给出了一种三分向地震台阵数据偏振分析的技术,并应用于挪威南部NORESS台阵三分向拾震器记录到的一大组区域事件。仔细分析了区域震相P_n、S_n和乙的偏振性质。该技术基于Flinn(1965)最先提出的一种时域算法,通过解协方差矩阵的本征值问题,在滑动时间窗内计算偏振椭圓,由运动椭圆导出表征质点运动的各种属性。通过对不同拾震器的协方差矩阵求平均,把上述方法扩展到了由多个三分向拾震器组成的台阵情况,这样,当噪声和地方散射效应不相关时,可使方差估计减小1/M(M为拾震器数)。此方法的一个重要特点是穿过台阵相干波前的相速度不需要有很高精度。数据分析的重要结果为:在整个短周期频带内,很好地确定了P_n和S_n的偏振;由P_n和L_g波动获得了源方位的估计及S_n和L_g波独特的偏振属性使得在多数情形下能把这些震相区别出来。 A polarization analysis of trishase seismic array data is given and applied to a large group of regional events recorded by NORESS array three-point pickup in southern Norway. The phase properties P n, S n and the polarization properties of B are carefully analyzed. The technique is based on the first time-domain algorithm proposed by Flinn (1965). By solving the eigenvalue problem of the covariance matrix, the ellipse is calculated in the sliding time window, and the properties of particle motions are derived by the motion ellipse. By averaging the covariance matrices of the different pick-ups, the above approach is extended to the case of a matrix consisting of multiple trishars, so that when the noise is not correlated with local scattering effects, the variance estimate can be reduced Small 1 / M (M is the number of pickups). An important feature of this method is that the phase velocity across the array’s coherent wavefronts does not need to be highly accurate. An important result of the data analysis is that the polarizations of P_n and S_n are well determined over the entire short-period band; the estimation of the source orientation and the unique polarization properties of the S_n and L_g waves are obtained from the fluctuations of P_n and L_g, so that in most cases Can distinguish these shock phase out.