move descript of orientation out of sources when geographical

[crotwell]

Many sources have "oriented" subsource identifiers that can be geographical. The spec would be cleaner if there was a section for just "how to code oriented sources" in one place. Adding information from #15 to this section as well.

Proposal, for source codes H, L, G, M, N, A, F, J, P, S change the "position Code" to be "see Geographic Orientations"

and add a "Geographic Orientation" section saying:

For sources that record data in a direction typically aligned with geographical coordinate systems, the subsource
identifier should follow the following conventions:

   .. table::
      :align: left

      ===================== ======
      **Z**, **N**, **E**   Traditional (Vertical, North-South, East-West), when with 5 degrees of true directions
      **A**, **B**, **C**   Triaxial (Along the edges of a cube turned up on a corner)
      **T**, **R**          For formed beams or rotated components (Transverse, Radial)
      **1**, **2**, **3**   Orthogonal components but non traditional orientations
      **U**, **V**, **W**   Optional components
      ===================== ======

[chad-earthscope]

The downside is that one cannot look at the Source code and "see" all the Subsource codes that go with the Source. One needs to scroll to see what the combinations are. It's already a bit like this but would become worse.

Also, not all of the Geographic Orientation codes would be appropriate for the few that are geographic, i.e. some of those make no sense for linear strain.

I'm also mostly ambivalent.

[crotwell]

My concern is that most orientable non-seismic channels currently have meaningless subsources. A, B, F, J and S all look to me like they would benefit from at least some of the geographical subsource descriptions, eg use 1,2 if not within 5 deg of north and east. J already looks almost like a copy paste from the seismometer.

Not sure I understand what you mean about linear strain? Which ones are not applicable? I get that Z doesn't make sense for tilt, but it would benefit from a N,E vs 1,2 convention.

Copy and past is fine with me if you prefer. We already have what looks like a copy paste for J, but it is already slightly different, having a Z12 line.

Unrelated, but R, rainfall, as "ZNE - Traditional" which makes no sense to me. What does Rainfall in the north direction mean?

[chad-earthscope]

Unrelated, but R, rainfall, as "ZNE - Traditional" which makes no sense to me. What does Rainfall in the north direction mean?

Typhoon :P

[crotwell]

touche, touche
:)

[chad-earthscope]

Not sure I understand what you mean about linear strain? Which ones are not applicable? I get that Z doesn't make sense for tilt, but it would benefit from a N,E vs 1,2 convention.

I was thinking:
T, R For formed beams or rotated components (Transverse, Radial)

but I suppose it could be rotated.

[crotwell]

Ah. T and R are already kind of weird as they are commonly used, but always derived I think. And so not "real" channels, which sort of goes against the philosophy of the rest of the naming system. Not sure I understand why they were included in the original SEED definition.

[WayneCrawford]

I do like @crotwell 's idea, it would clarify which codes are "reserved" for geographical and would avoid differences in definitions between different "source codes" (currently "Seismometer" says "when with [sic] 5 degrees of true..." whereas the others don't). But @chad-iris is right that we should have the codes listed in the appropriate section, plus it seems that certain sensors don't use all of the geographical codes.

I'd be for an intro to geographical codes at the top of the "subsource codes" section, including a table with the strict rules, but also leave the codes in their individual source code sections but with only a reference to the geographical codes section PLUS any source-specific qualifiers.

[chad-earthscope]

OK, first try. I'd like feedback before working through the rest of the source codes that may need geographic orientation code referencing.

A new section at the top of the source and subsource listing describes "Geographic orientation subsource codes" and then a modified section for Seismometer refers to them (in this case there are no others to describe).

http://docs.fdsn.org/projects/source-identifiers/en/draft/channel-codes.html#source-and-subsource-codes

Feedback welcome and needed to move forward with this one.

Some day, over a beer, perhaps in a far away land, I can regale you with tales of battling with the deceitful siren that is Sphinx/ReStructured text. Until then, I ask for your understanding that there is only so much that can be done beyond simple content in these pages.

[WayneCrawford]

Hi Chad, I looked at it and it looks good, but I would change the first sentence to  Traditional orientation values of North-South (N), East-West (E), and Up-Down (Z) The first change is just a typo, the second is important to be explicit that there is an implied polarity to the vertical (you do imply this in the table).  I labeled geophone channels (positive voltage is down) "Z" for a long time before seeing the havoc it wrought on pickers!   It might even be useful to explicit state this with a sentence like: "Z" implies a positive signal for upward motion.  Vertical orientations with a positive output for downward motion should be labeled "3". or maybe that's too long/explicit. I don't think that the seismometer Subsource Code section should have the pre-table sentence.  It only duplicates what is shown in the table and won't work for instruments that can have either geographic orientation codes or other codes. And I guess I would say "Deep sea differential pressure gauge" or Cox et al. [1984] differential pressure gauge" in the pressure section, to avoid confusion with the generic term. Cheers Wayne

…

_______________________________ Wayne Crawford CNRS Researcher Institut de Physique du Globe de Paris 1, rue Jussieu 75238 Paris Cedex 05 +33 6 5151 1054 On 15 September 2020 at 06:18:10, chad-iris ([email protected]) wrote: OK, first try. I'd like feedback before working through the rest of the source codes that may need geographic orientation code referencing. A new section at the top of the source and subsource listing describes "Geographic orientation subsource codes" and then a modified section for Seismometer refers to them (in this case there are no others to describe). http://docs.fdsn.org/projects/source-identifiers/en/draft/channel-codes.html#source-and-subsource-codes Feedback welcome and needed to move forward with this one. Some day, over a beer, perhaps in a far away land, I can regale you with tales of battling with the deceitful siren that is Sphinx/ReStructured text. Until then, I ask for your understanding that there is only so much that can be done beyond simple content in these pages. — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

[crotwell]

Would it be better to only have the positive directions, like:

Traditional orientation values of North (N), East (E), and Up (Z), must be within 5 degrees of true directions.

Also, there is an extra "the" in follow the these conventions

Maybe combine the two subsections Subsource codes for orientation and Geographic orientation subsource codes? It seems like they are really about the same thing?

I like having the link at the top of the table to the geographical, but think we should just reproduce the table in each needed section, instead of a link for each line? So have:

See Geographic orientation codes for descriptions of these subsource codes.

Subsource codes	Description
Z, N, E	Traditional orthogonal (Up-Down, North-South, East-West) When with 5 degrees of true directions
3, 1, 2	Orthogonal components, nontraditional orientations
Z, 1, 2	Orthogonal components, nontraditional horizontals
T, R	For rotated components or beams (Transverse, Radial)
A, B, C	Triaxial (Along the edges of a cube turned up on a corner)
U, V, W	Optional components, also used for raw triaxial output

[chad-earthscope]

• Merged the sections as @crotwell suggested.
• Added text suggestions
• Added appropriate parts of table to Seismometers, Tilt (A), Rotational Sensor (J), and Linear Strain (S)
• Removed NEZ orientation from rainfall code