Hyperspctral Remote Sensing Of Leaf Biochemical Contents In Phyllostachys Eduis

Leaf biochemical parameters of Phyllostachys edulis are critical to the physiological and ecological processes and plays an important role in the material and energy cycle of the ecosystem.A clear understanding of the plant growth dynamic is an essential way to improve its economic and ecological benefits.However,traditional measurements of biochemical contents are generally time consuming,expensive,and often unfeasible at regional scale.In this study,we focused on Phyllostachys edulis,and conducted synchronous measurements of biochemical and hyperspectral reflectance.This research explored a new way to estimate biochemical contents from hyperspectral remote sensing by combining hyperspectral indice methods with PROSPECT model.In details:?1?Based on the field synchronous measured data,the calibrated PROSPECT model was used to retrieve the biochemical contents of Phyllostachys edulis.The results show that the R² and RMSE of retrieved chlorophyll content and measured chlorophyll content were 0.36 and 6.82?g cm^-2.The R² and RMSE of retrieved water content between measured water content were 0.24 and 0.00078 g cm^-2.The R² and RMSE of retrieved dry matter content between measured dry matter content were 0.21 and 0.0012 g cm^-2.The calibrated PROSPECT model could be applicable for invertion of biochemical contents but had some limitations.?2?Hyperspectral remote sensing of biochemical contents based on synchronous field measured dataset revealed that the first-derivative spectra based indices were more effective for tracing biochemical contents compared with those models based on the original reflectance.The R² and RMSE of the identified best index dDDn?1220,975?using first-derivative spectra for estimating chlorophyll content were 0.75 and 3.39?g cm^-2.The R² and RMSE of the identified best index dSR?2205,1305?using first-derivative spectra for estimating water content were 0.53 and 0.00035 g cm^-2.The R² and RMSE of the identified best index dND?1670,1135?using first-derivative spectra for estimating dry matter content were 0.67 and 0.00048 g cm^-2.?3?This research combined a simulated database generated by PROSPECT model with the field measured dataset to develop a potentially general and robust index for biochemical contents estimation.The R² and RMSE of the identified best index ND?565,2245?for estimating chlorophyll content were 0.75 and 3.30?g cm^-2,verification in the field measured data set,R² was 0.60.The R² and RMSE of the identified best index D?770,1465?for estimating water content were 0.58 and0.00036 g cm^-2,verification in the field measured data set,R² was 0.35.The R² and RMSE of the identified best index SR?1425,2300?for estimating dry matter content were 0.60 and 0.00055g cm^-2,verification in the field measured data set,R² was 0.31.This research highlighted a promising way to retrieve leaf biochemical contents from the hyperspectral remote sensing data.The results obtained in this study also would have provided a basis for monitoring ecological safety and stability in Phyllostachys edulis forest using remote sensing data.