地震异常波反映了地下介质非连续性的重要信息,对小断层、弱构造、溶洞、孔洞等的准确定位有着重要的意义。在常规地震资料处理中,常常是使用频率域带通滤波和f-k滤波等方法将地震异常波压制掉,降低了成像的分辨率。因此,要想有效地利用异常波信息解决地质问题,首要任务是将它们完整的分离出来。由地震异常波理论分析和实际观测,借鉴常规地震波场分离与去噪技术,中值滤波和均值滤波方法,在共偏移距道集叠加剖面和原始地震数据中进行反射波预测,通过减去法提取地震异常波。将此方法应用于模型数据处理,证明了均值滤波方法在提取异常波所携带的地质信息时优于中值滤波。 Anomalous seismic waves reflect the important information of the discontinuity of underground media and are of great significance for the accurate positioning of small faults, weak structures, caves and holes. In conventional seismic data processing, it is often used to suppress the seismic anomaly waves by band-pass filtering in the frequency domain and f-k filtering to reduce the resolution of imaging. Therefore, in order to effectively use the abnormal wave information to solve geological problems, the primary task is to completely separate them. From the theory of seismic anomaly wave analysis and actual observation, the conventional seismic wave field separation and denoising technique, median filtering method and mean filtering method are used to predict the reflected wave in the superimposed section of common offset gathers and the original seismic data. Method to Extract Earthquake Anomalous Waves. Applying this method to the model data processing, it is proved that the mean filtering method is superior to the median filtering in extracting the geological information carried by the anomalous wave.