Research On The Inversion Method Of Aerosol Characteristics Based On Lidar Data

Aerosol is an important part of the atmospheric system.Although its relative content in the atmosphere is small,it has an important impact on the climate environment and atmospheric radiative transfer.Compared with the detection means of passive remote sensing and satellite,ground-based lidar,as an active optical remote sensing detection tool,has the advantages of high spatial-temporal resolution,long detection distance and high measurement accuracy,and it is widely used in aerosol measurement.Therefore,it is of great significance to research how to use lidar data to accurately measure aerosol related characteristics for atmospheric remote sensing,environmental monitoring and other fields.With the development of industrial manufacturing technology,lidar system is gradually mature,the cost is controlled,and the application is more common.The demands for lidar data processing and inversion model are more and more urgent in optical remote sensing industry.However,because of the large amount of noise in lidar signal,the research of lidar signal denoising method to extract the useful signal in the strong background noise is the premise to ensure the accurate inversion of lidar data.Furthermore,artificial intervention such as prior information assumption or parameters determination based on experience is often introduced when using lidar data,which is easy to cause large errors and difficult to achieve adaptive and automatic processing.These limitations are mainly reflected in the lidar data preprocessing,the parameters inversion of aerosol layer vertical structure,aerosol optical and physical properties.The research work of this dissertation will be devoted to solve these problems,which are shown as follows:(1)Lidar echo signals are easily contaminated by noise,particularly in strong background light,which severely affects the retrieval accuracy and the effective detection range of the lidar system.In this study,an adaptive lidar signal denoising method based on whale optimization algorithm(WOA)and variational mode decomposition(VMD)is proposed to overcome the interference problem of strong noise.The decomposition mode number and quadratic penalty factor are obtained by WOA method,which makes the VMD model achieve the better denoising effect.The experimental results show that this method can be successfully applied to the noise reduction of lidar signal,effectively improve the signal-to-noise ratio and extend the effective detection range of the lidar system used in the experiment from 6 to 10 km.(2)The planetary boundary layer height(PBLH)is a primary parameter characterizing the aerosol layer vertical structure,and its accurate estimation is critical for the prediction and research of weather and air quality.The wavelet covariance transform(WCT)method is a common inversion algorithm of the PBLH estimation,but is subject to the problems of dilation selection and interference from clouds,aerosols,etc.Based on the WOA and the top limit method,an improved WCT method is proposed in this dissertation.Without other auxiliary measurement equipment,this method can detect PBLH automatically,accurately and stably only based on simple micro pulse lidar without other auxiliary measuring equipment,which solves the above technical problems.(3)Aerosol extinction coefficient(AEC)is a crucial factor of aerosol optical properties,which has an important influencing in the atmospheric process.Mie scattering lidar is the most widely used lidar equipment.However,the empirical hypothesis and complex numerical calculation exist in AEC retrieval,which restricts the accuracy of inversion results.Combined with the high-precision retrieval results of high spectral resolution lidar,an inversion model based on deep belief network(DBN)is proposed to predict AEC.The experiment results indicate that the trained DBN model is robust and satisfactory,and provides a competitive solution for retrieving aerosol property parameters of Mie scattering lidar.(4)The microphysical parameters such as aerosol particle spectrum distribution can be used to monitor the evolution and study the temporal and spatial variation of atmospheric aerosol.Based on lidar equation and Mie scattering principle,the relationship between aerosol optical properties and microphysical properties is revealed.Aiming at the problems of regularization algorithm in retrieving aerosol particle distribution,the regularization parameters and aerosol complex refractive index are obtained by generalized cross-validation method and minimum deviation criterion.And then aerosol particle distribution inversion can be calculated based on multi wavelength lidar data.