Interesting pressure anomaly at RS01SBPS around end of May 2015

Sampling water column parameters at a high vertical resolution

Moderator: Staff

Post Reply
FriedrichKnuth
Posts: 5
Joined: Mon Apr 24, 2017 6:43 pm

Fri Apr 28, 2017 8:38 pm

During data evaluation, I came across a pressure anomaly in the CTD record, visible in both the CTD on the RS01SBPS 200 platform and the profiler, which had been sitting stationary right above the platform (not profiling during this time).

Oregon Slope Base Shallow Profiler Mooring (1A)
200m Platform (RS01SBPS-PC01A-4A-CTDPFA103)
Shallow Profiler (RS01SBPS-SF01A-2A-CTDPFA102)
Lat 44.529, Lon -125.389
Water Depth 2906 m
Time Period: 2015 05/27 - 05/31
ctd_pressure_anomaly.png
CTDPFA103 and 102 pressure plots
ctd_pressure_anomaly.png (556.44 KiB) Viewed 375 times
RS01SBPS_OR_Slope_Base_Shallow_Profiler.jpg
Mooring diagram
RS01SBPS_OR_Slope_Base_Shallow_Profiler.jpg (3.12 MiB) Viewed 375 times
location.png
Location of RS01SBPS
location.png (1.76 MiB) Viewed 375 times

FriedrichKnuth
Posts: 5
Joined: Mon Apr 24, 2017 6:43 pm

Fri Apr 28, 2017 8:54 pm

I plotted the data up via the UI as well, which shows that it looks to be a gradual signal over the course of a few days.
https://ooinet.oceanobservatories.org/d ... e43-sample
RS01SBPS-SF01A-2A-CTDPFA102-UIplot.png
CTDPFA201 UI plot
RS01SBPS-SF01A-2A-CTDPFA102-UIplot.png (146.37 KiB) Viewed 373 times
This anomaly can also be seen in the Single Point Velocity Meter sensor on the profiler.
https://ooinet.oceanobservatories.org/d ... ocity-data
veplt_anomaloy.png
VELPTD102 UI plot
veplt_anomaloy.png (721.01 KiB) Viewed 373 times
A colleague at Rutgers plotted up the ADCP data from RS01SBPS-PC01A-05-ADCPTD102, which show that the changes in pressure are correlated with the pitch, roll and currents in the East-West direction.
ADCPTD102.png
ADCPTD102 plot
ADCPTD102.png (205.17 KiB) Viewed 373 times

Should these data be marked as suspect because the platform appears to have moved downward in the water column by 4 meters, due to a current? It is my assessment that these data are good and the time range can be annotated as an event, but not as being suspect data.

elhunter
Posts: 1
Joined: Tue May 16, 2017 11:01 am

Tue May 16, 2017 1:03 pm

This was part of an email thread I decided to move to this forum:

Hi Friedrich,

Well, I'm not sure how the OOI defines suspect data vs an event. But seems advisable to flag this data as suspect. This sort of SSH anomaly (3-4m over days) is well outside of the global range of SSH and for major boundary currents and transient eddies. I don't think a sensor failure has to be limited to an electrical or mechanical failure of the sensor itself, but also includes interference with the sensor which results in questionable data. That seems to be the case here. For example, on my cabled ADCP in the Raritan, it recently began reading vertical velocities on the order of 10s of cm/s.This is clearly bad data. After a few days it corrected itself. Presumably there was interference with the sampling volume, but the sensor itself did not malfunction. Nevertheless I removed the data from the published data stream.

Of course, it's up to the OOI QA/QC crew, but in my opinion this "event" would not be part a science-quality dataset.

On another note. I noticed your figures from the single point velocity meter on the profiler and they look suspect. They are not consistent with the ADCP velocities. The eastward and northward velocities seem to be noise. (+/- 0.5 m/s). And the vertical velocities are -0.5 - -1.0 m/s. I wouldn't trust this data either.

Eli


Friedrichs response:

If the data product were SSH, then this data most certainly would be suspect. But since it is just measuring pressure, I don't think we can flag these data as suspect. The pressure sensor accurately recorded the changes in pressure in the water column, it is just that the platform moved due to a current, rather than SSH changing. So if someone is using this data to calculate SSH, we want to give them a heads up that events like this happen, but not flag the pressure data as suspect, which would imply fake data and an issue with the sensor or calibration of the instrument. Would you agree with this distinction?



Friedrich,

I understand the distinction you are trying to make. I guess the question is what data product the OOI is trying to provide. If the goal is provide a science quality dataset which reflects the state of the ocean, then I would consider this as suspect. Whether or not the pressure sensor measured pressure, the system did not measure the pressure as expected. i.e. the platform moved.

Of course, if the purpose of the pressure sensor is to simply monitor the vertical position of the instrument platform in the water column, then this is not suspect. it's doing it's job.

Eli

FriedrichKnuth
Posts: 5
Joined: Mon Apr 24, 2017 6:43 pm

Tue May 16, 2017 3:23 pm

Correct, in this case SSH isn't a data product created by this particular suite of OOI sensors, which is what you are referring to as a science quality dataset. I think this data product could be calculated, but that would be an enhancement and isn't in the current scope for this sensor. Its sole purpose is to measure pressure according to its position in the water column, which, as we have determined, can change with currents. This pressure data is then used in the data product calculation for multiple other co-located sensors, including dissolved oxygen, fluorometry and pH sensors. If we mark the pressure data as suspect, then these other data products would programmatically also be marked as suspect, which isn't accurate.

However, users can derive SSH from the existing data, corrected for periods of platform movement, which we would link to as a community developed "value added data product" under http://oceanobservatories.org/community-tools/.

Another sensor suite to consider when calculating SSH might be the collection at the Oregon Slope Base Seafloor (RS01SLBS). These sensors are fixed on the seafloor below RS01SBPS at ~2900 m depth, and are unaffected by currents. Below is a pressure record from the Tidal Seafloor Pressure sensor (RS01SLBS-MJ01A-06-PRESTA101), which does not show elevated pressure values during the same time period. As mentioned, this sensor is fixed to the seafloor (unaffected by current) and thus a more useful dataset to derived SSH from.
RS01SLBS-MJ01A-06-PRESTA101.png
RS01SLBS-MJ01A-06-PRESTA101.png (131.7 KiB) Viewed 320 times

SkipDenny
Posts: 1
Joined: Wed Jun 14, 2017 9:24 pm

Fri Jun 16, 2017 5:44 pm

Please note: this is data from a platforms moored at 200m depth in 2500 and 3000m water depth. As it happens, we modeled (OrcaFlex) the motion of the moorings before we put them in, based on some measurements, and some “worst case” scenarios. This effort was done to establish anchor weights, identifiy potential problems with some techniques used to make the EOM cable more neutral in the water, etc. We’re recently sent a set of ADCP measurements, particularly with the Endurance site (1B) as there is a 75 KhZ ADCP that is bottom mounted, with noted apparent blowdown data from the moorings to the people who did the original modeling for model validation. For the model, estimates had to be made of the Cd of the various components, and we need to solidify/verify those estimates. Strong eddies were noted when we were installing the moorings, and “jets” in the 100-400m depth range have been observed during our operations (e.g. the ROV could not fight the currents during some dives with >0.5 m/s speed capability). To understand platform motion, and how it influences measurements, is of intense interest to us and perhaps should be a project of its own. We think that "cuddies" may be providing this current, however in the deeper water we do not have full coverage of the current profile and ~90% of the drag on the mooring will be due to the cables.

Skip Denny
Principal Ocean Engineer
APL-UW, OOI-RSN Team

Post Reply
  • Information
  • Who is online

    Users browsing this forum: No registered users and 1 guest