zero amplitude spectrum

[Evanna]

Hello,
The signals from my simometers are jerky, so I'm having trouble calculating the H/V ratio of certain signals per Julian day. In Python I've set up a code to merge all the data onto a single trace, which is more practical for displaying the signals on Geopsy. Missing data is replaced by a 0. I'd like to do my H/V analysis on all the parts of the signal that aren't equal to 0. But I can't figure out how to do it. I tried using the bad sample tab but it didn't work. In fact, I don't know if there's any processing on geopsy before doing the H/V analysis or not.
Best regards,
Evanna

[admin]

Hi,

The best is keep the original signals as they are without merging them. Make sure that all individual portion of signal have the same sampling frequency, compatible starting times (differences are multiples of the sampling period) and the same name and the same coordinates (the portions that belong to one station). The gaps between the available ranges are automatically recognized and geopsy will create time windows only when the signal is available for the three components at the same time.

Merging and setting 0 is not the most efficient method. You have to add gaps manually (right click menu). This is tedious and error-prone.

Best regards,

Marc

[Evanna]

Hello,
First of all, thank you for your reply!
In fact, some Julian days are only recorded over a few hours (or even a few minutes... Certain technical problems with the seismometers persist...). However, I would still like to use this data. I'm sharing with you a screenshot of Julian Day 162 as an example of an attempt at H/V processing. The logs show that the samplings are incompatible, yet each portion of the signal has a sampling frequency of 100 Hz...
The names are identical, as are the coordinates...
I'm not sure how to identify this error.

Thanks,
Evanna

geopsy.PNG

(322.51 KiB) Not downloaded yet

download/file.php?mode=view&id=563

[admin]

The resolution of the snapshot is too low to clearly distinguish numbers for the starting times. However, I can read that they differ by a few ms which is lower than the sampling period. For H/V processing you can safely round them to the sampling period, because the phase is not exploited. That's strange that even the E, N and Z components do not have the same starting time.

You can run the following script with menu Edit/Set Header:

https://www.geopsy.org/wiki/index.php/G ... ing_period

[Evanna]

Hello,
is there anything I can do to get the startTime property recognized? applying the link code snippet, my log says :
[Header] Uncaught exception at line 6:
[Header] TypeError: Cannot read property 'startTime' of undefined

Thanks
Best regards,
Evanna

[admin]

How many signals do you have in the viewer? Thanks to your post I discovered while answering on June 28th that there was a stupid condition (I have no idea why it was there) to prevent the creating of the object signals[] if there were less than 10 signals. This is now fixed for future release. But meanwhile, unless you are willing to compile geopsy on your machine, you must have more than 10 signals to run the script. Add some fake signals or duplicate the signals to get at list ten rows in the table.