数字化台网的近震震相自动识别

AUTOMATIC RECOGNITION OF SEISMIC PHASES OF REGIONAL EARTHQUAKE

首次提出了数字化台网对于近震震相自动识别的“波形变化值增长”算法和“无后续震相”判据 .前者根据波形变化值来描述近震波形的变化 ,用该值的增长捕捉地震波振幅或频率的突变 ;后者在识别到初动后考察后续波形的变化情况 ,如果没有其它震相 ,初动识别无效 .该判据使算法在初动震相振幅过小识别不清时 ,不会误将后续振幅较大的震相识别为初动震相 .该方法经实例检验表明精度符合地震速报的要求 。

In order to recognize the seismic phases of a regional earthquake automatically, the algorithm of increasing waveform variation value and the criterion of no subsequent phase are put forward firstly. The former is to describe the waveform variation of a regional earthquake from its waveform variation value, for capturing the sudden variations of amplitude or frequency of seismic wave by using increase of its value. The latter is used to observe the variations of subsequent waveforms after determining the onset. If there are no other phases in the subsequent waveform, the recognition of onset is invalid.Using this criterion, it will not recognize subsequent phases with bigger amplitudes as the onset phase when the amplitude of the onset phase is too small to be recognized clearly.It shows by the example tests that the accuracy of the method meets the need of the prompt report of earthquake.The relevant programs are compiled.