初至智能拾取技术

地震记录的初至自动拾取一直是复杂地形和表层条件下地震勘探工作的一个难题。为此,提出了初至智能拾取技术,模拟人机交互初至半自动拾取过程进行初至拾取。首先通过改进Coppens的滑动时窗能量比方法进行"好初至波"拾取;然后通过模拟人工辅助线的"架桥"技术获得可用初至波;再采用波峰(谷)相位域追踪技术获得初至波峰点,进一步采用炸药震源初至波起跳点的相位域估算技术获得初至波起跳点。在南方复杂地区,采用自主研发的Hunter初至智能拾取软件对靶区地震资料进行了初至拾取,并与OMEGA商业软件初至自动拾取结果进行了比较。结果表明,初至智能拾取方法及软件不仅能够实现高精度、高命中率的智能化初至拾取,而且能够提供初至时间、初至波周期和初至波品质等更多的信息;在复杂地区低信噪比地震资料中更能体现Hunter初至智能拾取软件的优越性。

First break intelligent picking technique

The precise first break picking technique is the prerequisite for dramatic increasing amount of data processing.Under complex to- pography and near surface geological conditions,the automatic first break picking is always a knotty problem in seismic exploration. Therefore,a new intelligent first break picking method was pro- posed,which utilizes the new concept of simulating interactive semi-auto first break picking.The technique is consisted of four parts:first break picking with high S/N ratio data,bridging with complementary line,peaking (trough) tracking in phase domain, and estimating the earliest point of the first arrival for explosive source.Actual data application shows that this method is suitable for the low S/N data from complex areas and is able to realize the intelligent first break picking with a high precision and high focu- sing factor and to provide more information for users,including first break time,first break cycle,and first break quality.