High-resolution velocity analysis and program design based on Bootstrap algorithm

3rd International Conference on Energy Engineering and Environmental Protection

能源学;生态环境科学

Cai, Dechao^1 ; Fa, Guifang^1 ; Li, Jia^1 ; Xing, Xing^2 ; Yi, Yanjing^1

Research Institute of Petroleum Exploration and Development, Beijing, China^1

Research Institute of Petroleum Exploration and Development, Dagang Oil Field, Tianjin, China^2

关键词: Bootstrap algorithms;    Conventional methods;    High resolution velocity;    Matrix transformation;    Seismic data processing;    Seismic wave velocity;    Statistical distribution;    Theoretical derivations;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/227/4/042029/pdf
DOI  :  10.1088/1755-1315/227/4/042029

学科分类：环境科学（综合）

来源: IOP

【 摘 要 】

Seismic wave velocity is applied in all processes of seismic data processing and interpretation, and it is one of the most important parameters in seismic exploration. How to obtain high-resolution stack velocity spectrum [1], and then understand accurate velocity distribution of underground media is the key to seismic data processing. Conventional velocity analysis is based on multi-track cross-correlation within a time window defined along a hyperbola on a CDP gather, so it reflects the changes in signal stack energy within the window. The Bootstrap method is derived from mathematical statistics and is a mathematical method based on matrix transformation, which can accurately estimate the statistical distribution of given data. Based on conventional velocity analysis, this paper introduces how to use the Bootstrap method [2] to greatly improve the resolution of stack velocity spectrum, especially in processing data of shallow layers, large offsets and thin layers, based on old data and without additional investment. The application has provided high-resolution results and can effectively compensate many shortcomings of conventional methods. Based on theoretical derivation and process design, this paper presents the software designing process based on the data structure of an orthogonal list, then verifies the results using model and actual data, and finally compares the results with those from popular commercial software to verify the stability of the program and the reliability of the method.