Fault surface extraction based on computational topology

Abstract

Fault surface extraction is a key component for seismic interpretation. By constructing the spatial distribution of faults, fault surface extraction techniques provide critical clues for structural interpretation and modeling. There are two main challenges that would affect the final results in extracting fault surfaces from the fault attribute data: one is to ensure the integrity of the fault surfaces, and the other is to deal with complex faults. To address the above issues, we propose an automatic fault surface extraction method based on computational topology, which outperforms in maintaining the integrity of the extracted surfaces while effectively handling the complex faults such as intersecting faults in the field data test. Given a seed point on the target fault manually or automatically, we search for control points of this fault in the fault attribute data and obtain the fault boundary. Then we extract the fault surface from the fault attribute data using collapse, an operation in computational topology, under the constraint of the fault boundary. The topology-preserving property of collapse operation can ensure the integrity of the surface. Meanwhile, since our search process allows a relatively large distance, the segmented faults can be recombined. We test our method on the field data. The current result demonstrates the effectiveness of our method in terms of fault integrity, fault location accuracy, and the capability for dealing with intersecting faults.