Inversion Study Of Canopy Chlorophyll Content Of Jujube Trees Based On Sentinel-2 Data

In this study,a part of the 224th regiment of the 14th division of Xinjiang Production and Construction Corps was used as the study area,and the Sentinel-2 images of the budding,flowering,and maturity stages were used as the main data source.The land use situation in the study area was supervised and classified by SVM support vector machine classification method,and the planting areas of date palm in the study area were extracted.On this basis,thirteen single-band and combined-band factors,six vegetation index factors and nineteen other remote sensing factors were extracted as independent variables,and five data regression models,including multiple stepwise regression,BP neural network,decision tree,random forest and XGBoost,were established with ground-measured date palm canopy chlorophyll content data at different fertility periods in the study area as dependent variables to investigate the accuracy of the models,and to investigate the accuracy of The inversion of chlorophyll content in date palm at different fertility canopy scales was thoroughly investigated.The specific research work and conclusions are as follows:(1)In order to carry out the inversion of chlorophyll content of date palm canopy,it is necessary to first classify the land use types in the study area and extract the remote sensing factors of date palm planting areas.For this purpose,the Sentinel-2 remote sensing data can be preprocessed to obtain high-quality image data.Then,the support vector machine(SVM)algorithm is used to classify the land use of the study area,which is finally classified into four categories:building land,date palm area,water body,and bare land.Finally,the remote sensing factors of date palm plantation areas can be extracted based on the classification results for the inversion of the chlorophyll content of date palm canopies.(2)Among the 19 remote sensing variables extracted,for the correlation distribution of remote sensing factors in different fertility periods,all factors,except for the SWIR 1 band factor at the budding and new shoot stage,achieved significant or highly significant correlation with the chlorophyll content of date palm canopy in each period,among which,Blue,Green,Red,Aerosols,Red Edge 1,Red The selected vegetation index factors NDVI,EVI,DVI,GVI,RVI,SAVI,and OSAVI were mostly correlated with the chlorophyll content of date palm canopies.(3)The canopy chlorophyll content of date palm in the study area showed an increasing trend from the budding neonatal stage to the maturity stage.Through the comparison of five data regression models,including MSR,BP,DT,RF and XGBoost,for inversion of date palm canopy chlorophyll content at different fertility stages using four evaluation indexes,R~2,MAE,MSE and RMSE,it was found that the XGBoost model performed best at different fertility stages with high prediction accuracy and stability.At the budding and new growth stage,the XGBoost model had R~2of 0.974,RMSE of 0.822,MSE of 0.677,and MAE of 0.678,which were the best performances compared with other models.In the flowering stage,the XGBoost model had R~2of 0.752,RMSE of 3.858,MSE of 14.883,and MAE of 3.307,which also performed best among all models.The XGBoost model can invert the canopy chlorophyll content of jujube trees at different fertility stages quickly and accurately,and generate a spatial distribution of chlorophyll content for each fertility stage.The results were consistent with the actual growth of jujube in the study area,and the development of this study can provide theoretical support and technical guidance for the observation of jujube growth and development and the fine management of jujube woodlands in 224th Kunyu Mission,Xinjiang.