Correlation Analysis between Biochemical Contents of Camphor and in Situ Hyperspectral Measurements

The correlation analysis between biochemical contents of trees and in situ hyperspectral measurements provides the basis for forest remote sensing. Because trees especially arbor trees often are very tall and have a long growth period, it is very hard to collect hyperspectral measurements of crowns and thus it is difficult to analyze the relationships of their biochemical contents with hyperspectral data. This study focuses on the correlation analysis between the biochemical contents of young camphor trees and hyperspectral measurements obtained using a hyperspectrum analyzer from the Analytical Spectral Devices, (ASD). The results show that chlorophyll-a, chlorophyll-b, and chlorophyll-ab are significantly correlated with the maximum of the first differential of red-edge and the reflectance of low-point of red-edge, while the correlation of carotene, cellulose, phosphorus, and nitrogen with the hyperspectral location variables is very weak. Moreover, chlorophyll-a, chlorophyll-b, chlorophyll-ab, and phosphorus have significant correlation with the area of blue-edge but not with the areas of yellow-edge and red-edge. Carotene, cellulose, and nitrogen are weakly correlated with all the hyperspectral area variables. Chlorophyll-a, chlorophyll-b, and chlorophyll-ab have statistically significant correlation with the vegetation indices including Rg/Ro, (Rg-Ro)/ (Rg+Ro), SDr/SDb, (SDr-SDb)/(SDr+SDb), and weak correlation with SDr/SDy and (SDr-SDb)/ (SDr+SDb). Phosphorus is significantly correlated with SDr/SDb, (SDr-SDb)/ (SDr+SDb) and weakly correlated with other vegetation indices. In addition, the correlation of carotene, cellulose, and nitrogen with the vegetation indices is not significant. These findings provide the basis on which the models that account for the relationships of biochemical contents with hyperspectral variables can be developed.