Closed alebrovelli closed 5 years ago
I found another bug, somehow related to the one above. I am pretty sure the s-pick times are not correctly "sent" to hypo71. In the output of hypo71 I find values that are different from the one picked manually. The p-picks instead are correct. Let me know if/how I can help fixing these bugs, which are quite critical, as they result in wrong localizations.
A side note: I made a comparison between the standalone Hypo71PC and the version that comes with swarm, and after manually correcting the s-times the results are consistent.
Thank you regards Alessandro
I have added below prior to the hypo71.calculateHypo71 step in Hypo71Manager.calculate: controlCard.setPOS(SwarmConfig.getInstance().velocityRatio);
With regards to the S-arrival time, perhaps I misinterpreted the Hypo71 documentation on this:
My original interpretation was I could not just take the S-arrival seconds in case it crossed into the next minute, so I took (P-arrival seconds)+(P-arrival time - S-arrival time in seconds). Re-reading the "**" note above, it sounds to me like I need to apply 60 seconds to SEC and S if the S-arrival crosses the minute, but it is not clear to me. What is your interpretation?
2.8.13 is out with the Vp/Vs change for Hypo71 (plus a fix for bug introduced in 2.8.12). Please try it out.
I will await for your comments before addressing the second issue you mention. Actually, if you can make a second ticket for it that would be great. This way we can close this one for the Vp/Vs.
ok! thank you Diana
Dear Diana, from a few tests I made, I realized that hypo71 uses the same Vp/Vs in all runs (default value 1.78, "POS" variable, 5th entry in the ControlCard array), regardless of what is set in the swarm config file. I think this value should be instead taken from the "velocityRatio" setting in Swarm.config.
Kind regards Alessandro Brovelli