We would like to compare the result we obtained in this post to the same set of events but by using the adaptive stacking TCAS approach.



We plot the raw data, initial alignment by removing the predicted arrival travel-time, and the final alignment after the cross-correlation. We allow the final time-shift to be within +/- 2 seconds.



The residuals contained in the output are consistent considering the same event path. Furthermore they are in relative agreement with the residuals obtained by manual picking with the mean removed from the observed minus predicted values.



Firstly the result shows that tcas is robust as it gives a consistent output for equal path. Secondly it shows that it reflects our manual work by showing similar perturbation pattern in the stations. Therefore there is the feasibility to use tcas to automate pickings using more sources.