A regression approach for seismic first-break picking  

作　　者：Huan Yuan San-Yi Yuan Jie Wu Wen-Jing Sang Yu-He Zhao 

机构地区：[1]CNPC Key Laboratory of Geophysical Exploration,China University of Petroleum(Beijing),Beijing 102249,China [2]Research Institute of Petroleum Exploration&Development—Northwest(NWGI),Key Laboratory of reservoir characterization of CNPC,PetroChina,Lanzhou 730020,Gansu,China

出　　处：《Petroleum Science》2024年第3期1584-1596,共13页石油科学（英文版）

基　　金：financially supported by the National Key R&D Program of China(2018YFA0702504);the National Natural Science Foundation of China(42174152);the Strategic Cooperation Technology Projects of China National Petroleum Corporation(CNPC)and China University of Petroleum-Beijing(CUPB)(ZLZX2020-03);the R&D Department of China National Petroleum Corporation(2022DQ0604-01)。

摘　　要：The picking efficiency of seismic first breaks(FBs)has been greatly accelerated by deep learning(DL)technology.However,the picking accuracy and efficiency of DL methods still face huge challenges in low signal-to-noise ratio(SNR)situations.To address this issue,we propose a regression approach to pick FBs based on bidirectional long short-term memory(Bi LSTM)neural network by learning the implicit Eikonal equation of 3D inhomogeneous media with rugged topography in the target region.We employ a regressive model that represents the relationships among the elevation of shots,offset and the elevation of receivers with their seismic traveltime to predict the unknown FBs,from common-shot gathers with sparsely distributed traces.Different from image segmentation methods which automatically extract image features and classify FBs from seismic data,the proposed method can learn the inner relationship between field geometry and FBs.In addition,the predicted results by the regressive model are continuous values of FBs rather than the discrete ones of the binary distribution.The picking results of synthetic data shows that the proposed method has low dependence on label data,and can obtain reliable and similar predicted results using two types of label data with large differences.The picking results of9380 shots for 3D seismic data generated by vibroseis indicate that the proposed method can still accurately predict FBs in low SNR data.The subsequent stacked profiles further illustrate the reliability and effectiveness of the proposed method.The results of model data and field seismic data demonstrate that the proposed regression method is a robust first-break picker with high potential for field application.

关 键 词：First-break picking Low signal-to-noiseratio Regression BiLSTM TRAVELTIME Geometry Noisy seismic data 

分 类 号：P631.4[天文地球—地质矿产勘探]