Fwd: [cf-convention/vocabularies] Standard names: lake variables (#25)
Mathew Biddle - NOAA Federal
From: taylor13 <notifi...@github.com>
Date: Tue, Feb 4, 2025 at 9:14 PM
Subject: Re: [cf-convention/vocabularies] Standard names: lake variables (#25)
To: cf-convention/vocabularies <vocabu...@noreply.github.com>
Cc: Mathew Biddle <mathew...@noaa.gov>, Mention <men...@noreply.github.com>
Hi all,
Both of the above comments are quite helpful I think. First regarding Jonathan's:
I agree we shouldn't attempt to distinguish water as a substance based on salinity.
In the standard names, we want Roy's definition of a 'water_body' to refer to only the liquid found in bodies of liquid water lying atop an underlying solid surface and interfacing with either the atmosphere or layer of ice above; we exclude water in the atmosphere or underground aquifers (I think), I like the suggestion to use the term 'sea_or_inland_water' for these water bodies, and if only bodies of water associated with a certain area type should be considered, that should be specified using a "where directive" in the cell_methods (e.g., "where land" would indicate only inland bodies are considered and "where sea" would indicate only areas of the sea should be considered).
Regarding Roy's last paragraph above: in CMIP, sea_water was meant to refer to only the water that an ocean component of a coupled climate model simulated. Of the 2062 variables requested from models in CMIP6, about 17% included sea_water in their standard name, and only a few of those omitted specifying "where" the variable should be reported; most indicated "where sea", but for river transports into the sea from land "where land" was specified and for certain fluxes into and out of sea ice, "where sea_ice" was specified. And in a few cases there was no "where" specification given, but this will likely be corrected in CMIP7.
So, I'm coming around to the notion that we can interpret "sea_water" in all datasets written prior to version 1.13 as possibly referencing both inland water and sea water. Going forward, I would favor:
- for datasets compliant with versions of CF prior to 1.13, "sea_water" appearing in a standard name should by default be interpreted as a water body (found either on land or sea)
- for datasets compliant with versions of CF post 1.12,
- "sea_water" in a standard name refers to water "where sea"
- "sea_or_inland_water" in a standard name refers to water anywhere. (The old "sea_water" name should no longer be used for this purpose. Instead, a new standard name should be proposed replacing "sea_water" with "sea_or_inland_water".)
- no standard names should be created with only "inland_sea_water"; if the quantity described by the standard name should be reported only over land areas, then the standard name including the phrase "sea_or_inland_water" should be used and the
cell_methodsshould include "where land". - it is recommended that when considering a variable "where sea", a standard name with "sea_water" be used in preference to one with "sea_or_inland_water".
There are two reasons I'm so keen to retire the current interpretation that "sea_water" can refer to water found anywhere on earth:
- I would claim that the typical person reading CF variable standard names will assume that "sea_water" roughly refers to "water in the oceans" (and excludes inland bodies of water).
- I am attempting to clarify the definition of
area_type"sea" here and in that context limit "sea" to bodies of water (roughly at sea level) that interact directly with an ocean. This would exclude lakes and rivers over land and so would only be consistent with the use of "sea_water" in standard names if the standard names also were limited in this way.
—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you were mentioned.
--
NOAA/NOS
US Integrated Ocean Observing System Office
1315 East-West Highway
Silver Spring MD 20910