search

GeoNet / help

An issues repo for technical help questions.
6 stars 3 forks source link

Sc3 event XML to FDSN event XML: Units of amplitude measurements are missing #60

Open calum-chamberlain opened 5 years ago

calum-chamberlain commented 5 years ago

I wondered what units the amplitude measurements in GeoNet QuakeML's are in? QuakeML spec suggests these generic amplitudes should be stored in SI units, but they encourage the used of the unit field but it doesn't look like this is populated, and the amplitudes do not look like SI values. For example, this M 2.9 event has ML amplitudes > 1.0, where ML amplitudes are usually displacement (in m for SI units, 1.6m displacement at a decent distance for a 2.9 seems high). e.g. (copied from the QuakeML):

<amplitude publicID="smi:nz.org.geonet/Amplitude/20190704101041.098734.32827">
  <type>ML</type>
  <genericAmplitude>
    <value>1.620822695</value>
    <lowerUncertainty>0.2354822821</lowerUncertainty>
    <upperUncertainty>0.2354822821</upperUncertainty>
  </genericAmplitude>
  <timeWindow>
    <reference>2019-07-04T10:09:44.39313Z</reference>
    <begin>-14.999999</begin>
    <end>48.81</end>
  </timeWindow>
  <pickID>
    smi:nz.org.geonet/20190704.100934.39-AIC-NZ.WHZ.10.HHZ
  </pickID>
  <waveformID networkCode="NZ" stationCode="WHZ" locationCode="10" channelCode="HH"/>
  <creationInfo>
    <agencyID>WEL(GNS_Primary)</agencyID>
    <author>scamp@kseqp02.geonet.org.nz</author>
    <creationTime>2019-07-04T10:10:41.098811Z</creationTime>
  </creationInfo>
</amplitude>

Any chance the units field of the amplitudes can be populated please?

salichon commented 5 years ago

Hello @calum-chamberlain Please refer to https://docs.gempa.de/seiscomp3/current/apps/global_mlv.html#global-mlv the input units are mm from seiscomp3 at last then into the event quakeml

(actually i think it s rather mm.s-1 after filtering since measured over the trace of velocimeters)

... I guess in a ideal world we would be all in SI (and even no imperial units .. ) ;) *but then the usage the history and the codes..)

Cheers Jerome

calum-chamberlain commented 5 years ago

Hi Jerome, thanks for that - any chance the units field of the Amplitude measurement in quakeml can be populated so that users don't have to dig through Seiscomp docs?

Is the QuakeML generated by Seiscomp, or does that just generate an SC3ML file that you guys convert?

salichon commented 5 years ago

Kia Ora @calum-chamberlain I would deem that the unit may be deliberately not informed in general since amplitude measurements may span a fair amount of scale depending of the magnitude/instrument used. We re are using the provided conversion sheet xslt that you can find under for instance: https://github.com/SeisComP3/seiscomp3/blob/master/src/trunk/libs/xml/0.11/quakeml_1.2__sc3ml_0.11.xsl But yeah i think this coudl be done :).

you may ask this type of question to Quakeml/seisocmp3 for more details See https://quake.ethz.ch/quakeml/ there I think Fabian Euchner woudl provide you with more details thans me :). Cheers Jerome

salichon commented 5 years ago

having had a look at the qml schemes:

genericAmplitude Measured amplitude value for the givenwaveformID. Note that this attribute can describe different physical quantities, depending on the type and category of the amplitude. These can be, e.g.,displacement, velocity, or a period. If the only amplitude information is a period, it has to specified here,not in theperiod attribute. The latter can be used if the amplitude measurement contains information on, e.g., displacement and an additional period. Since the physical quantity described by this attribute is not fixed, the unit of measurement cannot be defined in advance. However, the quantity has to be specified in SI base units. The enumeration given in attributeunit provides the most likely units that could be needed here. For clarity, using the optional unit attribute is highly encouraged

--->>> this is a request that could be made to seiscomp3 xml inputs perhaps

calum-chamberlain commented 5 years ago

Yeah, I read that too. Do you want to suggest it to Seiscomp, or shall I?

salichon commented 5 years ago

I think it s always good if it s coming from you ha! - i ll back your request up ;)

calum-chamberlain commented 5 years ago

:+1: Will do.

calum-chamberlain commented 5 years ago

Issue opened over at Seiscomp - regarding #58 as well, I got feedback from Fabian that the methodID field can be used to provide links or DOIs to documentation on how fields are populated/measurements taken. I think this would be really helpful to provide a consistent and documented catalogue.

salichon commented 5 years ago

I ve made a question on their forum: https://forum.seiscomp3.org/t/sc3-xml-and-amplitude-units/1518

calum-chamberlain commented 5 years ago

See comment on Seiscomp 243. Looks like the unit field should be populated if the seiscomp event conversion tool is used. I guess this comes back here then?

salichon commented 5 years ago

yeap @calum-chamberlain :) ! The sc3xml to quakeml conversion is made there https://github.com/GeoNet/seiscompml-to-quakeml - i ll enquire - exploring the built in converter to compare FYI @nbalfour

salichon commented 5 years ago

@calum-chamberlain @nbalfour here is a difference between the built in quakeml converter and the Geonet conversion
Sc3 built-in conversion: ` ...

0.001620822695

...

ML 
    **<unit>m</unit>** `

Geonet quakeml conversion `

1.620822695... 
    <timeWindow>

`

So

I think there could be room for improvement :)

salichon commented 4 years ago

@JonoHanson @danieldooley @carolinelittle Kia Ora Morena about quakeml units (and read above discussion):

"Looks like the unit field should be populated if the seiscomp event conversion tool is used."

I reckon there is an opportunity here to improve the on-the-fly conversion of the sc3xml to quakeml event files available through the GeoNet FDSN webservice. and possibly also changing the way the GeoNet process is converting the events sc3xml to quakeml.

I let you check what s the most relevant :)

Cheers Jerome

salichon commented 4 years ago

FYI @JonoHanson just to keep this in mind (wrt FDSNws reference for a paper about magnitude - refering to Geonet station amplitude data source)

salichon commented 3 years ago

FYI @sue-h-gns ..Archeogit_ticketing for improvement in the fdsn event service https://github.com/GeoNet/tickets/issues/5259#issuecomment-567737676

salichon commented 3 years ago

As a note for later 2021 in the archive:

  • Quakeml Geonet provided: ML Amplitude Unit is millimeter
  • Sc3xml (original source): ML Amplitude is millimeter

--> No conversion of the amplitude into SI for any of the format

--> Document the XML fields ?
--> No Unit description is provided

salichon commented 2 years ago

UPDATE 2022: Can be worked out with other FDSN improvements when work will be planned

salichon commented 1 week ago