~.ttsprd File
A ~.ttsprd file is produced by every run of mloc. It is written by the module mloc_ttsprd.f90. Every phase that is present in the dataset at least twice is processed to obtain the robust estimate of its spread and the baseline offset relative to the theoretical travel-time model. These values can be read into subsequent runs with the command tfil. These values are used in the windowing algorithm (command wind) to provide a procedure for rejecting gross outliers that is more sophisticated than a simple single-limit criterion. That is, the width of the window for each phase is based on the observed spread of arrivals for that phase and the window is offset from the theoretical travel-time model by the observed baseline offset to avoid bias.
The example is the full ~.ttsprd file for a cluster:
P 13088 1.592 2.953 Pn 5423 2.179 -0.591 S 816 4.513 4.471 PcP 173 2.713 2.716 PcS 26 3.016 3.038 PnS 35 2.900 -0.493 SS 88 5.300 2.092 PP 211 3.962 3.770 pP-P 62 1.227 0.246 SKiKP 8 2.677 2.927 SKKSac 9 1.293 0.968 ScS 28 3.592 3.265 SKSac 15 2.789 2.090 PS 7 7.499 5.060 Sn 821 2.637 -1.075 Sg 464 1.355 0.032 Pg 1575 1.346 0.129 Lg 410 3.294 2.857 pP 19 1.298 -1.911 PKiKP 105 1.595 2.329 Pdiff 168 1.062 3.366 PKKPdf 11 3.327 1.861 PKKPbc 19 1.427 3.760 PKPdf 18 1.689 4.479 PKPab 6 1.756 -2.155 sP-P 28 1.220 -0.301 sP 22 1.220 -0.499 ScP 12 2.873 2.034 sS 21 5.003 4.181 P'P'df 8 6.270 -2.066 PKKPab 15 4.855 0.831
The columns are:
- Phase
- Number of samples
- Spread
- Baseline offset
Notice that the baseline offset of the P phase is significant: almost +3 seconds relative to the ak135 global mode. On the other hand, the phases Pg and Sg have quite small baseline offsets, because the hypocentroid of the cluster has been determined mainly with these phases (direct calibration), using a local velocity model that has been refined to match the observed data. In calibrated relocation analyses it is very common to observe offsets of this magnitude for teleseismic branches.