Spectral Characteristics Analysis And Digital Mapping Of Populus Ekphratic And Populus Pruinose

Populus euphratica and Populus pruinosa,as unique vegetation resources in the Tarim River Basin,maintain the stability of desert riparian ecosystems.As the habitat of P.euphratica forest gradually deteriorated,its physiological indicators also changed accordingly.Using remote sensing technology to invert and model the physiological indexes of P.euphratica forest can provide technical support for the health monitoring of P.euphratica forest,thus providing a theoretical basis for the ecological conservation of P.euphratica forest in the Tarim River Basin.In this study,the leaf samples of P.euphratica and P.pruinosa in different months were collected,and the spectral reflectance data were obtained.The variation characteristics of spectral reflectance and physiological indexes of P.euphratica and P.pruinosa in different periods were analyzed.The canopy spectrum was simulated by PROPECT-SAIL model,and the physiological and biochemical component model of P.euphratica and P.pruinosa canopy was constructed.The digital mapping of physiological and biochemical components of P.euphratica and P.pruinosa was completed by hyperspectral image.The main analysis conclusions are as follows :(1)The total variation tendency of leaf moisture(LWC)was similar between P.euphratica and P.pruinosa.The LWC of P.euphratica and P.pruinosa decreased in July.At this time,the transpiration or evaporation rate of P.euphratica and P.pruinosa were higher than that of water absorption,which led to a reduction in the amount of moisture in leaves.Overall,the LWC of P.pruinosa leaves was higher than that of P.euphratica,and the content of chlorophyll a,chlorophyll b and total chlorophyll in the leaves of P.euphratica were significantly higher(P < 0.01),and the content of chlorophyll a,chlorophyll b and total chlorophyll in the leaves of P.euphratica were obviously higher(P < 0.01).The N content in leaves of P.pruinosa was higher than that of P.euphratica(P < 0.01),and P content in leaves of P.pruinosa was significantly lower than that of P.euphratica(P < 0.01).These phenomena reflect the resistance of Populus euphratica and Populus pruinosa to drought stress.(2)The original spectral reflectance curves of P.euphratica and P.pruinosa leaves are basically the same.The leaf spectral reflectance curves of P.euphratica and P.pruinosa are quite different in the green peak,red valley and near-infrared band.The spectral curve after removing the envelope has obvious difference in the absorption valley of 400-700 nm.The first derivative has a significant difference in the blue to red band,and the second derivative spectrum has a significant difference in the partial band range of near infrared.(3)The PROPECT-SAIL model was used to simulate the canopy spectra of P.euphratica and P.pruinosa,and the LWC、Chl a、Chl b、Chl a+b、N and P estimation models based on support vector machine(SVM)and partial least squares(PLSR)were constructed.The results showed that the SVM estimation model constructed by the second derivative transform spectrum of P.pruinosa and LWC was the best,and the modeling parameters and verification parameters passed the accuracy verification.The R2 were 0.897 and 0.785,the RMSE were 0.015 and 0.378,and the RPD were 2.737 and 2.026,respectively.RPD is greater than 2,which can accurately estimate the sample.At the same time,the modeling accuracy of P.pruinosa is generally better than that of P.euphratica,and the accuracy of SVM model is better than that of PLSR model,which also provides a basis for the subsequent construction of image-based physiological index estimation model.(4)The support vector machine method was used to classify the Zhuhai-1(OHS)image in the study area.The total classification accuracy was 89.01 %,and the Kappa coefficient was 0.873.The classification accuracy meets the needs of subsequent research.At the same time,the SVM model of Populus euphratica physiological index was constructed by using OHS hyperspectral image.The results showed that the accuracy of Chl a+b and N estimation model was the highest.According to the constructed model,the inversion mapping of Chl a+b and N of Populus euphratica and Populus pruinosa was realized.