Study On The Retrieval Of Different Wetlend Vegetation LAI Using Multi Source Remote Sensing Data

Wetland vegetation plays an important role in maintaining the stability of wetland ecosystem.The leaf area index of wetland vegetation controls many biological and physical processes.LAI of wetland vegetation is a key index to measure primary productivity,ecological environment and health status of wetland ecosystem.Therefore,the study of wetland vegetation LAI is of great significance to the stability of ecosystem.Today,more and more remote sensing technology is presented in the forefront of the field,such as airborne LiDAR technology,hyperspectral remote sensing technology,satellite borne LiDAR technology,microwave radar technology,and so on.What's more,more and more remote sensing techniques have been applied to the study of wetland vegetation.The main innovations and features of this paper include the following aspects.First,the characteristic parameters of airborne LiDAR data,such as CHM and LPI,are integrated into the LAI inversion.The experimental results show that,the MSE accuracy of the LAI inversion model of reed was increased from 0.235 to 0.043 by CHM,and the MSE accuracy of the LAI inversion model with CHM of the salt artemisia and spartina alterniflora are increased by 0.084 and 0.160.Second,the ARDI which is defined as the classification feature in this paper,improves the overall classification accuracy by 9.2%.Compared with traditional algorithms,the accuracy of random forest algorithm based on object oriented has been improved significantly.Third,the RF algorithm and SVR algorithm are realized in the construction of LAI inversion model.The results show that the SVR models of LAI inversion of different wetland vegetations all have achieved the highest accuracy compared with stepwise regression models and RF models,while the accuracy of RF regression model is not stable.This paper not only provides a theoretical and technical support for the study of wetland vegetation,but also provides a strong basis for the application of remote sensing technology in the field of ecology.