote sensing data from 2000 to 2019 and ground survey data in 2014 the remote sensing estimation model of FVC was built up by empirical regression model.The spatial-temporal variation,stability and causes of variation over the past 20 years were analyzed in alpine meadow to provide a scientific basis of dynamic monitoring in the future.The results showed that,in sex vegetation index,with the exception of RVI,the correlation coefficient between vegetation index and vegetation coverage was greater than 0.65(P<0.01).Through model accuracy test found that the best inverse model was the NDVI quadratic polynomial model:y=-0.65 x 2+1.97 x-0.23(R 2=0.81,RMSE=7.33).The FVC of alpine meadow in Gansu Province showed the pattern of high value in the south and low value in the north.The FVC of alpine meadow increased,stable and decreased percentages of FVC were 52.76%,27.58%and 19.66%,respectively.The annual average FVC in last 20 years kept a high level and it showed an increasing trend with an average growth rate of 0.15%·a-1.The alpine meadow area with low 0.15 of coefficient of variation accounted for 86.57%over the last 20 years.In general,the FVC of alpine meadow in Gansu Province showed an increasing trend and kept a high stability from 2000 to 2019.