不同植被指数反演植被覆盖度的真实性检验及对比分析

Validation and comparative analysis of fractional vegetation cover inverted by different vegetation indexes

针对NDVI的饱和特性对植被覆盖度估算产生负影响的问题,使用DVI推导植被覆盖度反演模型,利用相对真值进行检验,并从统计指标、地形影响和不同覆盖度影响对比了2种反演模型的精度。结果表明:真实性检验必须克服尺度效应的影响,f_(NDVI)、f_(DVI)不适用于高植被覆盖度地区;基于DVI的像元二分模型反演结果显示,f_(DVI)总体精度56.55%优于f_(NDVI)总体精度49.68%,结果的鲁棒性更好。

This paper intends to address the negative impact of saturation characteristics of NDVI on the estimation of vegetation coverage.The study is focused on the efforts to derive a vegetation coverage inversion model by DVI and test it by relative truth values,and compare the accuracy of the two inversion models from three aspects:statistical index,terrain influence and different coverage influences.The results show that the authenticity test must overcome the impact of scale effects,f_(NDVI) and f_(DVI) are not suitable for high vegetation coverage areas;according to DVI inversion results,the overall accuracy of f_(DVI) is 56.55%,which is better than that of f_(NDVI) by 49.68%,and the robustness is better.