地震处理把在地表采集的通常为兆兆字节的大量数据转换成精确的地下图像,其关键步骤是子波处理(如噪声抑制、反褶积、多次波压制)和成像。 子波处理涉及到大量的数据管理,每个记录字的计算量中等,人的干预相对较少。三维成像则相反,每个记录字的计算量极大,数据管理量中等,在速度分析和模拟中要求大量人工干预。 Seismic Processing Converts large amounts of terabytes of data collected on the surface into accurate underground images, the key steps of which are wavelet processing (such as noise suppression, deconvolution, multipath suppression) and imaging. Wavelet processing involves a large amount of data management, with a moderate amount of calculations for each recorded word, with relatively few human interventions. In contrast to 3D imaging, the computational complexity of each recorded word is extremely large and data management is moderate, requiring extensive manual intervention in speed analysis and simulation.