Legacy Sea-floor depth parameter codes to be mapped to their atomic components (BODCNVS-1860)

[gwemon]

As part of the remodeling of the P01 parameter codes that belong to the P02 parameter group: Bathymetry and Elevation we need to decide on the best suited atomic component terms for them. The BODC Parameter Usage Vocabulary (PUV) makes a distinction between "Depth" as one of the dimensions of a physical entity (http://vocab.nerc.ac.uk/collection/S06/current/S0600169/) and "Depth" the vertical location of an object or process in a medium (http://vocab.nerc.ac.uk/collection/S06/current/S0600167/). For bathymetric depth we are proposing to remodel it as: "Depth (spatial coordinate) of sea-floor (relative to xx datum) in the water body..." using the S06 property http://vocab.nerc.ac.uk/collection/S06/current/S06001697/. while Sea level parameters (http://vocab.nerc.ac.uk/collection/P02/current/ASLV/) will be modeled following the pattern "Depth (relative to XX datum) of the water body..." using the S06 property http://vocab.nerc.ac.uk/collection/S06/current/S0600169/. Comments welcome :-)

[roy-lowry]

Just to check that you're not proposing changing terms like 'Surface elevation (Chart Datum) of the water body by fixed in-situ pressure sensor' to 'Depth (relative to Chart Datum) of the water body by fixed in-situ pressure sensor'. The former refers to the distance ABOVE the datum whereas the latter refers to the distance BELOW it.

[roy-lowry]

As became apparent in the VMG meeting there are issues - I couldn't understand the preferred labels and I composed them - so I'll take a look. Could somebody have a chat with Pauline and see what her ideal P01 preferred label would be for describing GEBCO elevations in a map running from the bottom of the sea through to the top of mountains and post the result here. Possibly a 5-minute chat would be more helpful than an e-mail (think @danibodc said she would do this) as Pauline isn't that familiar with P01. When I get back from Potugal I'll do a systematic analysis of the concepts mapped to ASLV and see if I can come up with some ideas for the mapping.

[roy-lowry]

Just had a quick look at concepts mapped to MBAN. The semantics here are "interesting" as they represent a processing chain that changes the nature of the measurement. It starts with the simple concept that we all know and love as 'water depth'. This is then corrected to a fixed datum (e.g. MSL) to give a water depth relative to a datum, which is then gridded. The final step is the integration of these grids into a digital terrain model of some sort together with terrestrial elevation data. The resulting measurements are either expressed as depths or heights.

'Depth (spatial coordinate) of sea-floor (relative to xx datum) in the water body' has a problem as we should be dealing with S0600169 which is 'Depth', not 'Depth (spatial co-ordinate)'.

The S0600169 definition is 'The vertical dimension through an object or a body of matter measured downward from an upper surface (or top) to its base (or bottom) (e.g. water column depth). Not to be used for depth as a spatial coordinate. 'Looking at this we should surely have 'Depth of water body', not 'Depth of sea-floor'

'Depth of water body' and 'Depth of water body (relative to xx datum)' could work for the first two stages of the processing chain were the definition of S0600169 changed to 'The vertical dimension of a body of matter measured downward from its upper surface to its base (e.g. water column depth). The upper surface may be the physical surface at the time of the measurement or corrected relative to a datum. Not to be used for depth as a spatial coordinate.

I'll see what you think of this before considering the heights in the stage two processed concepts

For the final stage in the processing chain I think we need something slightly different, which is where Pauline's input comes in. My straw man suggestions would be something like:

Elevation expressed as below datum of solid earth surface (relative to xx datum) - I think all GEBCO products are relative to MSL, but check with Pauline. Elevation expressed as above datum of solid earth surface (relative to xx datum)

Mull this over (no rush!) and let me know what you think.

[roy-lowry]

Thinking overnight the idea of making the water body surface a datum is too obscure and won't be well understood. It also creates a semantic car crash with heights because there is no 169 equivalent for height. So the 169 definition should stay at 'The vertical dimension of a body of matter measured downward from its upper surface to its base (e.g. water column depth). Not to be used for depth as a spatial coordinate.'

However. I also realised that the process of correcting a sea-floor measurement to a datum changes the measurement from a 'Depth of the water body' to a 'Depth (spatial coordinate) of sea-floor (relative to xx datum)'. I think this is is the key to the modelling of MBAN.

I'll think a bit more and do some rigorous testing when I get back from Olhao. My remaining uncertainty is what to do about the GEBCO grids and Pauline's input would really be appreciated here.

[danibodc]

Thanks @roy-lowry, Pauline is currently away until 6th November, so have sent her an email in the meantime and have offered to arrange a chat when she is back. Enjoy Olhao Roy!

[roy-lowry]

Getting back to this after the joys of Portugal.

Any contact with Pauline? My main uncertainties are:

1) Can the GEBCO product parameters can be described as 'Depth (spatial co-ordinate)/height of X (relative to mean sea level datum)'? In other words are the GEBCO depths/heights considered relative to MSL.

2) When can we use 'sea-floor' for 'X'? I think the earlier products only included data below sea level in which case 'sea-floor' works for me. However, the later products - I think the ones with 'heights' also include terrestrial topography and having the top of Everest described as 'Height of sea-floor (relative to mean sea level datum)' strikes me as a little daft. Maybe 'Height of solid earth surface (relative to mean sea level)' would be more acceptable.

3) Which of the GEBCO products include terrestrial topography and which don't?

I'm pretty sure Pauline has the answers to these. If she uses GitHub would it be helpful if she were to watch this ticket?

[roy-lowry]

Is it just the P01 concepts mapping to ASLV that include the word 'depth' in their preferred label that I am to look at?

[roy-lowry]

Initial musings on ASLV depths. Feel free to chip in. For measurements relative to instantaneous sea surface the difference between 'water body depth' in MBAN and ASLV is the space/time framework of the measurements. In MBAN time is fixed with each measurement at a different position. In ASLV position is fixed with each measurement at a different time. However, the thing that is measured (which is what P01 describes) is the fundamentally the same. There is another difference. Those making the measurements mapped to MBAN (usually with echo-sounders) cared that their measurements represented the actual physical distances between sea surface and seabed. However, those using underwater pressure sensors to male measurements mapped to ASLV were often only concerned with depth changes as a function of time and so many of the 'depth' measurements were actually 'depth+10m' as pressure sensors read approximately 10m in air due to atmospheric pressure and no account was taken of this in the pressure to depth conversion. This is why there are all the detailed methodology descriptions. I guess these could be carried through to the mapped concept descriptions and their consequences described clearly in their descriptions.

Next I'm going to trawl the memory to look at the duplicate preferred labels to see if they are in fact duplicates or if some semantic subtleties have got lost with time....

[roy-lowry]

Would anybody have time to have a look at the usage of code DEPHPRMN? Is it something like the average of the depths logged by a CTD during a water bottle firing?

Starting to think this could possibly belong together with DEPHPRST and DEPHPREN mapped to AHGT and including 'depth (spatial co-ordinate)' in the preferred label.....

[gwemon]

@roy-lowry @danibodc will provide an update on this. We met with Pauline and she was going to check with her GEBCO network whether our proposal was acceptable. We went through a number of options and the last email relative to the proposal contained what follows: Either Depth (spatial coordinate) of sea-floor relative to XX datum in the water body Where XX datum is defined in the controlled vocab: http://vocab.nerc.ac.uk/collection/S20/current/ (we currently only have a few but more will need to be added)

Proposed definition: The distance between the sea surface and seabed measured from the sea surface referenced to the specified datum and expressed as a positive value. When no datum is specified then the P01 code will refer to “unspecified datum” or “datum specified elsewhere”.

https://en.wikipedia.org/wiki/Bathymetry is a good resource or terms. And Topographic height of sea-floor relative to XX datum in the water body Definition: The distance between the sea surface and seabed (or elevation of the seabed) measured from the sea surface referenced to the specified datum and expressed as a negative value. When no datum is specified then the P01 code will refer to “unspecified datum” or “datum specified elsewhere”.

Alternatively we could also have: Elevation of Earth’s surface relative to XX datum Height of the Earth’s surface relative to Earth’s geoid?

To avoid using “Depth (spatial coordinate)” for this, we could also create a new S06 for “Bathymetric depth”. So we would have 2 new S06 “Bathymetric depth” and “Topographic height” and the P01 codes of the form:

Bathymetric depth of sea-floor relative to XX datum in the water body And Topographic height of sea-floor relative to XX datum in the water body

Notes:

• Keeping “seafloor” as “sea-floor” so that it is as close to what we currently have as possible.
• Definition of the height exactly the opposite as the definition of the depth to avoid confusion

[gwemon]

Would anybody have time to have a look at the usage of code DEPHPRMN? Is it something like the average of the depths logged by a CTD during a water bottle firing?

Starting to think this could possibly belong together with DEPHPRST and DEPHPREN mapped to AHGT and including 'depth (spatial co-ordinate)' in the preferred label.....

It was used in the series database. Looks like it's associated with a number of plankton optical counter series. We will try and check if this might be a P02 mapping mistake.

[roy-lowry]

I quite like:

Bathymetric depth of sea-floor relative to XX datum in the water body Topographic height of sea-floor relative to XX datum in the water body

as it has a consistent symmetry and separates bathymetry/topography from other depths/heights in a way that is intuitively clear (to me at least). It also covers all the various datums to which shallow water bathymetry is recorded. I would regard the Earth's geoid as a datum (think it's closely related to mean sea level). Remember vocabs can always include terms cover cases like unspecified datum.

We would then need the pair 'depth (spatial co-ordinate)', 'height (spatial co-ordinate)' to cover z co-ordinate channels of CTDs and radiosondes, aircraft height, etc., which leaves echo-sounder and sea-floor pressure gauge water depths to be covered. I wondered about a third S06 pair (depth/height) to do this, but maybe height isn't needed. However, 'depth' certainly is.

[danibodc]

I've just had a response from Pauline - she hasn't had a chance to discuss this with GEBCO colleagues but has said:

I think that I agree with Gwen to use 'Bathymetric depth of sea-floor relative to XX datum in the water body'

With regard to how to define a data set (such as the GEBCO grid) that has positive values for land and negative values for depths below the sea surface, I think 'Elevation of Earth’s surface relative to XX datum - Height of the Earth’s surface relative to a specified datum - with the distance between the sea surface and seabed expressed as a negative value and heights of the land surface above the sea surface expressed as a positive value.

[roy-lowry]

I regard 'Elevation' and 'topographic height' as synonyms for which I have equal preference. I prefer Elevation/topographic height of Earth's surface to Elevation/topographic height of sea-floor and even more comfortable with 'solid Earth's surface' than 'Earth's surface' to make it clear we're not talking about sea surface.

[gwemon]

I quite like:

Bathymetric depth of sea-floor relative to XX datum in the water body Topographic height of sea-floor relative to XX datum in the water body

as it has a consistent symmetry and separates bathymetry/topography from other depths/heights in a way that is intuitively clear (to me at least). It also covers all the various datums to which shallow water bathymetry is recorded. I would regard the Earth's geoid as a datum (think it's closely related to mean sea level). Remember vocabs can always include terms cover cases like unspecified datum.

We would then need the pair 'depth (spatial co-ordinate)', 'height (spatial co-ordinate)' to cover z co-ordinate channels of CTDs and radiosondes, aircraft height, etc., which leaves echo-sounder and sea-floor pressure gauge water depths to be covered. I wondered about a third S06 pair (depth/height) to do this, but maybe height isn't needed. However, 'depth' certainly is.

Hi Roy @roy-lowry thank you for this. We already have the Depth/Height S06 terms in place and used in P01. See: http://vocab.nerc.ac.uk/collection/S06/current/S0600167/, http://vocab.nerc.ac.uk/collection/S06/current/S0600169/, http://vocab.nerc.ac.uk/collection/S06/current/S0600166/, http://vocab.nerc.ac.uk/collection/S06/current/S0600264/

[gwemon]

Great thanks @roy-lowry we will review all this with Pauline.

[danibodc]

@roy-lowry @gwemon Pauline and GEBCO would be happy with: Bathymetric depth of sea-floor relative to XX datum in the water body Topographic height of sea-floor relative to XX datum in the water body

[gwemon]

S20 (datum) terms needed for this were also needed for https://github.com/nvs-vocabs/P01/issues/17. They have been added to S20 (https://github.com/nvs-vocabs/S20/issues/3). However new S29 combinations will be needed with S18='seafloor'