MQZ issue with broadband vs. accelerometer comparison

elidana commented 2 years ago

Jesse Hutchinson (University of Canterbury) has noted an issue with MQZ seismic data when comparing data from the broadband and strong motion sensors installed at the station.

His method calculates the difference of the natural log values of PGAStrong from PGABroadband instruments, for all M>5 events recorded by the station.

A plot he provided is showing a consistent discrepancy for all events recorded between ~2014 and ~2020.

pSA_diff_MQZ_datetime

Further investigation to understand the cause are needed

salichon commented 2 years ago

the graph is an indication though it does not help to troubleshoot really.

I encourage to have a look at the raw data (say using daily streams) as it should be done prior to any processing of some sort :) - especially when it is made with big batch (though tedious i know)

Would be helpful to have from Jesse:

A detailed step by step procedure of the processing of the data in order to frame the discrepancies and the actual source(s) -Or at least the earliest steps from :querying to deconvolving and using the data.
A list of the M5 earthqaukes would help to identify the most relevant periods

salichon commented 2 years ago

Summary of the PSA process above from what i understood

Input:

M5 event waveforms (time window length?)
Waveforms from high quality station (national sites)
Use in priority strong motion HN (200HZ) (then strong BN, and then weak HH 100Hz)

Process:

Remove instruments response
Compute Pseudo Spectral Accelerations spectra (PSA), From: 1 - Strong motion 2 - Weak motion
--> PSA methods for each data type to be described briefly

Tools and Sources:

Used Python packages to perform this computation
Query mechanisms ?

salichon commented 2 years ago

MQZ strong motion periods (delta/fdsnwebservice: https://service.geonet.org.nz/fdsnws/station/1/query?network=NZ&station=MQZ&level=channel&format=text&channel=?NZ)

NZ|MQZ|20| HNZ/BNZ |Accelerometer|427336.117780|1.000000|m/s2| 2009-08-06T00:00:01|2016-02-22T20:00:00**
NZ|MQZ|20|HNZ/BNZ|Accelerometer|427336.117780|1.000000|m/s2| 2016-02-24T01:02:00|2019-06-17T05:25:00**
NZ|MQZ|20|HNZ/BNZ|Accelerometer|427818.982500|1.000000|m/s2| 2019-06-17T23:20:00| NOW**

Periods of identified Strong Motion default in quality: 2016.054 till 2017.218 ( !!2017.219 is --> when Pb is solved on the ticket (to be detailed) )

Today: Raw archived data --> MQZ strong has "Stepped traces" ("long period Steps every couple of hours") as below :

Does this affect Computation on the event traces ? Are the data actually valid (over one single event for instance?) Any other issues identified with this installation over that time ? Can we explain the HHvsStrong observations with this ?

With regards to the comparison with the broadband HH It looks like the mount of the HH is old stable and performing since 2009:

MQZ|10|HHZ| Broadband Seismometer|838860800.000000|1.000000|m/s|100.000000|2009-08-06T04:00:00|

SquirrelKnight commented 2 years ago

The procedure is as follows: All stations for which an event arrival was detected and within a certain distance defined by a magnitude-distance scale are used to download waveforms. The time window extends from 10 seconds prior to the P arrival to the S arrival + the ds595 value * 1.2. The ds595 value has been modified from the original equations by Afshari and Stewart (2016) to include an additional term to capture more of the waveform coda for larger magnitude events.

After the raw mseed files are downloaded, they are then evaluated with a ground-motion classifier to determine whether the event meets certain quality criteria. High quality events are then pre-processed for intensity measures. In pre-processing, I first demean the data and pad the data for five seconds on either end. I then remove instrument sensitivity and differentiate any velocity data into acceleration. The data are then rotated to ZNE components and trimmed back to their original window length. Finally, the data are low-pass filtered at a corner frequency of 0.05 Hz.

The pre-processed data is used to compute intensity measures (PGA, PGV, pSAs, etc.) for horizontal, vertical, rotd100, and rotd50 components.

The list of M5 events used for MQZ are as follows (note that this analysis has only been performed between HN and HH channels): '2016p858069', '2016p858147', '2016p858156', '2016p858160','2016p858279', '2016p858848', '2016p858895', '2016p858951','2016p859051', '2016p859105', '2016p859270', '2016p859378','2016p859625', '2016p859628', '2016p859929', '2016p860053','2016p860215', '2016p860224', '2016p860287', '2016p860567','2016p860729', '2016p861632', '2019p809993', '2019p927023','2020p201743', '2020p475488', '2020p606562'

salichon commented 2 years ago

thanks for this description:

from the 27 events ID perspectives and the Period of time identified Strong Motion default in quality 2016.054 till 2017.218. It looks like 22 events may fall into the questionable period of the SM instrument issue.
the 5 others '2019p809993', '2019p927023','2020p201743', '2020p475488', '2020p606562' should have better data set with the strong motion

From the Broadband side of the process: MQZ broadband had a fairly stable sensor and datalogger since 2009 to now: Guralp Systems Ltd.,CMG-3ESPC,T35861,MQZ,10 ,2009-08-06T04:00:00Z,9999-01-01T00:00:00Z Kinemetrics,Q330/6,3386,Glencoe Station,,2009-08-06T00:00:01Z,9999-01-01T00:00:00Z

so let s focus on the most recent and 5 events first.

salichon commented 2 years ago

MQZ is located in Christchurch

https://www.geonet.org.nz/earthquake/technical/2019p809993 M5
https://www.geonet.org.nz/earthquake/technical/2019p927023 M5.4 are located to distance ~700/800km upper north island Strong motion tool does not provide SM data for both events. This is legitimate due to signal to noise ratio too low for relevant result afaik https://strongmotion.geonet.org.nz/#/event/2019p927023 https://strongmotion.geonet.org.nz/#/event/2019p809993

https://www.geonet.org.nz/earthquake/technical/2020p475488 M5.9 is located in Fjordland ~ ~400 KM from MQZ Strong motion tool does provide SM data https://strongmotion.geonet.org.nz/#/event/2020p475488
https://www.geonet.org.nz/earthquake/technical/2020p606562 M5.4 is located in Fjordland ~ ~400 KM from MQZ Strong motion tool does provide SM data https://strongmotion.geonet.org.nz/#/event/2020p606562
https://www.geonet.org.nz/earthquake/technical/2020p201743 M5.1 is located in Canterbury about 100km Strong motion tool does provide SM data https://strongmotion.geonet.org.nz/#/event/2020p201743

Out of 5, the 3 latest and closest have some reasonable data to compare with for a start

salichon commented 2 years ago

@SquirrelKnight In order to focus on the most relevant data sets: What are the certain distance criteria for using the data : "All stations for which an event arrival was detected and within a certain distance defined by a magnitude-distance scale are used to download waveforms." ? I find that the "certain distance" can be pretty loose for some distant events and a resulting acceleration can be pretty low for M5 > 500km distance
thanks :)

I guess also it would be interesting to get the raw PSA data derived from HH and HN ( as a text file) of the method to be able to compare the values for a least the events: 2020p475488, 2020p606562 , 2020p201743 cheers!

SquirrelKnight commented 2 years ago

Mag_dist_relatonship

Here is an image of the magnitude-distance relationship. As you can see, stations selected for lower magnitude data are selected from considerably smaller distances.

I can certainly provide the raw PSA data. Where would be a good place to upload it?

salichon commented 2 years ago

thanks a lot @SquirrelKnight
"I can certainly provide the raw PSA data. Where would be a good place to upload it?" here on git (as *.text) so that we some psa to compare :)

okay thanks for the distance relationship ! still for having MQZ used in far north 2019p809993 M5 and 2019p927023 M5.4 I reckon that the selection should be tighter the epicenter distance does not seem right.

cheers j

SquirrelKnight commented 2 years ago

As I mentioned above, I do compute IMs for detected arrivals as well. We will be providing a pared down version of the database that excises any IMs outside of this distance relationship, but I've made it an objective to supply as much data as possible! Looking at 2019p927023, the signal appears quite strong (these figures are both unfiltered with no instrument correction).

SquirrelKnight commented 2 years ago

I wanted to add that I have also produced similar plots, but in consideration of the hypocentral distance. Here is an example for MQZ: pSA_diff_MQZ_distance

salichon commented 2 years ago

MLZ Mavora Lakes is deemed providing some good results. MLZ is located close to Fjordland so should be relevant to compare to the 3 M>5 2020 events: 2020p475488, 2020p606562, 2020p201743
will use: 1- strong motion parameters from our operational tools 2- recompute parameters for both HN and HH data set from these stations ( with and without instrument response) this will provide sets of parameters to compare with.

salichon commented 2 years ago

--- Summary (strong motion parameters - reprocessed as in the strong motion tools - using response) 2020p201743 M5.1 is located in Canterbury about 100km of station MQZ

 <station code="MQZ" name="MQZ" netid="NZ" insttype="CMG-3ES |   <station code="MQZ" name="MQZ" netid="NZ" insttype="FBA-ES-
    <comp name="10.HHE">                                      |     <comp name="21.HNE">
      <acc value="0.0697800113" flag="0"/>                    |       <acc value="0.1903951915" flag="0"/>
      <vel value="0.0353069676" flag="0"/>                    |       <vel value="0.0591125614" flag="0"/>
      <psa03 value="0.1714820997" flag="0"/>                  |       <psa03 value="0.1674342303" flag="0"/>
      <psa10 value="0.0159153808" flag="0"/>                  |       <psa10 value="0.0176827828" flag="0"/>
      <psa30 value="0.0013179099" flag="0"/>                  |       <psa30 value="0.0015167207" flag="0"/>
    </comp>                                                         </comp>
    <comp name="10.HHN">                                      |     <comp name="21.HNN">
      <acc value="0.0680543813" flag="0"/>                    |       <acc value="0.1263234515" flag="0"/>
      <vel value="0.0265862312" flag="0"/>                    |       <vel value="0.0585187640" flag="0"/>
      <psa03 value="0.0982430136" flag="0"/>                  |       <psa03 value="0.2947429778" flag="0"/>
      <psa10 value="0.0162962367" flag="0"/>                  |       <psa10 value="0.0169652717" flag="0"/>
      <psa30 value="0.0012745212" flag="0"/>                  |       <psa30 value="0.0016538703" flag="0"/>
    </comp>                                                         </comp>
    <comp name="10.HHZ">                                      |     <comp name="21.HNZ">
      <acc value="0.1216350151" flag="0"/>                    |       <acc value="0.1318365404" flag="0"/>
      <vel value="0.0341038089" flag="0"/>                    |       <vel value="0.0369035893" flag="0"/>
      <psa03 value="0.1477548844" flag="0"/>                  |       <psa03 value="0.1561743035" flag="0"/>
      <psa10 value="0.0140938729" flag="0"/>                  |       <psa10 value="0.0147768619" flag="0"/>
      <psa30 value="0.0014342152" flag="0"/>                  |       <psa30 value="0.0014173057" flag="0"/>
    </comp>                                                         </comp>
  </station>                                                      </station>

location of the instruments are not exactly the same. SM is more tightly coupled to the ground. Both traces of the instruments HH/HN are pretty clear and sharp Main differences on these summary parameters are on pga/pgv/psa03(0.3s) values on the horizontal components Vertical values are consistent

Output on the Strong motion HN are consistent with the strong motion tools output for this event.

salichon commented 2 years ago

Here is a plot of the event 2020p201743 (M5 distance: 100km) derived Strong motion PSA parameters for MQZ HH and HN | 3 components no rotation | Orientation assumed the same (Note: this is a raw analysis to investigate if issues are existing and waht they are consisting of)

_Left: PSA graphs (HHZ and HNZ) 3 components | Right: PSA Ratio of the respectives components with tthe different instrumentations [ Left x: period in second; right x:period in second y:Ratio %g_psaHH/%gpsaHN ]

Beyond 4 seconds period i deem that the ratio values are not representative.
(4 seconds - 0.25 Hz This about corresponds to the natural frequency of a ~~ >30 stories high building - natural period increases with height) The values are in %g and beyond 4 seconds very small and hence with potential large difference uncertainties (e.g. at 4.4s psa_hhz 7.09876e-05 psa_hnz 4.78637e-05 is already 1.5).

From what is see here in that case: --> Comparing very small values does not make sense here to me :) (say here the decent limit would be 4 seconds) with the regard to the magnitude of the event and its distance
- Below 4 seconds periods for the PSA: The Z component PSA derived data is consistent from one to another instruments when using representative data set. The Horizontal components are showing some obvious differences. More specifically in the range below 2 seconds. Assumptions: site effect as in instrument disposition and installation. As mentioned above: The instruments are not exacly collocated and Strong motion instrument is fixed on the ground

Derived data avaiable here (PSA for the 3 components and different instruments. ) 20200315_2020p201743_PsaHH_HN.tar.gz

Will check these assumptions again and track other source of errors with the next relevant event.

salichon commented 2 years ago

---- Summary results: --- https://www.geonet.org.nz/earthquake/technical/2020p475488 M5.9 is located in Fjordland ~ ~400 KM from MQZ ~100km from MLZ Strong motion tool does provide SM data https://strongmotion.geonet.org.nz/#/event/2020p475488 Sta Pgav | Pgah |Pgvv |Pgvh MLZ 3.218 | 3.541 | 0.489 | 0.371 MQZ 0.068|0.188|0.022|0.018

MQZ and MLZ data are available and comparable (here below recomputed locally )

<stationlist created="" xmlns="ch.ethz.sed.shakemap.usgs.xml" <stationlist created="" xmlns="ch.ethz.sed.shakemap.usgs.xml"

  <station code="MLZ" name="MLZ" netid="NZ" insttype="CMG-3ES |   <station code="MLZ" name="MLZ" netid="NZ" insttype="FBA-ES-
    <comp name="10.HHE">                                      |     <comp name="21.HNE">
      <acc value="0.4229672999" flag="0"/>                    |       <acc value="0.3618856398" flag="0"/>
      <vel value="0.3659175560" flag="0"/>                    |       <vel value="0.3711099389" flag="0"/>
      <psa03 value="1.2677875227" flag="0"/>                  |       <psa03 value="1.2403931778" flag="0"/>
      <psa10 value="0.4350361154" flag="0"/>                  |       <psa10 value="0.4504940692" flag="0"/>
      <psa30 value="0.0610643821" flag="0"/>                  |       <psa30 value="0.0615814795" flag="0"/>
    </comp>                                                         </comp>
    <comp name="10.HHN">                                      |     <comp name="21.HNN">
      <acc value="0.3193711150" flag="0"/>                    |       <acc value="0.3347459452" flag="0"/>
      <vel value="0.2464728413" flag="0"/>                    |       <vel value="0.2503942757" flag="0"/>
      <psa03 value="0.9258431552" flag="0"/>                  |       <psa03 value="1.0410995613" flag="0"/>
      <psa10 value="0.2961127905" flag="0"/>                  |       <psa10 value="0.2863315358" flag="0"/>
      <psa30 value="0.0357367901" flag="0"/>                  |       <psa30 value="0.0395135999" flag="0"/>
    </comp>                                                         </comp>
    <comp name="10.HHZ">                                      |     <comp name="21.HNZ">
      <acc value="0.3365245925" flag="0"/>                    |       <acc value="0.3274415320" flag="0"/>
      <vel value="0.4820284336" flag="0"/>                    |       <vel value="0.4887159460" flag="0"/>
      <psa03 value="0.7003180371" flag="0"/>                  |       <psa03 value="0.7130276059" flag="0"/>
      <psa10 value="0.6049872591" flag="0"/>                  |       <psa10 value="0.6201437560" flag="0"/>
      <psa30 value="0.0700818604" flag="0"/>                  |       <psa30 value="0.0723492409" flag="0"/>
    </comp>                                                         </comp>
  </station>                                                      </station>
  <station code="MQZ" name="MQZ" netid="NZ" insttype="CMG-3ES |   <station code="MQZ" name="MQZ" netid="NZ" insttype="FBA-ES-
    <comp name="10.HHE">                                      |     <comp name="21.HNE">
      <acc value="0.0059879198" flag="0"/>                    |       <acc value="0.0150929550" flag="0"/>
      <vel value="0.0144443891" flag="0"/>                    |       <vel value="0.0183881625" flag="0"/>
      <psa03 value="0.0147165093" flag="0"/>                  |       <psa03 value="0.0171418621" flag="0"/>
      <psa10 value="0.0112023555" flag="0"/>                  |       <psa10 value="0.0075501583" flag="0"/>
      <psa30 value="0.0054860242" flag="0"/>                  |       <psa30 value="0.0048658806" flag="0"/>
    </comp>                                                         </comp>
    <comp name="10.HHN">                                      |     <comp name="21.HNN">
      <acc value="0.0050133942" flag="0"/>                    |       <acc value="0.0191646313" flag="0"/>
      <vel value="0.0175696390" flag="0"/>                    |       <vel value="0.0159122795" flag="0"/>
      <psa03 value="0.0101204720" flag="0"/>                  |       <psa03 value="0.0257699208" flag="0"/>
      <psa10 value="0.0075251001" flag="0"/>                  |       <psa10 value="0.0120140415" flag="0"/>
      <psa30 value="0.0048763954" flag="0"/>                  |       <psa30 value="0.0059326115" flag="0"/>
    </comp>                                                         </comp>
    <comp name="10.HHZ">                                      |     <comp name="21.HNZ">
      <acc value="0.0059313295" flag="0"/>                    |       <acc value="0.0069631843" flag="0"/>
      <vel value="0.0211748722" flag="0"/>                    |       <vel value="0.0218031156" flag="0"/>
      <psa03 value="0.0119290356" flag="0"/>                  |       <psa03 value="0.0129613151" flag="0"/>
      <psa10 value="0.0071544516" flag="0"/>                  |       <psa10 value="0.0070231955" flag="0"/>
      <psa30 value="0.0114401078" flag="0"/>                  |       <psa30 value="0.0117586620" flag="0"/>
    </comp>                                                         </comp>
  </station>                                                      </station>

MLZ: Derived summary data are consistent between HH and HN (all components) MQZ: Derived summary data are consistent on the Z component - Horizontal components differ in the same trend as the previous event.

salichon commented 2 years ago

Here is a plot of the event 2020p475488 M5.9 derived Strong motion PSA parameters for MLZ HH and HN ( distance: 100km) | 3 components no rotation _Left: PSA graphs (HHZ and HNZ) 3 components | Right: PSA Ratio of the respectives components with the different instrumentations [ Left x: period in second; right x:period in second y:Ratio %g_psaHH/%gpsaHN ]

Beyond 4 seconds values are very small and deemed not relevant to compare with.
Nevertheless MLZ comparison shows a better agreeement between the 2 intrumentations (ration is in the range 1 for a wider range of periods)
the Vertical component is the one that is the most consistent with regards to the horizontal ones
Assumption installation may differ - site effect)

parameters for MQZ HH and HN ( distance: 400km) _Left: PSA graphs (HHZ and HNZ) 3 components | Right: PSA Ratio of the respectives components with the different instrumentations [ Left x: period in second; right x:period in second y:Ratio %g_psaHH/%gpsaHN ]

MQZ displays a consistent ratio for its Vertical component (for HH/HN) to the difference of much significant discrepancies on the horizontal components. Assumptions: installation difference - site effects - also perhaps instrument relative orientation may play a role in the difference.

For both events above the comparaison between HH and HN is relevant at leasts on the Vertical component

Data available at 20200624_2020p475488_PsaHH_HN.tar.gz

salichon commented 2 years ago

---- Summary results: --- https://www.geonet.org.nz/earthquake/technical/2020p606562 M5.4 is located in Fjordland ~400 KM from MQZ and 100 km for MLZ Strong motion tool does provide data https://strongmotion.geonet.org.nz/#/event/2020p606562 Sta Pgav | Pgah |Pgvv |Pgvh MLZ 0.582 | 0.990 | 0.039 | 0.062 MQZ is not provided by the tool (signal to noise ratio)

MQZ and MLZ data are available and comparable ( when recomputed locally ) (upon threshold SNR on MQZ reduced)

<stationlist created="" xmlns="ch.ethz.sed.shakemap.usgs.xml"   <stationlist created="" xmlns="ch.ethz.sed.shakemap.usgs.xml"
  <station code="MLZ" name="MLZ" netid="NZ" insttype="CMG-3ES |   <station code="MLZ" name="MLZ" netid="NZ" insttype="FBA-ES-
    <comp name="10.HHE">                                      |     <comp name="21.HNE">
      <acc value="0.1097885017" flag="0"/>                    |       <acc value="0.1006030431" flag="0"/>
      <vel value="0.0638779698" flag="0"/>                    |       <vel value="0.0619786209" flag="0"/>
      <psa03 value="0.2081247422" flag="0"/>                  |       <psa03 value="0.2032801817" flag="0"/>
      <psa10 value="0.1010981966" flag="0"/>                  |       <psa10 value="0.1037319811" flag="0"/>
      <psa30 value="0.0098056856" flag="0"/>                  |       <psa30 value="0.0098380985" flag="0"/>
    </comp>                                                         </comp>
    <comp name="10.HHN">                                      |     <comp name="21.HNN">
      <acc value="0.0743964401" flag="0"/>                    |       <acc value="0.0848356426" flag="0"/>
      <vel value="0.0452266220" flag="0"/>                    |       <vel value="0.0508578654" flag="0"/>
      <psa03 value="0.1790381132" flag="0"/>                  |       <psa03 value="0.1976003936" flag="0"/>
      <psa10 value="0.0602227080" flag="0"/>                  |       <psa10 value="0.0585418850" flag="0"/>
      <psa30 value="0.0054504785" flag="0"/>                  |       <psa30 value="0.0065544639" flag="0"/>
    </comp>                                                         </comp>
    <comp name="10.HHZ">                                      |     <comp name="21.HNZ">
      <acc value="0.0552732264" flag="0"/>                    |       <acc value="0.0590570866" flag="0"/>
      <vel value="0.0401399186" flag="0"/>                    |       <vel value="0.0392074679" flag="0"/>
      <psa03 value="0.1200724508" flag="0"/>                  |       <psa03 value="0.1165112536" flag="0"/>
      <psa10 value="0.0595654031" flag="0"/>                  |       <psa10 value="0.0607342540" flag="0"/>
      <psa30 value="0.0068880807" flag="0"/>                  |       <psa30 value="0.0070995212" flag="0"/>
    </comp>                                                         </comp>
  </station>                                                      </station>
  <station code="MQZ" name="MQZ" netid="NZ" insttype="CMG-3ES |   <station code="MQZ" name="MQZ" netid="NZ" insttype="FBA-ES-
    <comp name="10.HHE">                                      |     <comp name="21.HNE">
      <acc value="0.0034485443" flag="0"/>                    |       <acc value="0.0110414716" flag="0"/>
      <vel value="0.0028511579" flag="0"/>                    |       <vel value="0.0041728100" flag="0"/>
      <psa03 value="0.0068244668" flag="0"/>                  |       <psa03 value="0.0117987345" flag="0"/>
      <psa10 value="0.0050337637" flag="0"/>                  |       <psa10 value="0.0041371845" flag="0"/>
      <psa30 value="0.0009252253" flag="0"/>                  |       <psa30 value="0.0007074712" flag="0"/>
    </comp>                                                         </comp>
    <comp name="10.HHN">                                      |     <comp name="21.HNN">
      <acc value="0.0032250805" flag="0"/>                    |       <acc value="0.0137111113" flag="0"/>
      <vel value="0.0030939045" flag="0"/>                    |       <vel value="0.0036033617" flag="0"/>
      <psa03 value="0.0071020525" flag="0"/>                  |       <psa03 value="0.0110308212" flag="0"/>
      <psa10 value="0.0041287137" flag="0"/>                  |       <psa10 value="0.0051677919" flag="0"/>
      <psa30 value="0.0006931947" flag="0"/>                  |       <psa30 value="0.0010194502" flag="0"/>
    </comp> 
    <comp name="10.HHZ">                                      |     <comp name="21.HNZ">
      <acc value="0.0018454810" flag="0"/>                    |       <acc value="0.0039726319" flag="0"/>
      <vel value="0.0037273912" flag="0"/>                    |       <vel value="0.0037179299" flag="0"/>
      <psa03 value="0.0048416273" flag="0"/>                  |       <psa03 value="0.0050028354" flag="0"/>
      <psa10 value="0.0042838425" flag="0"/>                  |       <psa10 value="0.0043137068" flag="0"/>
      <psa30 value="0.0019592930" flag="0"/>                  |       <psa30 value="0.0019983280" flag="0"/>

same observations as before : MLZ okay - MQZ Horizontals have the most noticeable differences between HH/HN Note: with regards to the distance and the noise levels MQZ HN is borderline for providinf relevant results.

salichon commented 2 years ago

plot of the event 2020p606562 derived Strong motion PSA parameters for MLZ HH and HN ( distance: 100km) | 3 components no rotation | resp used. _Left: PSA graphs (HHZ and HNZ) 3 components | Right: PSA Ratio of the respectives components with the different instrumentations [ Left x: period in second; right x:period in second y:Ratio %g_psaHH/%gpsaHN ]

Respective differences between HN/HH are likely satisfying and varies only within max ~~ 20% over the horizontals

parameters for MQZ HH and HN ( distance: 400km) | 3 components no rotation | resp used. _Left: PSA graphs (HHZ and HNZ) 3 components | Right: PSA Ratio of the respectives components with the different instrumentations [ Left x: period in second; right x:period in second y:Ratio %g_psaHH/%gpsaHN ]

Similar observations MQZ Vertical components are consistent from one instrument to the other Horizontal components differ

salichon commented 2 years ago

Hi @SquirrelKnight here above are the analysis of 3 relevant events M>5 at distance 100/400 km from MLZ and MQZ dual instrumented sites. It is expected some relevant measures from these OK observations (I had to lower the SNR for MQZ strong though for one of the 400km distant event)| (NB: I might need to redraw the left sides graphics for clarity - looks like they show only one type of instrument)

The results here look consistent relevant to what s provided with other tools.
The instrumentations are providing consistent results between Weak and Strong for these 3 events. (as assumed using the Vertical component)
About the noticeable discrepancies on the horizontal of MQZ: I reckon this would need to check if the differences are related to horizontal orientation and/or (very) local installation/site effects, (or "dodgy" instrument H components).

Further investigation: Get the closest dual instrumented station, a set of most significant though fairly distant events (Say Max ~400km). Compare Results with regards to local effects. Also Can the ratio observed on the horizontal be expected ?

cheers!

salichon commented 2 years ago

Assume (only) Part of the problem may come from a "potential" component reversal on one of the instruments: 1- The Z components HH/HN match 2- The respective Horizontal components comparison shows discrepancies
(computing the norm of the Horizontal Psa comp. gave insights) 3- for MQZ 2020.225 event inspection of the raw trace Applying a- Demean/detrend and filtering BP 07-5 Hz on the components b- Derivation filter (dx/dy) on the Broadband to get acceleration equivallent. Observations: Z traces phase and event onset OK Respective E and N a bit less better Matching of derived HHN to HNE is addressing the question of the component orientation

--> investigation in instrument orientation on going

salichon commented 2 years ago

From Conrad: BB orientation is legit mqz_orient Check of the SM instrument will be done soon

salichon commented 2 years ago

update: @SquirrelKnight field check about to confirm Strong motion orientation (and misleading component naming) . HNN would be 280 deg. (TBC)

When enacted would you check while rotating the horizontal signals. (hence shifting the metadata azimuth) ? Thanks!
j

salichon commented 2 years ago

FYI @mnaguit - probably can be useful for some strong motion derived data sets when eventually figured out :)

salichon commented 2 years ago

@SquirrelKnight first iteration of change for the station MQZ strong: 1- Legacy site code change (20 to 21) (location of instrument) 2-Correction of the orientation at least from Mid 2019 to orientation of N to azimuth 270 Warning: Just noticed that it may contain a bug into the metadata built for historical data: orientation of the strong is considered 270 deg since the start of the Strong motion as in https://service.geonet.org.nz/fdsnws/station/1/query?network=NZ&station=MQZ&level=channel&format=text&channel=HN?

Second stage will be : fix the metadata and rename the data component properly in the 12Z system

FYI @elidana

SquirrelKnight commented 2 years ago

Great. Thanks for digging in deep!

On Sun, Oct 17, 2021 at 7:29 PM jerry @.***> wrote:

@SquirrelKnight https://github.com/SquirrelKnight first iteration of change for the station MQZ strong: 1- Legacy site code change (20 to 21) (location of instrument) 2-Correction of the orientation at least from Mid 2019 to oreination of N to azimuth 270

Warning: Just noticed that it may contain a bug into the metadata built for historical data: orientation of the strong is considered 270 deg since the start of the Strong motio) as in

https://service.geonet.org.nz/fdsnws/station/1/query?network=NZ&station=MQZ&level=channel&format=text&channel=HN ?

Second stage will be : fix the metadata and rename the data component properly in the 12Z system

FYI @elidana https://github.com/elidana

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/GeoNet/help/issues/84#issuecomment-945311853, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJ5LPP44MTPVW6N2UYE5QV3UHOBB3ANCNFSM5EUZ5ZVQ . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

salichon commented 2 years ago

Addendum (due to confusing naming...)

Check 21 HNN is actually az 270 Deg for the period 2016-02-24T01:02:00|2019-06-17T05:25:00 (as per delta)
Check that 20 HNN is correct az 0 deg between 2009-08-06T00:00:01|2016-02-23T08:00:00
test: Swap components HNN <- > HNE for some events

Actions: When confirmed correct the bearing between 2016-02-24T01:02:00|2019-06-17T05:25:00 (metadata) When approved set relevant component naming. (HNN -> HN1) later process) (data)

salichon commented 2 years ago

The 2016 orientation is deemed correct ENZ, based on photos of the site from that period.

The Basalt from 2016 actually looks to be oriented correctly. The Basalt could have been configured incorrectly though and the channel 3 maybe have been configured as N (it should be E). Although now that it is pulled out we cannot say for sure, just that if everything (configuration) was as it should be then it would have been fine.

from the comment above, then the assumption is that the cabling/configuration 2016-2019 was perhaps messed up but with an install put away since 2019, Only testing data will bring a some ideas / and fix.

salichon commented 2 years ago

For checking the above assumption (or similar consequence)

use event 2020p2020p606562
swap PSA output components (for the sake of simplicity)
Compute ratio
plot respective component ratio HHE Vs HNN and HHN Vs HNE

Swapped PSA output

As opposed to https://github.com/GeoNet/help/issues/84#issuecomment-940482069 with: Screenshot from 2021-11-15 17-55-26

So several things here (still to be solidified) @elidana @SquirrelKnight

1 - HN strong motion Horizontal components are most likely Inverted HNE is confirmed actually HNN! (and HNN is HNE) or more precisely HHN aligns much better with HNE in terms of output (HHE with HNN) 2 - Bearing is verified on the field and being 270 which is another Added source of confusion for naming channel orientation being likely South and West 3- To be verified: It looks like that the process (strong motion tools Geonet) I used to check does not rotate the channels with regards to the azimuth.

Hence badly named and orientated component chanels are a nightmare to deal with overtime.

salichon commented 2 years ago

Checking now the MQZ Strong Season II 2016 Feb to 2019 June
Selected events (from Jesse' s batch) Good SNR. prior to 2016 only veet from Kaikoura aftershocks 14-15Nove (so limited)
Picked fro try '2016p858147' ( M5 ~100km Monday 14 November) , '2016p861632' ( M5 ~100km Tue Nov 15),

Derivation of the PSA parameters for the horizontal components in Nov. 2016 It looks like it s ok-ish (compared to 2019/2020) for the East components. The North Components looks like is not very much. (quake source? )

Check on Vertical comp and other events .

So it shows that the date 20161113 has some signal inconsistencies - As a reminder the main event kaikoura in the the same date at 2016-11-13T11:03:43.808392Z the event considered above is one hour after the main at 2016-11-13T12:20
the day after looks fine 14 November

So .. some actual carefull event and signal quality control is always recommended and could have been made on the selection of events - usually aftershock sequences are slightly noisy - this does not help to track along issues.

Some events traces can be inspected for confirmation: for the problematic event 2019p858147 the inspection of the traces gives some hints about when to compute strong motion parameters or when to remove other signal from an event to get the strong motion parameter:
Zoomin on the oeriod of interest 12:20 shows how likely not very relevant to compute the strong motion paramters without a bit of data QC

other events to check later on in between 2016 and 2019 to be found to confirm that the 2016-2019 seasonII has some consistent HN or BN vs HH psa. (hence most likely orientation is all good as well as instrumentaion) These ones are worthy Close enough and well after the aftershock sequence

publicid	eventtype	origintime	modificationtime	longitude	latitude	magnitude	depth	magnitudetype
2017p795065	earthquake	2017-10-22T04:10:59.674Z	2017-10-27T03:21:33.464Z	173.4589	-42.3581	5.588791	8.27472	M
2017p161601	earthquake	2017-03-01T19:01:02.699Z	2017-03-01T19:08:37.228Z	173.0949	-42.9601	5.222643	11.51696	M

salichon commented 2 years ago

2017p795065 psa comparison between BN and HH is consistent (does not mean that oreintation are fully matching ...:)) strong assumption the period _20162019 seasonII is good in terms of data/metadata - QC and critical trace inspection was required

FYi @elidana

salichon commented 2 years ago

--- Summary ---- MQZ HH Vs Strong motion (2016-2020)

HH Horizontals are most likely all good and properly orientated for that period
SM motion parameters derived from either instruments are okay and consistent in values, consistent with the GeoNet tools.
Discrepancies for the period 2019/2020 is thought to be related to SM Horizontals channel naming and their orientation with regards to the weak motion. (basically inverted channels ) - Very confusing.
Many of the processing do not account for Azimuth or do not account for it when geographical coordinates.
SM from 2016 to 2019 is looking like properly named and orientated
Discrepancies shown for the set of events in 2016 is thought to be related to the choice of data within an immediate aftershock sequences. SNR + coda + superimposed events + local site effects (2 different locations) + instrumentation Even though the data looks really good quality for both instruments at that time. it looks like some frequency content differences. as in

-Data from 2017 with similar instrumentation do not show major critical difference between both instrumentations derived parameters. As seen from the SM derived parameters.

Verification tools used (Geonet strong motion parameters processing on HH and HN/BN) - scwfparam

FYI @elidana @SquirrelKnight @mnaguit @ozym et al ...

elidana commented 2 years ago

summary

broadband data need to be corrected:

from 2019xx : do we need to update channel names or swap components? rename channels to HN1/HN2 would be preferable and more clear
[ ] stream renaming Should be implemented over an identified period of time (untill the orientation is sorted - see below)
possibly visit the site and rerotate the sensor (or check sensor config and cables)
[x] sensor oreientation will happen and should be realigned ENZ as per convention
if using BN HN : all good (@SquirrelKnight can you please add your latest graph for this station?) - at high frequency difference might be Nyquist..

salichon commented 2 years ago

@SquirrelKnight @elidana @mnaguit FYI @ozym MQZ strong was reorientated, NB: as you can check in https://service.geonet.org.nz/fdsnws/station/1/query?station=MQZ&level=channel&format=text&channel=HNE,HNN there is a long legacy of instrumentation orientation for this site.

The later being: NZ|MQZ|21|HNN|Azimuth |270.000000

Has been modified on the field ~10:30 on 23/02/2022 NZT. (and will be updated in Delta asap)

N and E as they should be - Someone to verify from present dat set ? PR merged into delta - ongoing data are deemed true ENZ
[x] check on fdsn station service https://service.geonet.org.nz/fdsnws/station/1/query?station=MQZ&level=channel&format=text&channel=HN?
We will task/work back the legacy from the period at least: 2016-02-24 to Feb 2022 on the data and the metadata side of the thing - this will need a bit of time and caution
- [ ] create a data metadata change ticket and proceed

upon completion close this painful ticket!! :)

cheers jerome

GeoNet / help

MQZ issue with broadband vs. accelerometer comparison #84

Summary of the PSA process above from what i understood