MRZ - Strong Motion Vs BB - comparison enquiry

salichon commented 2 years ago

MRZ HN/HH waveforms derived from instruments have differences:

could be polarities for older events
can be amplitudes for events 2020

[x] List of events of interest polarity check evid 2498995 and 2376455 amplitudes 2020p802823 2020p464185 other possible 2020p434252 2020p877108 2020p947812 2020p972758 other events were not recording relevant data. (source too far) checked: https://www.geonet.org.nz/earthquake/technical/2019p927023
[ ] Isolate if new issues arise and are confirmed
[ ] Propose solution / fix /Actions
[ ] Document

salichon commented 11 months ago

Task/Check list for investigation

[x] interpret the issue reported ( e.g. Amplitude Diff between BB and SM waveforms)
inspecting figures oper events per staiton as provided par jessie: in gm_compare-1\M_5\HN_HH\wf_figs Checked for events of interest (lot of noise and few good data)

[x] isolate periods of questioning for the period mentioned in that ticket

[x] Check event 2019p927023 waveform parameters (MRZ ~300 km away) https://strongmotion.geonet.org.nz/#/event/2019p927023

> 
>    <?xml version="1.0" encoding="UTF-8" standalone="yes"?>         <?xml version="1.0" encoding="UTF-8" standalone="yes"?>
> <!DOCTYPE earthquake SYSTEM "stationlist.dtd">                  <!DOCTYPE earthquake SYSTEM "stationlist.dtd">
> <stationlist created="" xmlns="ch.ethz.sed.shakemap.usgs.xml"   <stationlist created="" xmlns="ch.ethz.sed.shakemap.usgs.xml"
>   <station code="MRZ" name="MRZ" netid="NZ" insttype="CMG-3ES |   <station code="MRZ" name="MRZ" netid="NZ" insttype="FBA-ES-
>     <comp name="10.HHE">                                      |     <comp name="21.HNE">
>       <acc value="0.0336222543" flag="0"/>                    |       <acc value="0.0279376229" flag="0"/>
>       <vel value="0.0171557611" flag="0"/>                    |       <vel value="0.0118713551" flag="0"/>
>       <psa03 value="0.1252890259" flag="0"/>                  |       <psa03 value="0.0579332028" flag="0"/>
>       <psa10 value="0.0072364425" flag="0"/>                  |       <psa10 value="0.0078426039" flag="0"/>
>       <psa30 value="0.0011547390" flag="0"/>                  |       <psa30 value="0.0018025627" flag="0"/>
>     </comp>                                                         </comp>
>     <comp name="10.HHN">                                      |     <comp name="21.HNN">
>       <acc value="0.0261324049" flag="0"/>                    |       <acc value="0.0400683587" flag="0"/>
>       <vel value="0.0121421276" flag="0"/>                    |       <vel value="0.0174330617" flag="0"/>
>       <psa03 value="0.0735810139" flag="0"/>                  |       <psa03 value="0.1192592514" flag="0"/>
>       <psa10 value="0.0084328539" flag="0"/>                  |       <psa10 value="0.0078407303" flag="0"/>
>       <psa30 value="0.0008719387" flag="0"/>                  |       <psa30 value="0.0014621644" flag="0"/>
>     </comp>                                                         </comp>
>     <comp name="10.HHZ">                                      |     <comp name="21.HNZ">
>       <acc value="0.0127150652" flag="0"/>                    |       <acc value="0.0173768548" flag="0"/>
>       <vel value="0.0075368465" flag="0"/>                    |       <vel value="0.0387199230" flag="0"/>
>       <psa03 value="0.0368886509" flag="0"/>                  |       <psa03 value="0.0362138030" flag="0"/>
>       <psa10 value="0.0059454713" flag="0"/>                  |       <psa10 value="0.0060438004" flag="0"/>
>       <psa30 value="0.0012763991" flag="0"/>                  |       <psa30 value="0.0018354591" flag="0"/>
>     </comp>                                                         </comp>
>   </station>                                                      </station>
> </stationlist>                                                  </stationlist>

[x] Check polarities for cusp events as in MRZ | 20 | reversed | 2003-07-31T00:00:00Z to 2011-04-13T00:00:00Z OLD
- HNE is reversed WRT HHE ALL components are reversed - amplitudes are good. See https://www.geonet.org.nz/earthquake/2376455 https://www.geonet.org.nz/earthquake/2498995 example.

[x] Check station change history SM and BB if available
- [x] Isolate periods of change and potential issue with regards to instrumentation https://github.com/GeoNet/tickets/issues/4669 Reversed Strong motion https://github.com/GeoNet/delta/blob/main/install/polarities.csv#L80

[x] Compare Raw waveforms Horizontals and Vertical in velocity and acceleration for given reference events and periods
[x] Compare waveforms removed form instrument reposnse on Horizontals and Vertical in velocity and acceleration for given reference events and periods
[x] Check if all components are consistent with each other (Horizontal Vs Vertical)
[x] Check PGA values for events of reference over time periods for both instruments
[ ] Check additional waveform parameters if required

mnaguit commented 11 months ago

2020 M4.7 Levin (2020p802823)

PGA estimates

PGA (HNE) = 0.019081, PGA (HNN) = 0.028269, PGA (HNZ) = 0.012876 m/s/s 
PGA (HNE) = 0.001946, PGA (HNN) = 0.002883, PGA (HNZ) = 0.001313 g 
PGA (HHE) = -0.019427, PGA (HHN) = 0.011166, PGA (HHZ) = 0.009958 m/s/s 
PGA (HHE) = -0.001981, PGA (HHN) = 0.001139, PGA (HHZ) = 0.001015 g

Acc subplots
Vel subplots

mnaguit commented 11 months ago

2020 M4.6 Whakatane (2020p464185)

PGA estimates

PGA (HNE) = -0.003669, PGA (HNN) = 0.003887, PGA (HNZ) = 0.001751 m/s/s 
PGA (HNE) = -0.000374, PGA (HNN) = 0.000396, PGA (HNZ) = 0.000179 g 
PGA (HHE) = 0.003156, PGA (HHN) = -0.002764, PGA (HHZ) = 0.001142 m/s/s 
PGA (HHE) = 0.000322, PGA (HHN) = -0.000282, PGA (HHZ) = 0.000116 g

Acc subplots
Vel subplots

mnaguit commented 11 months ago

Discussion summary (as of Oct 2, 2023)

polarity issue confirmed, all components may be reversed from 2003 to 2011 (this is captured in polarities.csv, which is good!)
waveforms corrected to ACC & VEL look comparable, along with peak motions using two events in 2020 (but cannot generalize based on 2 events!)
cannot finalize an outcome just depending on selected events/scenarios
- what happens for longer epicentral distances?
- how both sensors perform over a longer range of time, not just event-based?
- wider discussion on polarities may include like delta vs xml, or the extent by which to 'red flag' stations like this other than in polarities.csv which may not probably give a heads up to users not familiar with delta)
this assessment is linked with a data user query. the events were selected by the data user.
most selected events have short epicentral distances (lower than 60km) : for reference, the data user files are in https://www.dropbox.com/s/30ix5mn28a80hnb/gm_compare.zip?dl=0

BOTTOMLINE:

quality control is confined on user's assessment and choice of events: this is true not only for MRZ but it looks like for all stations in https://github.com/GeoNet/help/issues/91
quality control checklist (as identified by @salichon ) provides a good framework for assessment

WHAT'S NEXT:

replicate this assessment in other 'problem' stations on the list?
@Thomas-Benson to do further digging to check how we can refine this assessment

salichon commented 11 months ago

@Thomas-Benson @mnaguit Even better about the "polarities.csv" of MRZ these polarities are well accounted for (we may need to review the definition of polarities (i forgot that bit see below from Mark)

when we know 100% that the channel is reversed, this is related to hardware. e.g. an STS-2 connected to a Q4120 will be reversed. The polarities is for when we don't expect it to be reversed but for some reason it is (wrong connection, bad cable etc).

For the period 2003-2011: The SM sensor has an Azimuth to 0 in delta The polarity change is accounted for in the "polarities.csv"
The final results is https://service.geonet.org.nz/fdsnws/station/1/query?network=NZ&station=MRZ&level=channel&format=text&channel=HN? and gets the right components orientation

Thanks @Thomas-Benson for pointing that out ! well done - a No PB for that period.

Thomas-Benson commented 11 months ago

This doesn't really contribute that much, but I was testing some scripts for PSA calculations and plotting and wanted to put something here. Example with 2019p927023 which Jerome has done above:

Testing with the March 2021 7.2 East Cape quake 2021p169083 at distance 524Km for a slighly larger epicentral distance. Acceleration waveforms look OK: Untitled PSA plots:

I get good PSA values for both the Strong motion and broadband sensors, which also matches up fairly well with the values available on the strong motion page for the event	SM	BB
HNE PSA0.3: PSA 0.151885	HHE PSA0.3: PSA 0.166159	0.152314
HNE PSA1.0: PSA 0.03727	HHE PSA1.0: PSA 0.041462	0.0391874
HNE PSA3.0: PSA 0.006473	HHE PSA3.0: PSA 0.006867	0.00626706
HNN PSA0.3: PSA 0.162689	HHN PSA0.3: PSA 0.172862	0.17351
HNN PSA1.0: PSA 0.045646	HHN PSA1.0: PSA 0.039204	0.0474634
HNN PSA3.0: PSA 0.006985	HHN PSA3.0: PSA 0.006272	0.00659691
HNZ PSA0.3: PSA 0.059772	HHZ PSA0.3: PSA 0.064563	0.0629257
HNZ PSA1.0: PSA 0.023733	HHZ PSA1.0: PSA 0.022504	0.0272351
HNZ PSA3.0: PSA 0.004603	HHZ PSA3.0: PSA 0.004533	0.00459017

mnaguit commented 10 months ago

As of 10 October 2023

last item to check: SM vs BB spectral acc ratios being too symmetric (we can close this ticket after this last check! 👍 )
to do next: replicate these steps, choose another station (https://github.com/GeoNet/help/issues/91#issuecomment-1068699372)
- next on the list: MLZ
- see relevant figures from the end user: https://www.dropbox.com/s/30ix5mn28a80hnb/gm_compare.zip?dl=0
final verdict in terms of data quality
- SM vs BB channel components have comparable processed peak motions
- wrt selected events, both sensors are working as expected (noting that the performance of these sensors at MRZ was checked at certain time periods only)

GeoNet / help

MRZ - Strong Motion Vs BB - comparison enquiry #100