Jens - 
     Again, thank you. I'll get on it. 
Tim 

-----Original Message-----
From: Jens Havskov [mailto:[email protected]] On Behalf Of Jens Havskov
Sent: Wednesday, July 29, 2015 01:48 AM
To: Sickbert, Tim <[email protected]>
Cc: [email protected]; Yihun, Manahloh <[email protected]>; Abdel 
Salam, Mohamed <[email protected]>; Campbell, Nathan 
<[email protected]>
Subject: Re: [seisan] SEISAN, continuous waveforms

tom

On Wed, 29 Jul 2015, Sickbert, Tim wrote:

> 
> Greetings.
>
>     We are working to set up an earthquake seismology research lab at 
> Oklahoma State University. A brief inspection of available data 
> showing a rapid increase in seismic activity in north-central Oklahoma 
> since about 2009 would show our motivation (figure below). We have ten 
> broadband triaxial seismographs (Guralp
> CMG-6TDs) rather densely deployed (long axis of the research area is 
> about 100 km) in north-central Oklahoma.  Our instruments are
> offline: they are logging data and we have to visit the stations to 
> collect them. In addition we are using data from the USGS and Oklahoma 
> Geological Survey (OGS) available through IRIS. Thus we want to 
> analyze raw waveform data from about twenty seismic stations to 
> locate, and analyze all earthquakes possible.
>
> 
>
>     We had been using Seismic Handler (SHM) to pick onsets and analyze 
> events but we are not entirely satisfied with it. OGS and others have 
> recommended SEISAN. I have worked through most of the tutorial with 
> the embedded TEST data, and some of the training
> (seitrain.pdf) with the additional data. I very much like what SEISAN 
> can do. The documentation, however, does not seem to me to directly 
> address our application. Ideally, we want to simultaneously display 
> fractions of 24-hour-long continuous waveform files from about twenty 
> stations, zoom in on shorter time segments, pick onsets, locate and 
> analyze events, determine magnitudes and maybe moment tensors, and 
> save the appropriate S-files for cataloging and further analyses.
>
> 
>
>     I have been able to load individual miniSEED files into MULPLT. 
> However, I have not yet been able to load more than one of our 
> continuous waveform files at a time. My reading of ?seisan.pdf,? 
> ?seisan-tutorial.pdf,? and ?seatrain.pdf? lead me to think that I have 
> to build BUD or SeisComP archives for SEISAN to access the data. The 
> SEISAN documentation regarding these archives also seems to imply that 
> they are event-oriented, unlike our situation where we want to 
> manually, visually, perform the initial analyses of the continuous 
> waveform files.
>
> 
>
>     Thus I have a few questions:
> 
> 1)      Is SEISAN appropriate for what we are trying to do?
>

  you should be able to do all of what you mention above

> 2)      Do the documents I have clearly present the information I need 
> to set up the database(s) for our application?

it is there, but maybe a bit hidden here and there
> 
> 3)      Do I need to archive our data into BUD or SeisComP to be able 
> to use it?
>

that would be the easiest way. the only probelm is making the structure and the 
day files. chad trabant has a program, msrouter, to do that, see seisan manual 
5.3.1. we have a version of the program which also makes SeisCompP structure. 
see manual 2.2.3 for a description on how to use and set up the archive.
once the data is archived, you can display all or some channels, see examples 
1.8.5 and 1.8.6 in the training document.

you can now mnaully go through the data, window by window, archive the 
intesting sections as events and thereby create the data base.

you can also autodetect the events from the archive using program condet, see 
exercise 10.1 in training document.

it is possible to always work directly with the archive and the s-files only 
have a reference to the archive. we recommend however to extract out the 
waveform files corresponding to the events and put them in the seisan 
structure. this can be done automatically with mulplt when registring the 
event. for futire work, it much easier not having always to deal with the large 
archive.


  > 4)      Do you have any other or additional recommendations on how > we 
should manage our data to make it available to SEISAN
> (especially MULPLT)?

see above. not sure there is anything special to say about mulplt. if you have 
a lot of data you might want to autopic the amplitudes for Ml and spectral 
level for Mw, use autopic for that.

jens


  > 
> 
> Thanks!
>
> 
> 
> Tim
>
> 
>
> 
>
> 
>
> 
> 
> [IMAGE]
>
> 
> 
> Tim Sickbert
> 
> Devon Labs Coordinator
> 
> Boone Pickens School of Geology
> 
> 105 Noble Research Center
> 
> Oklahoma State University
> 
> Stillwater OK 74078
>
> 
> 
> (405) 744-6358
>
> 
> 
> [email protected]
>
> 
> 
> 
>

_______________________________________________
seisan mailing list
[email protected]
http://mailman.uib.no/listinfo/seisan

Reply via email to