Dear Jonathan,
Based on some information from my colleagues and information on the
internet, I have gathered here some comments on wave periods. If you
need more information, I could consult the wave experts amongst my
colleagues.
As I see it, if someone refers to the "mean wave period" that period may
have been determine in a number of ways.
* One method is to take the average length of the periods between (up or
down)crossings of the reference (zero) level. It is also common to
derive other periods from time series such as average period of the 1/3
highest waves, T_{H_{1/3}}, often referred to as the significant wave
period (which is often close to the peak period as discussed in this
thread).
* Other methods derive a mean wave period from the energy spectrum.
There are 9 standard names containing the expression wave_mean_period
and all refer to these spectral methods. The spectrum itself may be
obtained from an observed (or computer generated) signal or be the
result of a spectral wave model (i.e. numerical model working in the
frequency domain). Three of the standard names refer to waves in
general, three of them refer to swell waves, and three of them refer to
wind waves. Their definitions clearly indicate the exact way in which
they should be computed from a swell, wind wave or general spectrum. If
we look only at the first three:
sea_surface_wave_mean_period_from_variance_spectral_density_first_frequency_moment
sea_surface_wave_mean_period_from_variance_spectral_density_inverse_frequency_moment
sea_surface_wave_mean_period_from_variance_spectral_density_second_frequency_moment
This list of spectral methods may be extended with a time series method,
e.g.
sea_surface_wave_mean_period_from_zero_crossings
Although for some spectra all these methods may give approximately the
same results, they may differ greatly for other types of spectra. So,
wave experts will be more careful about labeling the wave period that
they are referring to than the average user. They will refer to first
wave period T_{m01}, inverse wave period T_{m-10}, second wave period
T_{m02} and zero-crossing period T_z respectively; the standard names
definitions include these terms. For some common wave spectra there
exist certain relationships between the periods, e.g. for the common
Jonswap spectrum T_p = 1.19 T_m01 = 1.28 T_m02 and T_z = T_m02.
Sometimes these relationships are misused to derive other quantities in
a way that is not consistent with their formal definition, e.g. one may
use the aforementioned relations to estimate the peak period from the
zero-crossing period. The results can be quite inaccurate.
Often all four quantities mentioned above may all be casually referred
to as sea_surface_wave_mean_period, which may be considered a higher
level in a wave ontology.
A more extensive description can be found in section 1.3.3 and 1.3.8 of
the following document
http://www.wmo.int/pages/prog/amp/mmop/documents/WMO%20No%20702/WMO702.pdf
Best regards,
Bert
------------
Jonathan Gregory wrote:
> Dear Roy
>
>
>> I think the "mean" comes from the world of observations. Wave data are acquired with kit that takes multiple measurements so any parameterisation is subject to variance.
>>
>> However, I have only ever seen means used and I have seen a great deal of inconsistency in whether the fact that the numbers are averages is labelled explicitly or not.
>>
>
> Ah right, I see. In that case I think we could omit it from the stdname, as
> it's an aspect of measurement technique rather than a definition of the
> geophysical quantity.
>
> Thanks. Cheers
>
> Jonathan
> _______________________________________________
> CF-metadata mailing list
> CF-metadata at cgd.ucar.edu
> http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
>
>
Received on Tue Aug 11 2009 - 12:07:28 BST
