Dear Jonathon, Thanks a lot, that is really helpful. Whilst I appreciate your point about whether or not '_upwave/_downwave' are necessary if MSS is unsigned, I think it wold still be useful to have this since the _mean_square_slope_*_direction may well get compared with _wave_*_direction and/or _wind_*_direction. Since and these latter follow a convention it is useful/necessary not to have any ambiguity in how these are referenced. [Thinking even more long term, upwave/downwave slopes could be assigned different values based on wave asymmetry, but lets not go there yet...]
So, I think we have enough now to summarise the proposed new names and see how we are doing... Agreed so far (I think) from earlier mails: charnock_coefficient_for_surface_roughness_length_for_momentum_in_air Units: 1 Coefficient value, based on the Charnock (1955) empirical expression for deriving surface_roughness_length_for_momentum_in_air over the ocean. The surface called "surface" means the lower boundary of the atmosphere. sea_surface_wave_mean_square_slope Units: 1 Wave slope describes an aspect of sea surface wave geometry related to sea surface roughness. Mean square slope describes a derivation over multiple waves within a sea-state, for example calculated from moments of the wave directional spectrum. sea_surface_wave_x_mean_square_slope Units: 1 Wave slope describes an aspect of sea surface wave geometry related to sea surface roughness. Mean square slope describes a derivation over multiple waves within a sea-state, for example calculated from moments of the wave directional spectrum. "x" indicates that slope values are derived from vector components along the grid x-axis. sea_surface_wave_y_mean_square_slope Units: 1 Wave slope describes an aspect of sea surface wave geometry related to sea surface roughness. Mean square slope describes a derivation over multiple waves within a sea-state, for example calculated from moments of the wave directional spectrum. "y" indicates that slope values are derived from vector components along the grid y-axis. Testing the new 'upwave' names: sea_surface_upwave_mean_square_slope_direction Units: degree Wave slope describes an aspect of sea surface wave geometry related to sea surface roughness. Mean square slope describes a derivation over multiple waves within a sea-state, for example calculated from moments of the wave directional spectrum. " direction" is used to assign a directional axis along which wave energy is travelling, with "upwave" used to indicate that this is equivalent to a "from_direction". sea_surface_upwave_mean_square_slope Units: 1 Wave slope describes an aspect of sea surface wave geometry related to sea surface roughness. Mean square slope describes a derivation over multiple waves within a sea-state, for example calculated from moments of the wave directional spectrum. "upwave" indicates that slope values are derived from vector components along (parallel to) the axis from which waves are travelling. sea_surface_crosswave_mean_square_slope Units: 1 Wave slope describes an aspect of sea surface wave geometry related to sea surface roughness. Mean square slope describes a derivation over multiple waves within a sea-state, for example calculated from moments of the wave directional spectrum. "crosswave" indicates that slope values are derived from vector components across (normal to) the axis from which waves are travelling. Hope these are beginning to sound sensible :-) Cheers Andy -----Original Message----- From: CF-metadata <[email protected]> On Behalf Of Jonathan Gregory Sent: 01 October 2018 18:29 To: [email protected] Subject: Re: [CF-metadata] standard names for sea surface roughness variables Dear Andy Thanks. I think your suggestion of "upwind" is certainly clearer than "from" (and "downwind" would be much better than "to"). Your middle options would be fine. > Parallel component: > sea_surface_wave_mean_square_slope_along_upwave_direction > Normal component: > sea_surface_wave_mean_square_slope_across_upwave_direction and your first options would be OK too, except I wonder if they'd be better as > Parallel component: sea_surface_mean_square_upwave_slope > Normal component: sea_surface_mean_square_crosswave_slope since it's the slope which is along or across the direction, and I made crosswave into one word like upwave, upward, eastward, etc. I think I'd prefer these shorter ones myself. But I still have a question about whether upwave and downwave need to be distinguished anyway for a mean square slope. Isn't avg((dh/dx)^2) the same regardless of the sign convention of x, if x is the wave direction? If it's not, don't you have to say whether cross-wave is leftward or rightward, correspondingly? Best wishes Jonathan ----- Forwarded message from "Saulter, Andrew" <[email protected]> ----- > Date: Mon, 1 Oct 2018 09:30:03 +0000 > From: "Saulter, Andrew" <[email protected]> > To: "[email protected]" <[email protected]> > Subject: Re: [CF-metadata] standard names for sea surface roughness > variables > > Good morning Jonathon, > > Was nice to have a weekend's reflection on this, not least because I also got > a bit more feedback from some of my other waves colleagues (thanks Fabrice). > > A quick fundamental, the reason we need to have some form of 'along' and > 'across' follows the same argument as the 'spread' conversation. Basically, > wave energy in a given sea-state is not uni-directional, so we have a > dominant/mean direction that gets calculated, but there will be a component > of wave energy (with associated height, period, slope characteristics etc.) > that runs normal to this. > > In terms of what the "direction" really is, the suggestion I've been given is > "upwave", i.e. a wave equivalent of "upwind" and, therefore, same as > "wave_from_direction" (correcting my initial suggestion of "to" in the > previous post). > > This gives us a few choices for names I think? > > Least verbose: > Direction: sea_surface_upwave_mean_square_slope_direction / > sea_surface_wave_mean_square_slope_from_direction* > Parallel component: sea_surface_upwave_mean_square_slope > Normal component: sea_surface_cross_wave_mean_square_slope > > More verbose (but perhaps more clear?): > Direction: sea_surface_upwave_mean_square_slope_direction / > sea_surface_wave_mean_square_slope_from_direction* > Parallel component: > sea_surface_wave_mean_square_slope_along_upwave_direction > Normal component: > sea_surface_wave_mean_square_slope_across_upwave_direction > > More consistent with existing names (but possibly least clear?): > Direction: sea_surface_wave_mean_square_slope_from_direction > Parallel component: > sea_surface_wave_mean_square_slope_along_from_direction > Normal component: > sea_surface_wave_mean_square_slope_across_from_direction > > * if we use _from_direction in conjunction with _upwave, then we need to add > some text to link the two terms in the standard name definition. > > Any of these make sense? > Cheers > Andy > > PS. Devon is geographically 'up' from Cornwall - but definitely 'down' in > terms of the quality of pasties, clotted cream and beer.... > > > > -----Original Message----- > From: Jonathan Gregory <[email protected]> > Sent: 28 September 2018 13:46 > To: Saulter, Andrew <[email protected]> > Cc: [email protected] > Subject: Re: [CF-metadata] standard names for sea surface roughness > variables > > Dear Andy > > > Re the direction of the _mean_square_slope, the parameter and calculation > > method from the wave spectrum is sufficiently different from that for > > _wave_[to/from]_direction that it should stand alone. There has already > > been a precedent set for this with waves, where different forms of > > parameter calculation from the spectrum are given their own names because > > there is not only a calculation difference but a different physical > > interpretation of each parameter (e.g. the various type of wave period). > > OK, fair enough. So you need sea_surface_wave_mean_square_slope_to_direction. > > I'm still stuck with what this "direction" really is. Can we insert anything > else for ? in > sea_surface_wave_mean_square_slope_along_?_direction > sea_surface_wave_mean_square_slope_across_?_direction > Apparently you want to quantify the mean square slope along and across the > direction of the mean square slope. Is that right? I'm not sure what it means. > Without the "mean square", I'd think that the slope normal to the direction > of the slope must be zero, but it must be more subtle than that in this case! > > Is there really an ambiguity of to/from with a mean square slope? It seems to > me that it must be the same (unsigned) number regardless of whether you go > backwards or forwards on a particular direction. > > Is Devon up or down from Cornwall? > > Best wishes > > Jonathan > > _______________________________________________ > 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
