基于HJ-1A CCD1数据的台湾相思树叶面积指数反演

基于HJ-1A CCD1环境卫星数据,以福建沿海地区普遍分布的台湾相思树为研究对象,利用回归分析法(NDVI、OSAVI、EVI、HJVI)和PROSAIL辐射传输模型,构建台湾相思树LAI反演模型。同时,利用同步野外地面实测数据,将模型估算LAI值与实测LAI值进行对比。结果表明：(1)相比归一化植被指数NDVI、优化土壤调节指数OSAVI和增强型植被指数EVI 3种常用植被指数,引入修正大气、土壤背景影响的蓝、绿波段的环境植被指数HJVI来反演相思树LAI具有更高的精度(R²=0.7344,RMSE=0.1421);(2)本研究所选4种植被指数构建的最优反演模型均为非线性模型,其中,环境植被指数HJVI反演LAI最优模型为幂函数模型,表明相思树LAI与植被指数之间呈非线性变化;(3)PROSAIL辐射传输模型法比回归分析法反演相思树LAI的精度有较大提高(R²=0.7903,RMSE=0.1303),可见PROSAIL模型法构建反演模型能更好地反演相思树LAI。

Comparison of Different Methods for Retrieving Acacia Rachii Leaf Area Index Based on HJ-1A CCD1 Imagery

With Acacia Rachii（Acacia confusa） grown in coastal region in Fuzhou as research object, and based on HJ-1A CCD1 imagery which was acquired from China Center for Resource Satellite Data and Applications,the Acacia Rachii LAI was monitored in field using LAI-2000 canopy analysis system, and two kinds of universal LAI inversion methods through regression analysis method and radiative transfer model PROSAIL model separately were introduced and used in this study. The simulation precisions for different models were analyzed and evaluated through comparing the simulated LAI and measured LAI. Further, we compared the research output with those of previous researchers. The results showed that：（1） Compared with the three vegetation indices（NDVI, EVI and OSAVI）, HJVI vegetation index performed best in Acacia Rachii LAI inversion among all of the vegetation indices with R2=0.7344 and RMSE=0.1421, which introduced blue band and green band in order to weaken the effects of atmosphere and soil;（2） The optimal inversion models of the above four vegetation indices all were non-linear models, and the optimal regression model for Acacia Rachii LAI inversion based on vegetation indices was the power regression model of HJVI, indicating that there existed non-linear relationships between Acacia Rachii LAI and vegetation indices;（3） There had obvious improvement in the precision of LAI inversion through PROSAIL model compared with regression analysis method based on vegetation indices with R2=0.7903 and RMSE=0.1303, which indicated that PROSAIL model could better estimate Acacia Rachii LAI than regression analysis method to some extent. Therefore, radiative transfer model such as PROSAIL used to construct the inversion model is feasible. It could reflect ground condition better and possesses higher application value and broad application prospect.