AVO INVERSION WITH THE INVERSE OPERATOR ESTIMATION ALGORITHM

YIN Xing-Yao,DENG Wei,ZONG Zhao-Yun

DOI: https://doi.org/10.1002/cjg2.20235

2016-01-01

Chinese Journal of Geophysics

Abstract:Seismic inversion is generally implemented with certain optimization algorithm. However, the inverse operator estimation algorithm proposed in this study is to perform the inversion of data matrix directly under the hypothesis that the inverse mapping exists in the empirically constrained subspaces. The key point of the proposed approach is to search those subspaces instead of searching for the solution indirectly as optimization algorithms do and it's more efficient. AVO/AVA (amplitude variation with offset or angle) inversion is widely utilized in exploration geophysics, and the inversion process is restricted by the quality of seismic data. L1 norm is applied in the construction of the kernel function of inversion by combining the constraint from initial models, which is helpful in enhancing the efficiency and stability of the inversion. Model and field data examples indicate that the proposed AVO inversion algorithm based on inverse operator estimation is more accurate and reliable.