基于加权模型参数的归一化磁源强度三维反演

发布时间：2018-07-02 11:35

本文选题：归一化磁源强度 + 加权模型参数　；参考：《石油物探》2017年04期

【摘要】：强剩磁的存在使磁化方向与地磁场方向偏差很大,进而对常规的磁异常反演和解释产生很大的影响。归一化磁源强度(normalized source strength,NSS)是一种弱敏感于磁化方向的转换量,它与场源的中心对应性比同类型的转换量要好。基于模型参数加权,采用共轭梯度的反演算法,使所有反演网格单元的综合灵敏度一致,以消除NSS核函数随距离的四次方衰减的影响。将这种改进的正则化共轭梯度算法应用于剩磁条件下的NSS反演。在研究区内存在剩磁情况下,与直接反演磁异常相比,NSS反演提供了更为可靠和稳定的解;与常规正则化反演方法相比,这种模型参数加权的方式能更好地降低NSS反演中核函数随距离的四次方衰减对反演结果的影响,且其反演结果能更好地刻画场源体的几何形态与物性分布。模型和实测数据测试结果证明了该剩磁条件下的反演方法的有效性和适用性。
[Abstract]:The existence of strong remanence makes the direction of magnetization deviate greatly from the direction of geomagnetic field, which has a great influence on the inversion and interpretation of conventional magnetic anomalies. Normalized magnetic source intensity (normalized source) is a weakly sensitive conversion to the magnetization direction, and its correspondence to the center of the field source is better than that of the same type. Based on the weighted parameters of the model, the conjugate gradient inversion algorithm is used to make the comprehensive sensitivity of all inversion grid elements consistent, so as to eliminate the influence of the attenuation of the NSS kernel function with the fourth power of the distance. The improved regularized conjugate gradient algorithm is applied to the NSS inversion under remanent magnetic conditions. In the case of remanent magnetism in the study area, the NSS inversion provides a more reliable and stable solution than the direct inversion of the magnetic anomaly, and compared with the conventional regularization inversion method, NSS inversion provides a more reliable and stable solution than the direct inversion of the magnetic anomaly. The weighted model parameters can better reduce the influence of the decay of kernel function to the fourth power of distance on the inversion results, and the inversion results can better describe the geometric shape and physical property distribution of the field source. The validity and applicability of the inversion method under the condition of remanence are proved by the model and measured data.
【作者单位】：同济大学海洋地质国家重点实验室;
【基金】：国家科技重大专项(2016ZX05004-003,2016ZX05027-001-008) 国家重点研发计划深地专项项目(2016YFC0601104) 国家自然科学基金(41506053) 海洋地质国家重点实验室自主课题(MG20170204)共同资助~~
【分类号】：P631.2

【参考文献】