Opened 7 years ago

Closed 21 months ago

#61 closed enhancement (fixed)

2 new cell methods

Reported by: awalsh Owned by: cf-conventions@…
Priority: medium Milestone:
Component: cf-conventions Version: 1.4
Keywords: cell methods, common concept name Cc: j.m.gregory@…, alison.pamment@…


1. Title

2 new cell methods

2. Moderator

Moderator - Alison Pammet Proposer - Andrew Walsh. Contributors and discussion - Jonathon Gregory, Roy Lowry, Mark Kulmar

3. Requirement

Extend the cell methods table with 2 new cell methods.

4. Initial Statement of Technical Proposal

As a result of discussion around the adding 9 new sea surface wave data

parameters to the CF standard names list it was decided to reduce this proposal to 6 new standard names thereby reducing the proliferation of new names. This reduction was achieved by introducing a 'common concept' name called 'sea_surface_wave_height' qualified by one of 4 cell methods i.e.

mean maximum root_mean_square mean_of_upper_decile

which will describe 4 of the statistical wave height variables:

sea_surface_mean_wave_height sea_surface_maximum_wave_height sea_surface_root_mean_square_wave_height sea_surface_wave_mean_of_highest_one_tenth_waves


'mean' and 'maximum' are existing cell methods. 'root_mean_square' and 'mean_of_upper_decile' are the 2 new cell methods proposed.

The other 5 new standard names proposed were:

sea_surface_wave_mean_crest_period sea_surface_wave_significant_wave_period sea_surface_wave_period_at_second_largest_peak_of_variance_spectral_density sea_surface_wave_variance_spectral_density_zeroth_frequency_moment sea_surface_wave_root_mean_square_amplitude_from_variance_spectral_density

5. Benefits

This proposal would benefit the netCDF user community as it reduces the proliferation of new names by using a common concept standard name qualified by cell methods. These 2 new cell methods may also find useful application with

other data parameters wherever statistical methods of 'root_mean_square' or 'mean_of_upper_decile' were applied.

6. Status Quo

Status quo would be to use non-standard long_names for these sea surface wave data variables with the disadvantage of poor interoperability.

Change History (6)

comment:1 Changed 7 years ago by jonathan

I support this proposal. I note that sea_surface_wave_height is proposed as a standard name (common concepts aren't agreed yet). The required change to the convention is to add the two new cell_methods to Appendix E. Neither of them alters the units i.e. then "units" columnn will read u for these two new methods. root_mean_square is self-explanatory, I would say; mean_of_upper_decile could be defined as the average of the largest 10% of values.

comment:2 Changed 7 years ago by lowry

Has my support as well.

comment:3 in reply to: ↑ description Changed 6 years ago by cf3glang


This can be regarded as a comment to the suggestion with some additional questions - see below. We have typically to deal with data of the following type.

Tidal High Water - Definition

Maximum height reached by a rising TIDE. The height may be solely due to the periodic tidal forces or it may have superimposed upon it the effects of prevailing meteorological conditions. Nontechnically, also called the high tide.

Typical Data

  • HW(N): array of consecutive high waters within a period of time.
  • xHW : maximum HW within period == max(HW(:)).
  • nHW : minimum HW within period == min(HW(:)).
  • mHW : mean HW within period == sum(HW(:))/N.

Standard Names

  • Option 1: Starting with existing standard name sea_surface_height:
    • for HW(N) use sea_surface_maximum_height,
    • for xHW use maximum_sea_surface_maximum_height,
    • for nHW use minimum_sea_surface_maximum_height, and
    • for mHW use mean_sea_surface_maximum_height.
  • Option 2: New standard name, e. g. high_water_due_to_tides:
    • for HW(N) use high_water_due_to_tides,
    • for xHW use maximum_high_water_due_to_tides,
    • for nHW use minimum_high_water_due_to_tides, and
    • for mHW use mean_high_water_due_to_tides.


  1. Are both options compliant with the suggested concept?
  2. Which option would be the recommended one?
  3. Other alternatives, suggestions available?
  4. Should we define something similar to the climatology concept, e. g. time: maximum within tides time: mean over tides?

A short reply would be very welcome.



comment:4 Changed 5 years ago by jonathan

This is an old proposal, which unfortunately did not reach a timely conclusion. However, I think we should now accept the proposal as originally stated, because enough support was expressed then, and no-one objected to it. Andrew Walsh should be named as an additional author of the CF convention when this change is made. No change to the conformance document is needed because this is just an addition to vocabulary.

The third posting was more of a question and digression. Guenther, I suggest that if you wish to pursue this still, you could ask on the email list or open a new ticket.


comment:5 Changed 4 years ago by rhorne@…

See trac item #102 for markup required to the CF standard supporting the two new cell methods.

comment:6 Changed 21 months ago by markh

  • Cc changed from, to,
  • Resolution set to fixed
  • Status changed from new to closed
Note: See TracTickets for help on using tickets.