| Hyperspectral remote sensing not only inherits the large area,rapid and dynamic advantages of traditional remote sensing data,but also as another milestone in the development of remote sensing,can provides more abundant spectral information,which has a greater advantages in recognition of forest vegetation type and inversion of vegetation physical and chemical parameters.In this study,the maximum carboxylation rate Vcmax,light use efficiency LUE and the corresponding leaf spectral data were obtained by field measurement.Based on the correlation analysis of LUE and leaf reflectance,the traditional PRI was improved to PRI518 by using518nm instead of 531nm,and the correlation between PRI518 and LUE was improved.In order to further improve the accuracy of LUE estimation,PRI518 was further corrected to SRPRI518,R2 was increased from 0.6281 to 0.6861,RMSE was reduced from 0.0027 to 0.0024.In addition,based on the accuracy of Vcmax calculation,a remote sensing quantitative estimation model of Vcmax based on LUE is established.The R2,RMSE and the accuracy of model was0.7207,8.439?mol?m-2?s-1and 75.68%.The estimation model of Vcmax was established basing on the leaf spectral and the model can not be directly used to remote sensing image.In this study,the 4-Scale geometric optical model and the look-up table was used to the transformation of the canopy spectrum to the leaf spectrum,Therefore the inversion model was last used for estimating Leaf-level Vcmax,which was converted to canopy-level combining with leaf area index LAI.Statistical method was also used for leaf area index calculation.The linear model of leaf area index and NDVI was established by using the measured data.The R2,RMSE and the accuracy of model was 0.8306,0.523 and 84.58%.Canopy scale mapping of Vcmax was achieved by leaf area index from inversion of remote sensing data and Leaf-level Vcmax.The results show that the Vcmax inversion result was consistent with the actual situation. |
PDF Full Text Request