Standard names: *Invert Barometer : Ocean surface elevation due to atmospheric pressure*

abiardeau commented 4 months ago

Hello,

I am Aurore BIARDEAU from Mercator Ocean international, working for Copernicus Marine service. Date : 07-03-2024

We are in the process of implementing a new variable in the Copernicus Marine service, and we would be happy to have your feedback on our proposal :

standard_name : sea_surface_height_amplitude_due_to_invert_barometer unit : m definition : The specification of a physical process by the phrase due_to_process means that the quantity named is a single term in a sum of terms which together compose the general quantity named by omitting the phrase. "Invert barometer" means variations in air pressure that give rise to corresponding variations in sea surface topography.

The definition is inspired by the definition of the standard_name sea_surface_height_correction_due_to_air_pressure_at_low_frequency, which corresponds to the IB correction in level 2 altimetry data.

I can provide more information from my expert colleagues if needed,

Kind regards, Aurore

github-actions[bot] commented 4 months ago

Thank you for your proposal. These terms will be added to the cfeditor (http://cfeditor.ceda.ac.uk/proposals/1) shortly. Your proposal will then be reviewed and commented on by the community and Standard Names moderator.

efisher008 commented 4 months ago

Hi Aurore,

I have added the name to the editor, you can view it here: https://cfeditor.ceda.ac.uk/proposal/5277/edit

I will wait for others to comment on this proposed name. As a starting point I would suggest adding the following phrase for sea surface height: "Sea surface height" is a time-varying quantity.".

Best regards, Ellie

efisher008 commented 2 months ago

Hi Aurore,

Just a note that I think it will be worth keeping track of the discussion relating to the use of invert_barometer in this name, which is happening in issue #285. As the two names are closely related and we want to be consistent in the format and definitions.

Best wishes, Ellie

abiardeau commented 1 month ago

Hi Ellie, Thanks for your feedback, To be consistent with issue #285, we could apply this correction : sea_surface_height_above_geoid_amplitude_due_to_invert_barometer

What do you think ? Thanks, Aurore

JonathanGregory commented 1 month ago

Dear Aurore @abiardeau

Please could you explain what you mean by "amplitude" in this proposed standard name and the two similar ones (#287 #288)?

Thanks

Jonathan

abiardeau commented 1 month ago

Dear @JonathanGregory After discussions, we agree that amplitude has no use here (and in the two similar ones), we can remove it from the standard name :) We can write inverse instead of invert indeed. Then, the proposition is now sea_surface_height_above_geoid_due_to_inverse_barometer

JonathanGregory commented 1 month ago

I still have a conceptual difficulty in understanding these proposals, because "sea surface height above height" is the level of a geophysical surface. It's not really "due to" something, like (for instance) the altitude of the tropopause and the density of sea water aren't due to something. The sea surface height, the tropopause and sea water density are always present. "Due to" is used to describe processes, changes and transient phenomena, not states. You must mean the change in SSH, I think, but with respect to what? When this quantity is zero, at what level is the SSH? It's not the geoid, I think, because atmosphere pressure variations are not the reason that SSH departs from the geoid.

slawchune commented 1 month ago

Hi Jonathan, I'm working with Aurore on the production of this new data. If I understand correctly, the current name of the variable might suggest that the “due to” is related to the reference height “the geoid”.

But this is not the case.

The “due to invert barometer” refers to the variable itself. In fact, it's the contribution of the inverse barometer to the sea surface heigh (the atmospheric pressure that presses on the ocean in the case of a high pressure system, for example, and causes water levels to drop).

Is “sea_surface_height_due_to_inverse_barometer_above_geoid” better?

I'm am not sure that a reference level is necessary. This field is more about something missing to our ocean model system than can be added afterwards. If this contribution is 0, the total sea level remains unchanged.

Please tell me if that answer your question.

Best regards, Stéphane

JonathanGregory commented 1 month ago

Unfortunately this morning I wrote the comment below in #311 instead of here by mistake, and misled Stéphane @slawchune and Karl @taylor13 into replying there as well. Sorry about that! I'm copying the messages here, where they belong.

At 0837 UTC on 31st May, I wrote

Dear Stéphane @slawchune

Thanks for your explanation. That clarifies what you mean for me i.e. the IB correction itself, not the corrected SSH. In that case, as you say, a reference level is not needed. We have two existing standard names

sea_surface_height_correction_due_to_air_pressure_and_wind_at_high_frequency sea_surface_height_correction_due_to_air_pressure_at_low_frequency

On the pattern of these, I think the quantity you mean would be named sea_surface_height_correction_due_to_air_pressure. However, others have made the comment, on our sea level terminology paper, that "correction" isn't a good phrase to use. The data is no more "correct" if the "correction" is applied; it's just a different quantity! Hence I will propose renaming these existing standard names.

Would you and others find sea_surface_height_increase_due_to_air_pressure_decrease a clear name for this quantity?

Best wishes

Jonathan

At 1003 UTC, @slawchune wrote:

Hi Jonathan, thank you very much with your help with this !

I would prefer "sea_surface_height_correction_due_to_air_pressure" because it stays general.

"sea_surface_height_increase_due_to_air_pressure_decrease" gives the idea that only low pressure system are considered.

actually we compute the inverted barometer at a 1h frequency. So do you think it would make sense to use "sea_surface_height_correction_due_to_air_pressure_and_wind_at_high_frequency"

Cheers, Stéphane

At 1435 UTC, @taylor13 wrote:

On the small point of "change" vs. "rise" for sea level, I think we should be explicit as to the sign, so "rise" would be better.

Similarly, we should find a way to word "sea_surface_height_correction_due_to_air_pressure_and_wind_at_high_frequency" to make it also unambiguous. I agree with Jonathan, that "correction" is vague (and possibly inappropriate), and that we need to specify what is considered "positive" vs. "negative". And "high frequency" depends entirely on context, which leaves it open to misinterpretation.

The frequency should be evident from the time coordinate attached to the variable, so I would leave the "frequency part" off.

Perhaps "adjustment" rather than "correction" might somehow be incorporated?

JonathanGregory commented 1 month ago

Dear Stéphane @slawchune and Karl @taylor13

The phrase "at high frequency" sounds vague to me, as it does to Karl, but the definition of sea_surface_height_correction_due_to_air_pressure_and_wind_at_high_frequency defines it precisely, as referring to timescales shorter than 20 days. If you're correcting for wind as well as IB, I agree that this existing standard name would be the right one for you to use, Stéphane. If it's just air pressure you consider, you would need a new standard name of sea_surface_height_correction_due_to_air_pressure_at_high_frequency, corresponding to the existing standard name of sea_surface_height_correction_due_to_air_pressure_at_low_frequency. It's fine to use these names, because if we decide on replacements for them, they will be aliased to the replacements; we do not invalidate existing standard names, for the sake of data already written. By the way, I notice that only the definition of the low frequency name (>20 days) says it's also called "inverse barometer".

As a separate matter, like Karl, I think we ought to work out better standard names for these quantities, for two reasons:

The sign convention is not clearly stated, and "correction" isn't a good word since SSH without these "corrections" is also a valid quantity - in fact, the quantity which is measured. I understand your concern about the implication that "negative air pressure" might exclude the other sign (though that's not what I intended). What do you think of sea_surface_depression_due_to_change_in_air_pressure? This is a positive number for a negative SSH anomaly (due to a positive anomaly in air pressure). Is that the sign convention you use for the "correction"?
It would be better to specify the timescale with a scalar coordinate, though we could retain 20 days as a default, if that is common practice. I assume this refers to a filter frequency or timescale for air pressure variations (low-pass or high-pass) - is that right, Stéphane?

Best wishes

Jonathan

slawchune commented 1 month ago

Hi Jonathan,

sorry it took me a while to reply.

Regarding the time scale, we use an analytical relationship of calculation with respect to atmospheric forecasts (surface pressure) in real time produced by the ECMWF. I'm not sure if it's appropriate to suggest a high/low frequency cutoff in the definition?
“sea_surface_depression_due_to_change_in_air_pressure” seems good to me, but there may be a sign error. This variable should be negative (respectively positive) when there's a positive (respectively negative) pressure anomaly. The variable should then be added to the total elevation term without changing its sign.

Does this make sense to you?

Best wishes Stéphane

JonathanGregory commented 1 month ago

Dear Stéphane

I agree with you that it is not necessary to specify the timescale in the definition or the standard name.

You're right, I was assuming the opposite sign i.e. a positive number, meaning SSH is lower than average, if the air pressure anomaly is positive, meaning higher than average. Should we make it change_in_sea_surface_height_due_to_change_in_air_pressure? There are many standard names with change_in, where the sign convention is the obvious one, that a positive number means the quantity increases. Hence this name should indicate a quantity whose values are negative, because SSH decreases when air pressure increases. Is that what you intend?

Best wishes

Jonathan

slawchune commented 1 month ago

Hi Jonathan; I think It's perfect ;)

Does it work for @efisher008 and @taylor13 ?

Cheers !

JonathanGregory commented 1 month ago

Oh good, I'm glad that sounds right. Meanwhile, I was working on #285, which made me think of another version of this one, namely change_in_sea_surface_height_due_to_air_pressure_anomaly. We already have air_pressure_anomaly as a standard name, meaning the difference from climatological air pressure.

larsbarring commented 1 month ago

As an outsider, I can only add that the new versions including "air_pressure" are much easier to understand! Many thanks @JonathanGregory

taylor13 commented 1 month ago

Regarding Jonathan's suggestion change_in_sea_surface_height_due_to_change_in_air_pressure. I fear it's not obvious what the sign should be (to writers or readers of data). Apparently the "change_in_air_pressure" has no sign assigned to it (i.e., we're talking about its absolute value or just that there is some air pressure change), but the "change_in_sea_surface_height" would be positive when there is an increase in sea surface height and negative when there is a decrease. I think having no sign associated with "change in air pressure", but have an sign determined by convention for "change in sea surface height" could be confusing.

If both changes had an implied sign, then I think the resulting quantity would be invariably negative.

Perhaps we can go with Jonathan's suggested standard name, but in the definition make it clear that the result is expected to be positive when the change in air pressure is negative and negative when the change in air pressure is negative.

Or perhaps, I'm just needlessly confused.

JonathanGregory commented 1 month ago

Dear Karl

Earlier I suggested specifying the sign of both quantities with constructions like sea_surface_height_increase_due_to_air_pressure_decrease, but Stéphane felt that it could be misunderstood as implying that only decreases in air pressure were included. How do you react to change_in_sea_surface_height_due_to_air_pressure_anomaly? There is a standard name of air_pressure_anomaly already. Its sign convention is not explicitly stated in the name, but would anyone have a doubt about its sign i.e. that this is a positive number if air pressure is higher than the climatological mean? If the sign of air_pressure_anomaly is obvious, I think the sign of change_in_sea_surface_height_due_to_air_pressure_anomaly ought to be clear. You get a negative change in SSH for a positive air pressure anomaly. I agree that this should also be stated in the definition, for the avoidance of doubt.

Best wishes

Jonathan

taylor13 commented 1 month ago

Thanks for the further explanation. An explicit statement in the definition would certainly be sufficient to satisfy me. Thanks.

JonathanGregory commented 4 weeks ago

OK, thanks, Karl. My preference is for change_in_sea_surface_height_due_to_air_pressure_anomaly, because I think it helps avoid the sign ambiguity. What do you think, Stéphane @slawchune and Ellie @efisher008?

JonathanGregory commented 4 weeks ago

I've changed my mind again, back to change_in_sea_surface_height_due_to_change_in_air_pressure, because this would allow us to have an analogous change_in_mean_sea_level_due_to_change_in_air_pressure for climate-change timescales. On those timescales, the change in pressure can't be described as air_pressure_anomaly, because an anomaly is defined wrt a climatology, and doesn't make sense when climate is changing.

slawchune commented 3 weeks ago

Hi Jonathan, I think we can go with "change_in_sea_surface_height_due_to_change_in_air_pressure" if it helps generalize this quantitiy :) I think it avoid also to think too much about the inversion sign relationship between air pressure and the sea surface height response.

efisher008 commented 1 week ago

Dear Stéphane,

I have changed the name to the agreed suggestion, and also removed the phrase describing "invert barometer" as this was not referenced in the name. The proposal can be viewed in the editor here: https://cfeditor.ceda.ac.uk/proposal/5277/edit

Would you like your name to be added to the author for these proposals along with Aurore (https://github.com/cf-convention/discuss/issues/285, https://github.com/cf-convention/discuss/issues/286, https://github.com/cf-convention/discuss/issues/287, https://github.com/cf-convention/discuss/issues/288, and https://github.com/cf-convention/discuss/issues/314), as you have had significant input on them?

Best regards, Ellie

cf-convention / discuss