PGA value

03-Unexpected behaviour/errors
Post Reply
arasch
Posts: 3
Joined: 04 Apr 2019, 15:56

PGA value

Post by arasch » 04 Apr 2019, 15:59

Why the maximum PGA values differ from the values seimosignal gives?

huffte
Posts: 840
Joined: 22 Jul 2011, 10:19
Location: Cookeville, Tennessee, USA
Contact:

Re: PGA value

Post by huffte » 05 Apr 2019, 00:56

I am not sure what you mean arasch. If I copy and past data from SeismoSignal (or SeismoSpect) to Excel and manually determine the maximum acceleration, I get the same value as is reported by SeismoSignal in the ground motion parameters. Of course, as it should be, SeismoSignal reports the maximum absolute vale. The reported PGA in the ground motion parameters is always positive, but may actually be the negative of the minimum acceleration in the accelerogram if that is greater than the maximum positive. So, for example, for the Chi-Chi record provided with SeismoSignal, the reported PGA is 0.361. The maximum acceleration in the record is 0.181. The minimum acceleration in the record is -0.361. So, SeismoSignal is reporting the correct value. Do you have a particular record for which this does not hold true.
Tim Huff

arasch
Posts: 3
Joined: 04 Apr 2019, 15:56

Re: PGA value

Post by arasch » 05 Apr 2019, 02:02

Hey Tim
The point is the PGA value which SeismoSignal gives differs from the value which SeismoSpect gives. Try with at2 files in both software.

huffte
Posts: 840
Joined: 22 Jul 2011, 10:19
Location: Cookeville, Tennessee, USA
Contact:

Re: PGA value

Post by huffte » 06 Apr 2019, 11:42

I get exactly the same PGA in both programs - SeismoSignal and SeismoSpect. Do you have a particular accelerogram for which the discrepancy exists? Or does the discrepancy occur every time on your machine and installation?

Even with PEER .AT2 files, I get exactly the same PGA, PGV, and PGD values with both programs, even if I specify 9 decimal places in the settings.

Do you have a record for which this is not true? If so, it would be interesting to find out what is happening with the record.
Tim Huff

arasch
Posts: 3
Joined: 04 Apr 2019, 15:56

Re: PGA value

Post by arasch » 08 Apr 2019, 11:09

With these files, which you can get from the Peer, SeismoSignal gives the PGA as 0.1802857 and 0.1588854 but SeismoSpect gives like 0.18295 and 0.15575. The file names are RSN15_KERN_TAF021.AT2 and RSN15_KERN_TAF111.AT2. And it happens for couples of other Peer data too.

huffte
Posts: 840
Joined: 22 Jul 2011, 10:19
Location: Cookeville, Tennessee, USA
Contact:

Re: PGA value

Post by huffte » 08 Apr 2019, 13:38

For the Kern County (RSN 15 PEER) record you mention, I get identical values in SeismoSignal and SeismoSpect for PGA, PGV, and PGD, . . .

Could it be that in one of the programs you are filtering/correcting the record and in the other program you are analyzing the raw, unfiltered/uncorrected accelerogram? Check the settings to see if this is the case. This will help to determine if there is, in fact, an issue. The ground motion parameters do indeed change when baseline correction and filtering is applied.

In general, PEER records have already been properly baseline adjusted and filtered and should not be re-corrected and re-filtered. Open the accelerogram with no baseline correction and filtering applied (see the 'Baseline Correction and Filtering' tab after opening the file). If the velocity and displacement return to zero at the end of the record, and the histories appear realistic, then the record has already been filtered and you should not re-apply the correction and filtering. This is specified in the Settings for SeismoSpect, on the 'Baseline and Correction' tab for SeismoSignal.
Tim Huff

Post Reply