3D Joint Inversion Of Airborne Electromagnetic And Magnetic Data Based On Pearson Correlation Constraints

Airborne electromagnetic(AEM)and airborne magnetic(Air Mag)methods are common methods used in geophysical exploration.AEM method has the advantages of high vertical resolution,while Air Mag method has the advantages of large detection depth with high horizontal resolution.They play an important role in the fields of mineral resources exploration,groundwater investigation and geological disaster prevention and control.At present,the data of the AEM method and the Air Mag method can be measured at the same time,thus,the measured time and the measured location of the above data are highly coincident,which provides very favorable conditions for joint inversion.Joint inversion can combine the advantages of two methods,which can reduce the multi-solution of separate inversion and improve the resolution of exploration.Although the joint inversion method has such observation advantages,the data of AEM and Air Mag methods are still interpreted separately,and the complementarity of their data information is not exploited.Therefore,this paper proposes to carry out a three-dimensional joint inversion study of AEM and Air Mag data.The main research contents are as follows:1)Based on previous research,this paper introduces the Pearson correlation coefficient,which is one of the three major coefficients in statistics,as the correlation constraint of parameters to realize the 3D joint inversion of AEM and Air Mag data.The Pearson correlation coefficient is not affected by the dimension or magnitude of the data,which is more suitable for joint inversion between different physical parameters in geophysics.In practical calculations,firstly,we divide the inversion region into multiple intersecting 3D subdomains.Then,calculate the Pearson correlation coefficient between the electrical parameters and magnetic parameters in each subdomain,and in the subsequent iterative inversion process,we use inversion constraints to enhance the linear correlation,to realize joint inversion constraints in the overall region.2)In this paper,the commonly used SEG/EAGE model is designed to verify the effectiveness of the joint inversion algorithm.In the inversion test of the theoretical model,we calculate the separate 3D inversion results of the AEM method and the Air Mag method,and compare the results of the cross-gradient 3D joint inversion methods with the 3D joint inversion results based on the Pearson correlation coefficient.The comparison shows that both the traditional cross-gradient joint inversion and the joint inversion based on the Pearson coefficient proposed in this paper can improve the inversion resolution,but the joint inversion of the Pearson coefficient has a greater advantage in the depth resolution.3)In order to more clearly show the influence of subdomain size on the inversion results,we tested the influence of three different subdomain sizes on the final inversion results.The results show that the selection of subdomains with different sizes has a great influence on the inversion results.If the size of the subdomain is too small,the two parameter models cannot be coupled,leading that the improvement of the joint inversion is not obvious;on the contrary,if the size of the subdomain is too large,the relationship between the two parameters in the subdomain is not a simple linear correlation,which is not conducive to the structural coupling of parameters.Therefore,we give an empirical selection method through testing.4)Based on the measured data in Lofoten-Vesteralen area of northern Norway,we calculate the results of the separate AEM inversion,the separate Air Mag inversion and the joint inversion based on the Pearson correlation coefficient.By comparing the results with the actual geological conditions obtained by predecessors,it can be found that the joint inversion based on local Pearson correlation coefficient can effectively reduce the multi-solution problem caused by traditional separate inversion,and can more accurately determine the location and shape of underground abnormal body.A novel 3D joint inversion method based on local Pearson correlation constraint is proposed in this paper.Through numerical test,our method shows obvious improvement in the inversion of theoretical models and measured data,thus,it is a joint inversion technology with broad application prospects.