位场数据反演中深度分辨率及改进措施

The depth resolution in potential-field inversion and some improvement measures

位场勘探是一种较为成熟的地球物理勘探手段,然而其数据反演中存在的深度分辨率不足问题阻碍了其进一步发展.文章在深入研究位场数据反演理论的基础上,详细分析了位场反演中普遍存在的深度分辨率不足问题,分析了造成这种现象的主要客观原因.研究利用深度分辨率成图技术对给定位场反演问题的深度分辨率进行定量刻画.针对造成分辨率不足的原因提出了几点改进措施:利用离散Picard定理分析问题的适定水平,结合傅立叶系数与奇异值对比图进行合理的噪声剔除;反演计算之前结合DRP技术,给出较为合理的地下剖分方式;合理的数据加权.文章使用理论二维重力模型进行说明,最后将其应用于复杂三维重力模型,试验表明上述措施对于改进位场反演深度分辨率有一定的效果.

Potential field detection is one of the mature geophysical prospecting techniques. However,the limited depth resolution in the inversion damped its further development.On the basis of intensive study of potential-field inversion,we analyzed the depth resolution in potential-field inversion.The reasons which make sense are analyzed in detail. During our research,the depth resolution plot introduced by Fedi is used.This plot can character the depth resolution of a given problem quantitatively.Faced with this problem,some improvement measures are discussed here.The discrete Picard theorem can be used to evaluate the amount of noise in a given measured data.Combined Fourier coefficient with singular value plot,we can use proper methods to filter out the contaminated noises.Meanwhile,the depth resolution plot can be used to guide us to take a more reasonable discrete way.A proper weighting can also be used to achieve a better depth resolution.In the paper,a 2Dgravity model is used to illustrate the theory.At last,a more complex 3Dgravity is used to test all the reasons and it illustrates that better depth resolution can be acquired if all of the conditions are satisfied.