Inversion And Application Of Complex Dielectric Constant Of Antarctic Ice Sheet Based On FMCW Rada

Mass loss from the Antarctic Ice Sheet will lead to a significant rise in global sea level under the influence of climate warming.In order to understand the response of the ice sheet to global warming and its contribution to sea level rise,it is necessary to monitor the quality of the East Antarctic ice sheet for a long time.The surface mass balance(SMB)model is essential for understanding the past and current state of the polar ice sheet and predicting its development mechanism under future climatic conditions.The SMB model includes parameters such as density,complex permittivity,crystal structure.At present,a large amount of ice sheet data is still needed to improve the accuracy of the SMB model.Since the density,complex permittivity,and crystal structure data depend on the collection and analysis of ice cores,large-scale research cannot be carried out due to logistical constraints;in addition,the process of estimating historical SMB data based on radar is cumbersome,and the automated processing technology for radar data is immature,which limits the efficiency of data analysis.Ice radar can study the temporal and spatial characteristics of the ice sheet on a large scale,but the method of estimating SMB based on radar is cumbersome,and the automatic processing technology has not been maturely applied,which limits the efficiency of data analysis.Based on the propagation characteristics of radar signals to invert the dielectric properties of the ice sheet,and the research on the inversion of the ice sheet parameters has broad application prospects for improving the accuracy of the model.By more fully analyzing the radar data obtained by domestic and foreign scholars,a large amount of SMB model parameter data can be obtained,which can effectively alleviate the problem of lack of ice sheet data,and provide a data basis for predicting the development trend of ice sheet quality.By investigating the research progress and development trend of SMB estimation methods and radar inversion technology,and focusing on the parameters of SMB models,this paper uses Frequency Modulated Continuous Wave(FMCW)radar to study the inversion method of ice sheet permittivity,and established the ice sheet permittivity inversion and density calibration algorithm,proposed the full waveform inversion scheme of crystal structure and conductivity,studied the reconstruction method of horizontal continuous isochronous layer,and finally realized the automatic estimation of SMB.The main works are summarized as follows.(1)This paper analyzes the complex dielectric parameters,density and radar propagation characteristics of snow/ice within 100 m of the ice sheet.Based on the empirical formula to realize the equivalent exchange of permittivity and density,combined with the propagation characteristics of the FMCW radar beat signal,it is clear that the ice sheet surface can be regarded as a one-dimensional horizontal layered structure;The peak amplitude of the echo signal reflects the local standard deviation of the ice sheet permittivity within the radar depth resolution.We use the stochastic medium model of the autocorrelation function to construct the internal crystal structure of the ice sheet,and combine with high-resolution ice core data to find the standard deviation of the permittivity can be replaced by the densification rate,which lays a theoretical foundation for the subsequent inversion algorithm;Based on the ice sheet stochastic medium model,the influence of model parameters on the echo signal is studied,which provides a computational model for the improvement of traditional full waveform inversion.(2)Aiming at the problem of low efficiency in continuous inversion of the permittivity/density based on a common center point inversion algorithm,this paper proposes an algorithm for continuous permittivity inversion based on a single radar reflection signal.This algorithm extracts the echo signal amplitude and converts it into reflection coefficient,and combines the layer stripping inversion method and the principle of FMCW radar to invert the permittivity.After converting the permittivity to density,the inversion density is combined with the densification model to calibrate the density error of stripping inversion method.The experimental results show that the root mean square error between the measured density of the ice core and the inversion result is within 5.54%.Continuous inversion of the 88 km density profile on the southwest side of the ice core DT401,and comparison with SMB data prove the effectiveness of the algorithm,which laid the foundation for conductivity inversion and SMB automatic estimation algorithm.(3)In order to obtain the characteristics of the crystal structure inside the ice sheet,this chapter proposes a full-waveform inversion scheme for inversion of the crystal structure roughness and electrical conductivity.Aiming at the dielectric properties of ice sheets showing gradual change in average permittivity and random distribution of microstructures,this chapter uses the permittivity inversion results based on the traditional full-waveform inversion method as prior information to construct a high-precision initial model,and establish boundary constraints on relative permittivity and conductivity based on statistical data to improve inversion performance;The stochastic medium model is used to characterize the crystal structure,and the ice crystal distribution characteristics are parameterized based on the roughness factor.We demonstrate the validity of the inversion results through data experiments and estimate the electrical conductivity and roughness factor at 39 km near LGB69 for reconstruction of the subglacial crystal structure.(4)This chapter introduces an automatic SMB estimation algorithm based on permittivity inversion,density calibration,and isochronous layer reconstruction.Firstly,on the basis of conventional image processing,deep learning denoising algorithm,image enhancement algorithm and contour detection algorithm are introduced to extract isochronous contour and horizon slope matrix.Establish a horizon contour model,which uses horizon shape and energy of radar image as objective functions,and is constrained based on isochronous layer contours,and uses dynamic programming algorithm to reconstruct continuous isochronous layer distribution based on slope matrix;We perform SMB automatic estimation based on density inversion and reconstructed isochronous layers,which have an average error of 6.49 kg/m~2/a from the traditional method;The automatic estimation of SMB and the analysis of the temporal and spatial variation characteristics of SMB are carried out around the ice core DT263 and the Dome A,which proves the effectiveness of the algorithm.