VSP初至逐层递推层速度反演研究及应用

Research and application of interval velocity inversion for VSP data by layer-by-layer recursion algorithm

利用VSP资料准确获取地层速度,对地震数据时深转换和偏移成像具有重要的作用。受激发接收条件变化、近源动态干扰和初至拾取误差等因素,VSP资料初至时间往往含有一定程度的随机干扰。笔者研究了当VSP资料初至时间存在随机噪声干扰和仅有深部接收数据的情况下,不同井源距VSP资料逐层递推反演层速度的误差特性。研究表明:①井源距越大,其对应资料初至反演速度误差越小;②当仅有深部资料时,井源距越小,其资料反演结果与实际地层速度更接近。为此,提出初至时间平滑后再反演,多井源距反演结果求平均的方法,从而提高层速度反演精度。利用该研究结论,在某海上Walkaway-VSP资料处理中,取得了较好的效果。

The use of VSP data to accurately acquire formation velocities plays an important role in time-depth conversion and migration imaging of seismic data.Due to factors such as changes in receive and excitation conditions,near-source dynamic interference,and first arrival time pick-up errors,the first arrival time of VSP data often contains a certain degree of random interference.The authors studied the error characteristics of inverting interval velocity for VSP data with different offsets by layer-by-layer recursion algorithm when VSP data have random noise at the first arrival time and only deep data are received.Some conclusions have been reached:①The larger the offset,the smaller the inversion velocity error of the corresponding data;②When the data is only recorded from the deep level,the smaller the offset,the closer the inversion result to the actual formation velocity.For this reason,the first arrival time should be smoothed before being inverted by layer-by-layer recursion algorithm,and then the average of inversion results is taken from multi offset VSP data as the final interval velocity.This method can improve the inversion accuracy of the interval velocity.By using this research conclusion,good results were obtained in the processing of marine Walkaway-VSP data.