细分表层速度场约束层析静校正技术在西昆仑山黄土区的应用

Application of Sub-divided Near-surface Velocity Field Constrained Tomography Statics in Loess Areas of Western Kunlun Mountain

塔里木盆地西昆仑山前覆盖巨厚的黄土,黄土厚度大且纵横向分布差异大,存在严重的静校正问题。要搞清黄土区的静校正问题,就要利用野外地震采集资料的大炮初至时间与黄土厚度变化有关联的特点,把拾取的大炮初至时间在横向上根据黄土厚度大小不同细分成多个区域的初至时间数据体,对每个区域数据体在纵向上根据黄土覆盖层呈连续介质特征把低降速层细分成多个速度层,通过横纵向细分表层速度场,建立符合黄土山近地表初始速度模型来约束层析反演,反演后的模型更可靠,计算的静校正量精度更高。在黄土区三维和弯线实际资料处理应用中,细分表层速度场约束层析静校正技术较好地解决了巨厚黄土区的静校正问题,使得资料成像效果明显提高,该方法具有较好的应用前景。

Loess of significant thickness covers the foreland of western Kunlun mountain in Tarim basin. The loess is considerably thick and variable both vertically and horizontally, which brings serious challenges to static correction. To solve the problem of static correction in such area, we should make good use of the relationship between first ar rivals of shot records and loess thickness. Laterally based on different loess thicknesses, first arrivals of shot records picked up throughout the whole survey is divided into several first arrival datasets. Then vertically each lateral sub dataset is divided into several sub layers, on the basis that low velocity layer （LVL）, loess overburdened appears as a continuous medium. Then an initial near surface velocity model suitable for mountainous areas with loess covered is built to constrain tomographic inversion by both lateral and horizontal velocity field division. Therefore, the inverted velocity model is more reliable, which leads to more precise statics calculation. This approach is employed in practi cal 3D and crooked line seismic data processing in loess area, and it provides a good solution to statics problem in areas with thick losses covered. The evident improvement of imaging quality proves that this method can be applied widely in such situation.