The Comparative Study Of The Regularization Inversion And Of The Focusing Inversion Of Gravity Gradients Abnormal

Currently airborne gravity gradient has a rapid development.In the country,the processing and explaining work of gravity gradient start.Through foreign scholars Zhdanov's(1999,2000,2004)article,we studied the focusing inversion,programmed using FORTRAN,and used it in the inversion of the gravity gradient.Focusing inversion can get a more clearly and focused massive results than a most smooth results.Then through the example of the theoretical model,we compared the results with the single-component data of the focusing inversion and regularization inversion.The comparison of single-component data showes:1.In the determination of the top and bottom boundary of the geological body,the results of focusing inversion is significantly better than the results of the regularization inversion;2.Model for different depths,the two methods both identify shallow model better than the deep,identify the upper boundary better than the bottom boundary.However,with the increasing depth,the effect of identifying the bottom boundary become worse;3.For the same size and different depth combined model,the boundary determination of results focusing inversion is better to the lager one,but get worse to the small one with the proportion increasing;the determination of the regularization inversion results get better to the small one;4.For the complex shape model,both results have limitation,but the focusing inversion results are better;5.Focusing inversion has many advantages,but is slow because of iterative calculation(usually 10 min with 50 iterations).The regularization inversion has some drawbacks,but it runs very quickly(usually 30 s).After that we do some comparative study of the joint Uxz,Uzz two components data inversion.The results show that the joint use of the multi-component data increase uniqueness and enhance inversion resolution.Last the joint inversion results of the multi-component data with noise show that the results can still be satisfying.