Efficient 3D inversion of potential field data using fast proximal objective function optimization algorithm

One of the key issues in three-dimensional (3D) potential field data inversion is computational efficiency. We have developed an efficient 3D potential field data inversion method using fast proximal objective function (FPOF) optimization algorithm. A distinguishing feature of the proposed inversion...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Journal of applied geophysics Jg. 159; S. 108 - 115
Hauptverfasser: Peng, Guomin, Liu, Zhan, Xu, Kaijun, Bai, Yongliang, Du, Runlin
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier B.V 01.12.2018
Schlagworte:
3D inversion
Fast proximal objective function optimization
Potential field
3D inversion
Potential field
Fast proximal objective function optimization
ISSN:0926-9851, 1879-1859
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:One of the key issues in three-dimensional (3D) potential field data inversion is computational efficiency. We have developed an efficient 3D potential field data inversion method using fast proximal objective function (FPOF) optimization algorithm. A distinguishing feature of the proposed inversion method is to calculate each unknown element in the model space independently at each iteration. In addition, based on subsequent data misfits in the iterative process, we design a computationally efficient method to choose an appropriate regularization parameter. We take gravity data inversion as an example and apply the proposed inversion method to synthetic data contaminated with noise. The results demonstrate that the proposed method provides almost identical results as the conventional conjugate gradient (CG) method, however, with a significant reduction in the execution time. Lastly, an application on real field data acquired in the Western China is carried out, yielding a 3D density model which is in good agreement with the available geological knowledge. •Each unknown element in the model space is calculated independently at each iteration.•An appropriate regularization parameter is determined through subsequent data misfits.•A significant reduction in the execution time is obtained in comparison to the traditional CG method.
ISSN:0926-9851
1879-1859
DOI:10.1016/j.jappgeo.2018.08.002