【摘要】 本文提出了一种在时间域用Wiener滤波方法提取台站接收函数的方法 ,用远震P波波形的垂直分量为输入 ,接收函数作为滤波因子 ,远震P波波形的径向和切向分量作为期望输出 ,通过期望输出与实际输出的均方误差达极小 ,来提取接收函数。接收函数的计算可归结为Toeplitz方程的求解 ,可以采用Levinson递推算法。Toeplitz方程的非奇异性保证了Wiener滤波反褶积方法的稳定性。合成地震图与观测地震图的检验表明 ,用Wiener滤波方法测定台站接收函数是一种有效的时间域反褶积方法。更多还原

【Abstract】 Wiener filtering is provided to estimate receiver function in time-domain. With the vertical component of 3-component teleseismic P waveform as the input of Wiener filter, receiver function as the filter response, and radial and tangential components as the expected outputs, receiver function is estimated by minimizing the error between expected and actual outputs. Receiver function can be obtained by solving Toeplitz equation through Levinson algorithm. The non-singularity of Toeplitz equation keeps the stability of Wiener deconvolution. Both synthetic and observational seismogram checks show that Wiener deconvolution is an effective time-domain method to estimate receiver function from teleseismic P waveform.更多还原