从小湾水库台网2005—2008年记录的众多地震中,挑选出最少被4个台站接收到的高质量地震数780个,一共5 230条P波和4 883条S波到时资料。利用Kissling方法得到了小湾水库库区最小一维P波和S波速度模型以及台站校正值。反演后的最小一维P波速度模型走时均方根残差从0.81s减少到0.12s,数据方差从1.64s2减少到0.04s2;地震震源深度比原来增加大约1公里,震源分布更加集中;不同台站的校正值差异表征了小湾库区速度结构存在横向不均匀性。最后利用得到的最小一维速度模型和台站校正值进行重定位,结果地震的走时均方根残差明显减少,表明得到的最小一维速度模型可信度较高。
Using 5 230 Parrivals and 4 883 Sarrivals from 780 high quality events,which can be recorded at least by 4stations,selected from events from the Xiaowan reservoir seismic network from 2005 to 2008.The minimum 1Dvelocity model including P wave model,S wave model and station correction are obtained with the Kissling method.After inversion,the RMS of the minimum 1Dvelocities decrease from 0.81 sto 0.12 s,data variances decrease from 1.64s2 to 0.04s2,and depths of earthquakes increase about 1km with source distribution is more concentrated.Different station correction values represent lateral heterogeneity of surface velocity structure in the Xiaowan reservoir area.Finally,we use the minimum 1Dvelocity and station correction to relocate the earthquakes,and the RMS reduce abviously,which shows the high quality of the 1D minimum velocity.