Dear Andy

Thanks for this, and to Roy for support. I agree, I think we've pretty much
worked out what we mean now.

Instead of chart_datum I would prefer that we had standard names mentioning
the specific datums you want e.g. lowest_astronomical_tide, because it is more
informative and reduces the need for other information. Or is there a need to
be vague about it?

We have a mechanism, viz grid_mapping, to identify datums more precisely. If
we can't do that currently for MSL we could extend the convention.

As well as steric changes, I think non-tidal includes ocean dynamics (on all
timescales). I think your definitions look right.

Best wishes

Jonathan

----- Forwarded message from "Saulter, Andrew" 
<[email protected]> -----

> Date: Tue, 8 May 2018 08:08:13 +0000
> From: "Saulter, Andrew" <[email protected]>
> To: "[email protected]" <[email protected]>
> Subject: Re: [CF-metadata] proposed new standard name for storm surge
>       residual
> 
> Dear Jonathon,
> 
> Thanks for all the constructive inputs - this feels like we are getting 
> somewhere :-)
> 
> I'd be very happy with "non_tidal_elevation_of_sea_surface_height"; elevation 
> references positive when SSH is increased but also acknowledges suppression 
> of SSH with high pressure/negative surge; non-tidal gives us the very useful 
> catchall that there are various processes at play.
> 
> Regards the tidal SSH, you are absolutely right about assumptions. Normally 
> the tidal predictions we would be using (either directly or as boundary 
> inputs to a model) would be derived from a harmonic analysis of data covering 
> a given averaging period (varies, but anything from a month as a minimum that 
> captures a single spring-neap cycle, to 19-odd years to capture a full set of 
> sun-moon variations that cause the very largest tides). It is generally 
> assumed that the averaging will cancel out many of the other ocean effects - 
> any errors where this hasn't happened correctly will manifest themselves in 
> the non_tidal part (which is why I'm comfortable that this name makes more 
> sense than surge for observed values).
> 
> Being the wild optimist that I am, I've taken the below and tried to flesh it 
> out a bit for the tide and non-tidal variables we would have in our use 
> cases. Please let me know what you reckon - the descriptions might still need 
> some work?
> 
>  tidal_sea_surface_height_above_mean_sea_level
> Units: m
> "Sea surface height" is a time-varying quantity. "Height_above_X" means the 
> vertical distance above the named surface X. "Mean sea level" means the time 
> mean of sea surface elevation at a given location over an arbitrary period 
> sufficient to eliminate the tidal signals. The tidal component of sea surface 
> height describes the predicted variability of the sea surface due to 
> astronomic forcing (chiefly lunar and solar cycles) and shallow water 
> resonance of tidal components; for example as generated based on harmonic 
> analysis, or resulting from the application of harmonic tidal series as 
> boundary conditions to a numerical tidal model.
> 
> tidal_sea_surface_height_above_chart_datum
> Units: m
> "Sea surface height" is a time-varying quantity. "Height_above_X" means the 
> vertical distance above the named surface X. "Chart datum" describes a local 
> vertical reference from which depths displayed on a nautical chart are 
> measured and which differs from mean sea level. For example, chart datum 
> based on "lowest astronomical tide" or "mean lower low water". The tidal 
> component of sea surface height describes the predicted variability of the 
> sea surface due to astronomic forcing (chiefly lunar and solar cycles) and 
> shallow water resonance of tidal components; for example as generated based 
> on harmonic analysis, or resulting from the application of harmonic tidal 
> series as boundary conditions to a numerical tidal model.
> 
> non_tidal_elevation_of_sea_surface_height
> Units: m
> "Sea surface height" is a time-varying quantity. The "non_tidal_elevation" of 
> sea surface height describes the contribution made by processes other than 
> astronomic forcing of the ocean and shallow water resonance of tidal 
> components. These processes include storm surge effects, due to 
> meteorological forcing of the ocean and interaction between the generated 
> surge and tides, steric changes in the water column and, at higher 
> frequencies, effects of surface ocean waves. The contribution of these 
> processes vary according to the averaging time of the variable as described 
> by the "time_bounds" attribute.
> 
> Reading around on vertical datum a little more, the "above_mean_sea_level" 
> example should cover Ordnance Datum Newlyn (a MSL covering a specified period 
> at Newlyn and used as the UK's vertical reference) whilst "above_chart_datum" 
> should cover the LAT based predictions used for navigation in UK waters - 
> just need to specify these via "time_bounds" or "comment" attributes I think?
> 
> Cheers
> Andy
> 
> 
> > > > > > > > "Sea surface elevation" is a time-varying quantity denoting the 
> > > > > > > > height of the sea surface relative to a given datum. 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. Tides are a significant 
> > > > > > > > contributor to the observed sea surface height; here “tide” 
> > > > > > > > denotes a generic variable describing the time varying tidal 
> > > > > > > > signal, for example as generated based on a summation of 
> > > > > > > > harmonically analysed components, or resulting from the 
> > > > > > > > application of such components as boundary conditions to a 
> > > > > > > > numerical tidal model.
> 
> -----Original Message-----
> From: CF-metadata [mailto:[email protected]] On Behalf Of 
> Jonathan Gregory
> Sent: 04 May 2018 19:40
> To: [email protected]
> Subject: Re: [CF-metadata] proposed new standard name for storm surge residual
> 
> Dear Andy
> 
> Thanks for this. I have few comments to add because our thoughts are pretty 
> much the same, I'm glad to say.
> 
> I am a bit uncomfortable with the phrase "SSH due to X" anyway, because SSH 
> has no meaning except wrt a datum, not in an absolute sense. The phrase "due 
> to X"
> aptly describes contributions to something which could be zero in an absolute 
> sense, so you can add up all the contributions to get the whole.
> 
> SSH is a fairly precise geophysical quantity (the level of the interface 
> between air and water) - although not quite precise! - but as you say, the 
> value it takes depends on the meaning period. CF describes meaning periods 
> with time-bounds, so that's no problem.
> 
> "Residual" is maybe slightly jargon. We could spell it out more plainly.
> I would suggest
> 
> sea_surface_height_above_X = tidal_sea_surface_height_above_X + 
> non_tidal_elevation_of_sea_surface_height
> 
> Sometimes sea surface elevation is a synonym of sea surface height, but I 
> hope that "elevation of SSH" obviously means making SSH higher - does it to 
> you?
> Alternatively, "increase in SSH" might be better. I assume that this term 
> must include the seasonal non-tidal variation in SSH (due to the seasonal 
> cycle in ocean dynamics, density and mass).
> 
> I still have a query about the tidal SSH above X. It must include assumptions 
> for all the other effects - a particular surface pressure, surface winds, 
> ocean currents, etc. What are they? Perhaps they are all assumed to be annual 
> mean climatological values.
> 
> Best wishes
> 
> Jonathan
> 
> 
> ----- Forwarded message from "Saulter, Andrew" 
> <[email protected]> -----
> 
> > Date: Fri, 4 May 2018 08:15:00 +0000
> > From: "Saulter, Andrew" <[email protected]>
> > To: "[email protected]" <[email protected]>
> > Subject: Re: [CF-metadata] proposed new standard name for storm surge
> >     residual
> > 
> > Dear Jonathon,
> > 
> > Thought I was so close... :-)
> > 
> > But concerned that I am missing something, so I could do with some 
> > clarification before suggesting another iteration. So please can you 
> > comment on the below.
> > 
> > The thing that I am failing to understand is the relationship between 
> > sea_surface_height and the components that should make it up. So here is 
> > what I think I know:
> > 
> > 1. sea_surface_height should always be related to some vertical datum (MSL, 
> > geoid, arbitrary reference) in order to be useful. 
> > 
> > 2. sea_surface_height is a time-varying quantity so must be made up of 
> > a whole load of components (tide, steric effects, surge, ocean waves) 
> > - and the parameter name/description is vague enough that how we 
> > define this ends up in the eye of the beholder a bit; so a climate 
> > scientist working with deep ocean data might choose a long temporal 
> > averaging period that reduces sea surface height variations to 
> > seasonal and longer effects only, whilst I'd be using a 15 minute 
> > averaging period with my estimates so will need to include tide and 
> > surge
> > 
> > 3. at some point the sum of these components must include an offset (Z) 
> > that makes sea_surface_height relative to its vertical datum. This is where 
> > the problem seems to lie in our discussion so far. For my users the answer 
> > is simple - the referencing is included in the tide values and all other 
> > components are relative. For the model tide and MSL case this becomes 
> > simplified as Z=0 and the tide rises and falls around this point, but for 
> > lowest astronomic tide as a datum then we either need to consider Z as some 
> > finite value or understand that the tide component will nearly always be 
> > positive ('cos you don't get much lower than lowest astronomic tide).
> > 
> > 4. If you go with idea that tide is referenced against a datum, which 
> > I is really what I need to do in order to give my users a 
> > straightforward dataset, then this implies that tide values should 
> > link to a specific sea_surface_height_above_X (i.e. I can't link a 
> > tide referenced against LAT with an SSH referenced against MSL)
> > 
> > Its these last two points where 'due_to' implying a relative quantity is 
> > messing with my mind. From our conversations, I get than sense that 
> > 'due_to' really implies that any component process we define should be 
> > transferable universally between different sea_surface_height_above_Xs, and 
> > that this is because there is another value that defines what Z is. For 
> > this to work for tide and surge,  this should actually mean that 'due_to' 
> > implies 'relative to MSL for any processes occurring on a higher frequency 
> > than the MSL definition'?
> > 
> > If that is the case, then 'due_to' will work fine for surge (which is a 
> > transferable quantity) but not for tide in the context I want to use it in. 
> > So do we actually need something like:
> > 
> > sea_surface_height_above_X = tidal_sea_surface_height_above_X + 
> > sea_surface_height_due_to_surge
> > 
> > The other question arising from our discussions is the use of the word 
> > 'surge'. I'd originally proposed this as it’s a terminology my users will 
> > understand, but perhaps its not specific enough for the purposes of CF 
> > and/or I should describe it better. In observation terms, surge is nearly 
> > always 'what I get when I subtract a tidal prediction from my observed sea 
> > surface height'. This includes a myriad of effects, from the meteorological 
> > forced variations (seasonal, daily, the lot), to the tide-surge interaction 
> > effects and the errors inherent in the tidal predictions;  on this basis 
> > 'non tidal residual of sea surface height' is a more precise description, 
> > its just that we use the surge label because surge is generally expected to 
> > be the dominant effect. When we model, life is a lot neater and surge is 
> > more simply a combination of the met and tide-surge interactions and we 
> > call our model a 'surge model' rather than a 'NTROSSH model' as a result!
> > 
> > Thinking this through, I'd be comfortable with using 'non_tidal_residual' 
> > instead of 'surge' if that helped - but don't think this this should be a 
> > 'due_to' any more since we are not pinning the changes in sea surface 
> > height to a specific process. In other words we would have something like:
> > 
> > sea_surface_height_above_X = tidal_sea_surface_height_above_X + 
> > non_tidal_residual_of_sea_surface_height
> > 
> > Cheers
> > Andy
> > 
> > 
> > -----Original Message-----
> > From: CF-metadata [mailto:[email protected]] On Behalf 
> > Of Jonathan Gregory
> > Sent: 03 May 2018 17:52
> > To: [email protected]
> > Subject: [CF-metadata] proposed new standard name for storm surge 
> > residual
> > 
> > Dear Andy
> > 
> > > 'height_above_X' means the vertical distance above the named surface 
> > > X
> > > - there are a whole set of these for 'sea_surface_height' 
> > > ('above_geoid', 'above_geopotential_datum', 'above_mean_sea_level') 
> > > that set the vertical reference
> > 
> > Yes.
> > 
> > > Then, as I think we've all agreed, 'due_to_surge', 'due_to_tide' are 
> > > simply relative quantities.
> > 
> > Yes, I agree, such a definition is possible, where due_to_X means the 
> > difference between SSH with X and SSH without X.
> > 
> > > for my surge model, which is referenced to mean sea level and has a tide 
> > > and surge part that I can decompose, I could generate variables with 
> > > standard names:
> > > 
> > > 1. sea_surface_height_above_mean_sea_level - the total
> > > 
> > > 2.sea_surface_height_due_to_tide - the tide part
> > > 
> > > 3. sea_surface_height_due_to_surge - the surge part
> > 
> > And you also said
> > > a summation of these quantities should lead us to a 
> > > 'sea_surface_height_above_X', but on their own they are generic.
> > That's not true here, is it? (2) + (3) is not equal to 1. There must be 
> > another term (Z, say) in the sum, for SSH above MSL when there is no tide 
> > and no surge.
> > 
> > (1) SSH above MSL = (Z) SSH above MSL with no tide and no surge
> >   + (2) SSH elevation due to tide + (3) SSH elevation due to surge.
> > 
> > Or maybe what the model produces is Z + 2 i.e. SSH above MSL if there 
> > is no surge? We could call that 
> > sea_surface_height_above_mean_sea_level_assuming_no_surge
> > 
> > Or maybe Z=0? That is what I suggested initially about MSL, in effect.
> > 
> > Which term contains the seasonal cycle of SSH?
> > 
> > > the reference_datum should a) be stipulated and b) have a way of being 
> > > mapped to the reference_ellipsoid we would use for the coordinate system. 
> > > A) would need a new descriptive variable, e.g. 'reference_datum_name'. B) 
> > > already has a precedent in 'water_surface_reference_datum_altitude' but 
> > > actually using the existing 'height_above_reference_ellipsoid' might be 
> > > more appropriate (basically I'm not sure if a 
> > > 'reference_datum_height_above_reference_ellipsoid' is necessary)?
> > 
> > I'm not too keen on the generic reference_datum, and feel it would be 
> > better to add names for the specific datums you use, if they are 
> > geophysical surfaces.
> > If they are arbitrary benchmarks, I agree that some way to name them is 
> > needed.
> > 
> > > Hopefully I'm not talking total nonsense.
> > 
> > Of course not. Sorry this is so hard.
> > 
> > Best wishes
> > 
> > Jonathan
> > 
> > > -----Original Message-----
> > > From: CF-metadata [mailto:[email protected]] On 
> > > Behalf Of Jonathan Gregory
> > > Sent: 02 May 2018 14:27
> > > To: [email protected]
> > > Subject: [CF-metadata] proposed new standard name for storm surge 
> > > residual
> > > 
> > > Dear Andy
> > > 
> > > Thanks for your email. This is surprisingly mind-bending. Although MSL 
> > > could mean time-means on various periods, I believe that when we refer to 
> > > it as a surface in CF standard names we mean a very long-term mean, to 
> > > get rid of all variations. Of course, that's still not well-defined, 
> > > because on very long timescales other things change like climate and 
> > > ocean basin bathymetry.
> > > 
> > > I don't think this is the point at issue. What I'm struggling with is 
> > > whether the elevation of the sea surface due to tide has a datum (MSL, 
> > > reference ellipsoid, geoid, etc.) or not. If, like "due to surge", it has 
> > > no datum, it means the difference between SSH with tide and without tide. 
> > > What does "without tide" mean, then? It could mean "with permanent tide 
> > > but no time-varying tide", for instance. If you include a datum you get 
> > > something like "elevation of sea surface above reference elliposid due to 
> > > tide", but I'm not sure what the attribution to tide means in that case. 
> > > What's the difference between this quantity and 
> > > sea_surface_height_above_reference_ellipsoid, which is already a standard 
> > > name? The difference would seem to be the part "due to tide". But that 
> > > returns us to the question of what "without the tide" means.
> > > 
> > > Best wishes
> > > 
> > > Jonathan
> > > 
> > > 
> > > 
> > > ----- Forwarded message from "Saulter, Andrew" 
> > > <[email protected]> -----
> > > 
> > > > Date: Wed, 25 Apr 2018 07:08:01 +0000
> > > > From: "Saulter, Andrew" <[email protected]>
> > > > To: "[email protected]" <[email protected]>
> > > > Subject: Re: [CF-metadata] proposed new standard name for storm surge
> > > >         residual
> > > > 
> > > > Thanks Jonathon,
> > > > 
> > > > From the below:
> > > > 
> > > > We use the term "sea" - agreed. "due_to_surge" has no need for a datum 
> > > > reference - agreed.
> > > > 
> > > > MSL implies no tide and no surge. I'd disagree with this; sea level at 
> > > > high frequency will comprise contributions from lots of different 
> > > > components and mean sea level is therefore a quantity where we have 
> > > > chosen to average these effects out as best as possible but they 
> > > > haven't gone away - for example a monthly mean sea level from a coastal 
> > > > tide gauge may still comprise some tidal signal (for example an 
> > > > asymmetry where the equinoxes occur in different months) and will 
> > > > certainly include variation due to a seasonal changes in the surge 
> > > > contribution.
> > > > 
> > > > So from my perspective, the only difference between tide and surge is 
> > > > that we would expect tide to always reference some form of fixed datum 
> > > > (which is preferably more flexible than just MSL) in order to allow us 
> > > > to construct a sea level series that is vertically referenced, whereas  
> > > > surge and other contributions will be relative quantities.
> > > > 
> > > > Cheers
> > > > Andy
> > > > 
> > > > 
> > > > 
> > > > -----Original Message-----
> > > > From: CF-metadata [mailto:[email protected]] On 
> > > > Behalf Of Jonathan Gregory
> > > > Sent: 24 April 2018 18:30
> > > > To: [email protected]
> > > > Subject: Re: [CF-metadata] proposed new standard name for storm 
> > > > surge residual
> > > > 
> > > > Dear Andy
> > > > 
> > > > > - I'm only going to do this for sea water levels, so from my 
> > > > > point of view using the term "sea" is fine; I'm just aware that 
> > > > > what comes below could be applied in other water bodies
> > > > 
> > > > Yes. However, we make our job simpler (as a principle in CF) by doing 
> > > > only what we need to for the current use-cases.
> > > > 
> > > > > - "due_to_surge" will either a) be derived as a residual value 
> > > > > calculated after taking a measured sea level value (referenced 
> > > > > to some fixed datum) and subtracting a predicted tide height 
> > > > > (referenced to same datum), or b) be a quantity that we would 
> > > > > expect to add to a predicted tide height in order to create a 
> > > > > total water level (again referenced to some fixed datum)
> > > > 
> > > > In both cases the datum is not relevant to the elevation due to surge.
> > > > 
> > > > > - "due_to_tide" will be the tide values mentioned above which will 
> > > > > have to be referenced against a datum or common benchmark, e.g. chart 
> > > > > datum, mean sea level, Ordnance Datum Newlyn, in order to make sense. 
> > > > 
> > > > ... whereas here the datum *is* required.
> > > > 
> > > > So these cases seem different after all, and may need different sorts 
> > > > of name - at least, that's my first reaction. It's because there isn't 
> > > > a situation of "no tide", but there is a situation of "no surge". On 
> > > > second thoughts, I'm not sure about this distinction. No tide, I 
> > > > suppose, means MSL. On the other hand, no surge isn't uniquely defined 
> > > > - something must be assumed about the MSLP and the wind when there 
> > > > *isn't* a surge. What is that?
> > > > 
> > > > > So far these are variables that give us what we might term 'still 
> > > > > water level', i.e. neglecting wave effects. However, thinking about 
> > > > > future requirements you could easily see an extension to higher 
> > > > > frequency parameters such as "due_to_wave_induced_setup" (minutes to 
> > > > > hours), "due_to_run_up" (seconds to minutes), "due_to_waves" 
> > > > > (seconds) if you were looking at a detailed approach to evaluating 
> > > > > total water levels. All these would work like surge, in that these 
> > > > > aren't referenced to a datum themselves but will contribute to some 
> > > > > total water level value that does need to be.
> > > > 
> > > > Right. If we can work out how to deal with the surge, I agree the 
> > > > others will follow when they are needed.
> > > > 
> > > > Best wishes
> > > > 
> > > > Jonathan
> > > > 
> > > > > Instinctively when I plot summations of these types of variables in 
> > > > > time-series I would write 'sea_surface_elevation' on the y-axis 
> > > > > (since the water goes down as well as up) but that, definitely, is 
> > > > > just me! Personally I have no objection to "elevation_of_sea_surface" 
> > > > > either - it seems clear what it means and if we are all happy that 
> > > > > "sea" can be generic for "water" I'd be good with this.
> > > > > 
> > > > > Thanks
> > > > > Andy
> > > > > 
> > > > > 
> > > > > -----Original Message-----
> > > > > From: CF-metadata [mailto:[email protected]] On 
> > > > > Behalf Of Jonathan Gregory
> > > > > Sent: 24 April 2018 17:08
> > > > > To: [email protected]
> > > > > Subject: [CF-metadata] Fwd: Re: Fwd: Re: proposed new standard 
> > > > > name for storm surge residual
> > > > > 
> > > > > Dear Andy
> > > > > 
> > > > > > "elevation_of_sea_surface_due_to_X" sounds most appropriate.
> > > > > OK.
> > > > > 
> > > > > Since we already have
> > > > >   water_surface_height_above_reference_datum
> > > > >   water_surface_reference_datum_altitude
> > > > > in the table, I agree that water_surface is OK to use. In general in 
> > > > > standard names we have made the word "sea" signify all bodies of 
> > > > > water, as we've not been able to find another neat and clear word for 
> > > > > them. However, we've already departed from that rule in this case. Do 
> > > > > you need to use these names for lakes?
> > > > > If your use is just sea, I would rather stick to sea names, since 
> > > > > we've got a lot more of those.
> > > > > 
> > > > > I would say that the reference_datum names should be avoided if your 
> > > > > datum is something that can be geophysically defined, such as 
> > > > > mean_sea_level or the geoid. We have names mentioning those levels, 
> > > > > which are more specific and useful. I think the reference_datum names 
> > > > > are for arbitrary levels, indicated by some physical benchmark.
> > > > > 
> > > > > I feel that
> > > > >   elevation_of_water|sea_surface would be better than
> > > > >   water|sea_surface_elevation
> > > > > because to me the former sounds like "making the water surface 
> > > > > higher", which is what we mean, while the latter means "how high the 
> > > > > water surface is". That is rather subtle and language-dependent, so 
> > > > > I'm a bit nervous about it. It also might just be me! How does it 
> > > > > sound to you?
> > > > > 
> > > > > Best wishes
> > > > > 
> > > > > Jonathan
> > > > > 
> > > > > > 
> > > > > > -----Original Message-----
> > > > > > From: CF-metadata [mailto:[email protected]] On 
> > > > > > Behalf Of Jonathan Gregory
> > > > > > Sent: 24 April 2018 14:26
> > > > > > To: [email protected]
> > > > > > Subject: [CF-metadata] Fwd: Re: proposed new standard name for 
> > > > > > storm surge residual
> > > > > > 
> > > > > > Dear Andrew and John
> > > > > > 
> > > > > > I hadn't noticed that sea_surface_elevation is already in use as an 
> > > > > > alias.
> > > > > > That's a pity, but maybe it would be confusing anyway, given John's 
> > > > > > comment.
> > > > > > 
> > > > > > I think that what Andrew needs is terms that say how much higher 
> > > > > > the sea surface is because of influence X relative to how high it 
> > > > > > would be in the absence of influence X. Such terms do not need any 
> > > > > > datum (like geoid or MSL). The difference in z is the same 
> > > > > > regardless of what datum would be used for z itself. I suggested 
> > > > > > before that change_in would be a possibility but it doesn't sound 
> > > > > > quite right, because we aren't comparing SSH before and after a 
> > > > > > storm surge for example, which is what I'd understand by "change in 
> > > > > > SSH due to storm surge". Other ideas:
> > > > > > 
> > > > > > elevation_of_sea_surface_due_to_X 
> > > > > > increment_to_sea_surface_height_due_to_X
> > > > > > increase_of_sea_surface_height_due_to_X
> > > > > > 
> > > > > > What others occur to you?
> > > > > > 
> > > > > > Best wishes
> > > > > > 
> > > > > > Jonathan
> > > > > > 
> > > > > > 
> > > > > > ----- Forwarded message from "Saulter, Andrew" 
> > > > > > <[email protected]> -----
> > > > > > 
> > > > > > > Date: Tue, 24 Apr 2018 07:17:48 +0000
> > > > > > > From: "Saulter, Andrew" <[email protected]>
> > > > > > > To: "[email protected]" <[email protected]>
> > > > > > > Subject: Re: [CF-metadata] proposed new standard name for storm 
> > > > > > > surge
> > > > > > >   residual
> > > > > > > 
> > > > > > > John,
> > > > > > > 
> > > > > > > I see where you are with that, but my understanding from Jonathon 
> > > > > > > Gregory's email earlier is that the 'due_to' part of the phrasing 
> > > > > > > identifies a component process that contributes to an overall 
> > > > > > > quantity. In the case below 'due_to_storm_surge' is a 
> > > > > > > contribution to 'sea_surface_elevation' and that quantity is what 
> > > > > > > needs to be referenced to some datum. Or maybe I'm not getting 
> > > > > > > it? Steep learning curve this...
> > > > > > > 
> > > > > > > Anyway, having thought about datum's now I have done some further 
> > > > > > > searching and noted the following already exist as standard names:
> > > > > > > 
> > > > > > > water_surface_height_above_reference_datum - this denotes 
> > > > > > > the quantity
> > > > > > > 
> > > > > > > water_surface_reference_datum_altitude - references the 
> > > > > > > datum to the
> > > > > > > (grid_mapping) geoid
> > > > > > > 
> > > > > > > These look much more like what I was after, so the question is 
> > > > > > > can the 'due_to_storm_surge' and 'due_to_tide' be sensibly 
> > > > > > > appended to 'water_surface_height_above_reference_datum'??
> > > > > > > 
> > > > > > > Cheers
> > > > > > > Andy
> > > > > > > 
> > > > > > > 
> > > > > > > -----Original Message-----
> > > > > > > From: John Graybeal [mailto:[email protected]]
> > > > > > > Sent: 23 April 2018 17:57
> > > > > > > To: Saulter, Andrew <[email protected]>
> > > > > > > Cc: CF Metadata List <[email protected]>
> > > > > > > Subject: Re: [CF-metadata] proposed new standard name for 
> > > > > > > storm surge residual
> > > > > > > 
> > > > > > > 
> > > > > > > I actually find this new name/definition internally inconsistent. 
> > > > > > > An elevation that is ‘due to storm surge’ seems to be 
> > > > > > > relative to the elevation without the storm surge, which makes 
> > > > > > > the datum irrelevant.
> > > > > > > Unless the change due to the storm surge would be measured 
> > > > > > > differently under different datums, but I can’t imagine that.
> > > > > > > (Taking the other way, if it’s an elevation relative to some 
> > > > > > > normal datum, then “due to storm surge” is irrelevant.)
> > > > > > > 
> > > > > > > In any case, under the new definition, the description needs to 
> > > > > > > include exactly how the datum is specified. The computers and 
> > > > > > > people will need to know where to look for that information, and 
> > > > > > > ideally it should be a unique identifier that the computers can 
> > > > > > > recognize and understand.
> > > > > > > 
> > > > > > > 
> > > > > > > john
> > > > > > > 
> > > > > > > 
> > > > > > > > On Apr 23, 2018, at 01:43, Saulter, Andrew 
> > > > > > > > <[email protected]> wrote:
> > > > > > > > 
> > > > > > > > Apologies, a little bit more to add to the below following 
> > > > > > > > up from Jonathon's first email,
> > > > > > > > 
> > > > > > > > For both tide and surge I would actually prefer to go with 
> > > > > > > > Jonathon's suggestion that the 'height_above_mean_sea_level' 
> > > > > > > > part of my suggestions is replaced with 'elevation'. This is a 
> > > > > > > > much more compact and flexible way of expressing things and 
> > > > > > > > means, particularly with tide that we can reference this to 
> > > > > > > > whichever datum we like (for example Chart Datum, Ordnance 
> > > > > > > > Datum, MSL) dependent on source elsewhere in the metadata. I 
> > > > > > > > think it is also appropriate that we think of 
> > > > > > > > "sea_surface_elevation" as a quantity that can be contributed 
> > > > > > > > to via processes with many different timescales, e.g. tides, 
> > > > > > > > surges, individual ocean waves.
> > > > > > > > 
> > > > > > > > This would take us to:
> > > > > > > > 
> > > > > > > > Proposed standard name: 
> > > > > > > > sea_surface_elevation_due_to_storm_surge
> > > > > > > > Units: m
> > > > > > > > "Sea surface elevation" is a time-varying quantity denoting the 
> > > > > > > > height of the sea surface relative to a given datum. 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. Storm surge effects, due to 
> > > > > > > > meteorological forcing of the ocean and interaction between the 
> > > > > > > > generated surge and tides, are a significant contributor to the 
> > > > > > > > observed sea surface height.
> > > > > > > > 
> > > > > > > > Proposed standard name: 
> > > > > > > > sea_surface_elevation_due_to_tide
> > > > > > > > Units: m
> > > > > > > > "Sea surface elevation" is a time-varying quantity denoting the 
> > > > > > > > height of the sea surface relative to a given datum. 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. Tides are a significant 
> > > > > > > > contributor to the observed sea surface height; here “tide” 
> > > > > > > > denotes a generic variable describing the time varying tidal 
> > > > > > > > signal, for example as generated based on a summation of 
> > > > > > > > harmonically analysed components, or resulting from the 
> > > > > > > > application of such components as boundary conditions to a 
> > > > > > > > numerical tidal model.
> > > > > > > > 
> > > > > > > > However, I have one concern in that "sea_surface_elevation" is 
> > > > > > > > presently given as an alias for 
> > > > > > > > "sea_surface_height_above_geoid". My worry is that the latter 
> > > > > > > > has implications for the vertical datum and that we might 
> > > > > > > > choose to disconnect this from other aspects of the 
> > > > > > > > grid_mapping variable (e.g. where my station positions are in 
> > > > > > > > WGS84, but the vertical reference is to chart datum) in which 
> > > > > > > > case we are not strictly referencing against the geoid any 
> > > > > > > > more. In addition, the term "sea_surface_height" has more 
> > > > > > > > usually been associated with altimeter and model products where 
> > > > > > > > high frequency signals are generally excluded? 
> > > > > > > > 
> > > > > > > > So some consensus as to whether "sea_surface_elevation" is the 
> > > > > > > > phrasing to go for would be very helpful...
> > > > > > > > 
> > > > > > > > Cheers
> > > > > > > > Andy
> > > > > > > > 
> > > > > > > > -----Original Message-----
> > > > > > > > From: CF-metadata 
> > > > > > > > [mailto:[email protected]]
> > > > > > > > On Behalf Of Saulter, Andrew
> > > > > > > > Sent: 20 April 2018 17:04
> > > > > > > > To: [email protected]
> > > > > > > > Subject: Re: [CF-metadata] proposed new standard name for 
> > > > > > > > storm surge residual
> > > > > > > > 
> > > > > > > > Jonathon, Helen,
> > > > > > > > 
> > > > > > > > Thanks for the feedback.
> > > > > > > > 
> > > > > > > > I'd looked at the existing 'sea_surface_height' terms but had 
> > > > > > > > the same worry as Jonathon that the use of 'amplitude' 
> > > > > > > > restricted these to some (unspecified) time integral. What I'm 
> > > > > > > > after is definitely a variable that varies as a function of 
> > > > > > > > time. It's also unusual in the coastal forecasting community to 
> > > > > > > > want to split the various contributions to tide up.
> > > > > > > > 
> > > > > > > > The 'due_to_air_pressure_and_wind' term captures the primary 
> > > > > > > > meteorological processes that induce surge. However, these do 
> > > > > > > > not capture the effect of tide-surge interaction in shallower 
> > > > > > > > waters (for example the extra surge elevation enhances the 
> > > > > > > > speed at which the tide propagates so a 'surge residual' can 
> > > > > > > > include the propagation speed delta as well as the background 
> > > > > > > > super-elevation) nor other secondary variability that we often 
> > > > > > > > see in surge residuals, such as steric changes of the water 
> > > > > > > > column. So I feel that using a catchall term 'storm_surge', 
> > > > > > > > although less specific would have a lot less potential to 
> > > > > > > > mislead a user. The option exists, I assume, in the comments 
> > > > > > > > attribute for a variable to be more precise about its 
> > > > > > > > derivation/generating processes.
> > > > > > > > 
> > > > > > > > So overall, I couldn't find a goldilocks term for either surge 
> > > > > > > > or tide that would fit my users understanding of the variables 
> > > > > > > > - hence the new suggestions.
> > > > > > > > 
> > > > > > > > Have a good weekend
> > > > > > > > Andy
> > > > > > > > 
> > > > > > > > 
> > > > > > > > 
> > > > > > > > -----Original Message-----
> > > > > > > > From: CF-metadata 
> > > > > > > > [mailto:[email protected]]
> > > > > > > > On Behalf Of Jonathan Gregory
> > > > > > > > Sent: 11 April 2018 18:37
> > > > > > > > To: [email protected]
> > > > > > > > Subject: [CF-metadata] proposed new standard name for 
> > > > > > > > storm surge residual
> > > > > > > > 
> > > > > > > > Dear Helen and Andy
> > > > > > > > 
> > > > > > > > I noticed the sea_surface_height_amplitude_due_to_X_tide names 
> > > > > > > > as well, and I wondered, what does "amplitude" mean here? The 
> > > > > > > > definitions of these names don't say, and I feel that we should 
> > > > > > > > be clear. I guessed it might mean the amplitude of SSH due to 
> > > > > > > > the tidal cycle, whereas I think Andy means the actual tidal 
> > > > > > > > height as a function of time. Are you able to clarify?
> > > > > > > > 
> > > > > > > > It's a good point about due_to_air_pressure[_and_wind], thanks. 
> > > > > > > > That may not obviously mean "storm surge", which maybe could be 
> > > > > > > > inserted in the definition.
> > > > > > > > 
> > > > > > > > Best wishes
> > > > > > > > 
> > > > > > > > Jonathan
> > > > > > > > 
> > > > > > > > ----- Forwarded message from "Snaith, Helen M." 
> > > > > > > > <[email protected]>
> > > > > > > > -----
> > > > > > > > 
> > > > > > > >> Date: Tue, 10 Apr 2018 13:14:16 +0000
> > > > > > > >> From: "Snaith, Helen M." <[email protected]>
> > > > > > > >> To: "Saulter, Andrew" <[email protected]>
> > > > > > > >> CC: "[email protected]" <[email protected]>
> > > > > > > >> Subject: Re: [CF-metadata] proposed new standard name for 
> > > > > > > >> storm surge
> > > > > > > >>        residual
> > > > > > > >> x-mailer: Apple Mail (2.3445.6.18)
> > > > > > > >> 
> > > > > > > >> Hi Andy
> > > > > > > >> 
> > > > > > > >> Many of the sea_surface_height terms have been used in 
> > > > > > > >> satellite altimetry for some time.
> > > > > > > >> The tidal components have been split out into
> > > > > > > >> sea_surface_height_amplitude_due_to_equilibrium_ocean_tide<javascript:
> > > > > > > >> void(0)>
> > > > > > > >> sea_surface_height_amplitude_due_to_geocentric_ocean_tide<javascript:
> > > > > > > >> v
> > > > > > > >> oid(0)>
> > > > > > > >> sea_surface_height_amplitude_due_to_non_equilibrium_ocean
> > > > > > > >> _t
> > > > > > > >> id
> > > > > > > >> e<
> > > > > > > >> ja
> > > > > > > >> va
> > > > > > > >> sc
> > > > > > > >> r
> > > > > > > >> ipt:void(0)>
> > > > > > > >> 
> > > > > > > >> And the pole tide
> > > > > > > >> sea_surface_height_amplitude_due_to_pole_tide<javascript:
> > > > > > > >> vo
> > > > > > > >> id
> > > > > > > >> (0
> > > > > > > >> )>
> > > > > > > >> 
> > > > > > > >> In these terms, amplitude has been used to identify the 
> > > > > > > >> ‘above mean level’ and sea_surface_height is as alias of 
> > > > > > > >> sea_surface_heigth_above_mean_sea_level
> > > > > > > >> 
> > > > > > > >> 
> > > > > > > >> Also included are the terms 
> > > > > > > >> sea_surface_height_correction_due_to_air_pressure_and_win
> > > > > > > >> d_
> > > > > > > >> at
> > > > > > > >> _h
> > > > > > > >> ig
> > > > > > > >> h_
> > > > > > > >> fr
> > > > > > > >> e
> > > > > > > >> quency<javascript:void(0)> 
> > > > > > > >> sea_surface_height_correction_due_to_air_pressure_at_low_
> > > > > > > >> fr
> > > > > > > >> eq
> > > > > > > >> ue
> > > > > > > >> nc
> > > > > > > >> y<
> > > > > > > >> ja
> > > > > > > >> v
> > > > > > > >> ascript:void(0)>
> > > > > > > >> 
> > > > > > > >> The former of which is related to surge I think - it is 
> > > > > > > >> normally determined from a tidal model and is the response of 
> > > > > > > >> sea level to changes in air pressure and wind.
> > > > > > > >> 
> > > > > > > >> Even if these are not the correct terms, as you are not 
> > > > > > > >> determining a 'correction’ but a value - they should be 
> > > > > > > >> related to the surge components, so do they give the ‘due to’ 
> > > > > > > >> component you need?
> > > > > > > >> 
> > > > > > > >> Helen
> > > > > > > >> 
> > > > > > > >> 
> > > > > > > >> On 4 Apr 2018, at 17:13, Saulter, Andrew 
> > > > > > > >> <[email protected]<mailto:[email protected]>>
> > > > > > > >>  wrote:
> > > > > > > >> 
> > > > > > > >> Dear all,
> > > > > > > >> 
> > > > > > > >> First posting to this list, so please forgive me if I’m 
> > > > > > > >> doing it wrong…
> > > > > > > >> 
> > > > > > > >> I’d like to request an addition to the standard name list to 
> > > > > > > >> include storm surge residual and tide. These variables are 
> > > > > > > >> generated for the purpose of coastal flood prediction and will 
> > > > > > > >> be available in future, netCDF based, operational products 
> > > > > > > >> from the Met Office.
> > > > > > > >> 
> > > > > > > >> Proposed standard name: 
> > > > > > > >> sea_surface_height_above_mean_sea_level_due_to_storm_surg
> > > > > > > >> e
> > > > > > > >> Units: m
> > > > > > > >> "Sea surface height" is a time-varying quantity. 
> > > > > > > >> "Height_above_X" means the vertical distance above the named 
> > > > > > > >> surface X. "Mean sea level" means the time mean of sea surface 
> > > > > > > >> elevation at a given location over an arbitrary period 
> > > > > > > >> sufficient to eliminate the tidal signals. 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. Storm surge effects, due to meteorological forcing 
> > > > > > > >> of the ocean and interaction between the generated surge and 
> > > > > > > >> tides, are a significant contributor to the observed sea 
> > > > > > > >> surface height.
> > > > > > > >> 
> > > > > > > >> Proposed standard name: 
> > > > > > > >> sea_surface_height_above_mean_sea_level_due_to_tide
> > > > > > > >> Units: m
> > > > > > > >> "Sea surface height" is a time-varying quantity. 
> > > > > > > >> "Height_above_X" means the vertical distance above the named 
> > > > > > > >> surface X. "Mean sea level" means the time mean of sea surface 
> > > > > > > >> elevation at a given location over an arbitrary period 
> > > > > > > >> sufficient to eliminate the tidal signals. 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. Tides are a significant contributor to the 
> > > > > > > >> observed sea surface height; here “tide” denotes a generic 
> > > > > > > >> variable describing the time varying tidal signal, for example 
> > > > > > > >> as generated based on a summation of harmonically analysed 
> > > > > > > >> components, or resulting from the application of such 
> > > > > > > >> components as boundary conditions to a numerical tidal model.
> > > > > > > >> 
> > > > > > > >> Many thanks
> > > > > > > >> Andy
> > > > > > > >> 
> > > > > > > >> 
> > > > > > > >> Andy Saulter
> > > > > > > >> Surge, Waves and Metocean Projects Manager Met Office 
> > > > > > > >> FitzRoy Road Exeter  Devon EX1 3PB
> > > > > > > >> Tel: +44 (0)1392 884703  Fax: +44 (0)1392 885681 
> > > > > > > >> [email protected]<mailto:andrew.saulter@met
> > > > > > > >> of
> > > > > > > >> fi
> > > > > > > >> ce
> > > > > > > >> .g
> > > > > > > >> ov
> > > > > > > >> .u
> > > > > > > >> k
> > > > > > > >>> http://www.metoffice.gov.uk<http://www.metoffice.gov.uk/
> > > > > > > >>> >
> > > > > > > >> 
> > > > > > > >> 
> > > > > > > >> --
> > > > > > > >> This message has been scanned for viruses and dangerous 
> > > > > > > >> content by MailScanner<http://www.mailscanner.info/>, and 
> > > > > > > >> is believed to be clean.
> > > > > > > >> _______________________________________________
> > > > > > > >> CF-metadata mailing list
> > > > > > > >> [email protected]<mailto:[email protected]>
> > > > > > > >> http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > > > > > > >> 
> > > > > > > >> ________________________________ This message (and any
> > > > > > > >> attachments) is for the recipient only. NERC is subject to the 
> > > > > > > >> Freedom of Information Act 2000 and the contents of this email 
> > > > > > > >> and any reply you make may be disclosed by NERC unless it is 
> > > > > > > >> exempt from release under the Act. Any material supplied to 
> > > > > > > >> NERC may be stored in an electronic records management system.
> > > > > > > >> ________________________________
> > > > > > > > 
> > > > > > > >> _______________________________________________
> > > > > > > >> CF-metadata mailing list
> > > > > > > >> [email protected]
> > > > > > > >> http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > > > > > > > 
> > > > > > > > 
> > > > > > > > ----- End forwarded message ----- 
> > > > > > > > _______________________________________________
> > > > > > > > CF-metadata mailing list
> > > > > > > > [email protected]
> > > > > > > > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > > > > > > > _______________________________________________
> > > > > > > > CF-metadata mailing list
> > > > > > > > [email protected]
> > > > > > > > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > > > > > > > _______________________________________________
> > > > > > > > CF-metadata mailing list
> > > > > > > > [email protected]
> > > > > > > > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > > > > > > 
> > > > > > > _______________________________________________
> > > > > > > CF-metadata mailing list
> > > > > > > [email protected]
> > > > > > > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > > > > > 
> > > > > > ----- End forwarded message ----- 
> > > > > > _______________________________________________
> > > > > > CF-metadata mailing list
> > > > > > [email protected]
> > > > > > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > > > > > _______________________________________________
> > > > > > CF-metadata mailing list
> > > > > > [email protected]
> > > > > > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > > > > 
> > > > > ----- End forwarded message ----- 
> > > > > _______________________________________________
> > > > > CF-metadata mailing list
> > > > > [email protected]
> > > > > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > > > > _______________________________________________
> > > > > CF-metadata mailing list
> > > > > [email protected]
> > > > > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > > > 
> > > > ----- End forwarded message -----
> > > > _______________________________________________
> > > > CF-metadata mailing list
> > > > [email protected]
> > > > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > > > _______________________________________________
> > > > CF-metadata mailing list
> > > > [email protected]
> > > > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > > 
> > > ----- End forwarded message -----
> > > _______________________________________________
> > > CF-metadata mailing list
> > > [email protected]
> > > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > > _______________________________________________
> > > CF-metadata mailing list
> > > [email protected]
> > > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > > _______________________________________________
> > > CF-metadata mailing list
> > > [email protected]
> > > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > 
> > ----- End forwarded message -----
> > _______________________________________________
> > CF-metadata mailing list
> > [email protected]
> > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> > _______________________________________________
> > CF-metadata mailing list
> > [email protected]
> > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> 
> ----- End forwarded message -----
> _______________________________________________
> CF-metadata mailing list
> [email protected]
> http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
> _______________________________________________
> CF-metadata mailing list
> [email protected]
> http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata

----- End forwarded message -----
_______________________________________________
CF-metadata mailing list
[email protected]
http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata

Reply via email to