the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Process-based modelling of nonharmonic internal tides using adjoint, statistical, and stochastic approaches. Part I: statistical model and analysis of observational data
Abstract. A substantial fraction of internal tides cannot be explained by (deterministic) harmonic analysis. The remaining nonharmonic part is considered to be caused by random oceanic variability, which modulates wave amplitudes and phases. The statistical aspects of this stochastic process have not been analysed in detail, although statistical models for similar situations are available in other fields of physics and engineering. This paper aims to develop a statistical model of the nonharmonic, incoherent (or nonstationary) component of internal tides observed at a fixed location, and to check the model's applicability using observations. The model shows that the envelope-amplitude distribution approaches a universal form given by a generalization of the Rayleigh distribution, when waves with non-uniformly and non-identically distributed amplitudes and phases from many independent sources are superimposed. Mooring observations on the Australian North West Shelf show the applicability of the generalized Rayleigh distribution to nonharmonic vertical-mode-one to mode-four internal tides in the diurnal, semidiurnal, and quarterdiurnal frequency bands, provided that the power spectra show the corresponding tidal peaks clearly. These results demonstrate the importance of viewing nonharmonic internal tides as the superposition of many random waves. The proposed distribution can be used for many purposes in the future, such as investigating the statistical relationship between random internal-tide amplitude and the occurrence of nonlinear internal waves, and assessing the risk of infrequent strong waves for offshore operations. The proposed statistical model also provides the basis of investigating processes and parameters controlling nonharmonic internal-tide variance in Part II.
Competing interests: Dr. Matt Rayson (editor) and other oceanographers at University of Western Australia are involved in an on-going collaborative project with my company (involving myself) on the topic of this manuscript. Also, I have a competitive relationship with them for industry-funded projects on topics related to this manuscript (in Australia).
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.- Preprint
(5466 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2024-4192', Jonas Nycander, 21 Feb 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2024-4192/egusphere-2024-4192-RC1-supplement.pdf
-
RC2: 'Comment on egusphere-2024-4192', Michael Cox, 27 Feb 2025
General Comments
Internal tide observations consist of a harmonic part, well-modelled by a sum of sinusoids at tidal frequencies, and a nonharmonic part. The author models internal tide observations at a given frequency by the sum of many waves with random, time-dependent amplitudes and phases. The harmonic amplitudes and phases are given by the expectation value of the total complex amplitude of the wave, equivalent to a least-squares fit which is standard in literature. Subtracting this least-squares-fitted harmonic part from the total internal tide signal leaves the nonharmonic remainder.
By considering the internal tides at a single frequency to be the sum of many waves, the author allows for generation of tides from multiple sources, a novel contribution of this study. The waves propagate through the ocean, a random media, which justifies the random amplitudes and phases of the signal.
The author derives the total probability density functions (PDFs) of nonharmonic amplitude and phase from the individual amplitude and phase distributions. Statistical assumptions on the form of individual wave amplitude and phase PDFs allow for relatively straightforward computation of the total PDFs. These total PDFs compare favourably to observation (figure 6), seemingly justifying the author’s multi-source, statistical approach.
The work is well-written and provides an interesting contribution to the field of nonharmonic-tide modelling. I am happy to recommend publication after the author has addressed the points below.
Specific Comments
- I am in support of the suggested alternative derivation by Jonas Nycander for section 2.1., providing the author agrees that the two are equivalent. (Please ignore points 2--5 below if the alternative derivation is implemented.)
- Lines 147--8: For those of us who aren't statisticians, please can you state what the Fourier transform convention in statistics is?
- Abramowitz and Stegun (cited line 158) has a digital successor, the Digital Library of Mathematical Functions (DLMF). This is a citable resource and much easier for the online reader to use. Please consider changing your reference to the relevant DLMF equations.
- Eqn 9 can be obtained using just DLMF 10.9.2 and 10.4.1 rather than Eqs. 9.1.5, 35, 44, and 45 of Abramowitz and Stegun (line 158).
- Eqn 12: a reference to the Fourier-Bessel series used would be useful for the reader.
- Please define "celerity" or change to "phase speed" if the two are equivalent.
- As I understand it, the author's vertical mode formulation cited as Shimizu (2017, 2019) on lines 282--3 and Shimizu (2011) on line 306 is somewhat equivalent to that of Kelly et al. (2012) and subsequent works. If this is the case, the author should cite the work of Kelly and co-authors as well.
- The author should mention the SWOT mission in his introduction (e.g. Morrow et al. 2019), given it is a key reason for current interest in the nonharmonic tide.
- Kachelein et al. (2024) consider amplitude modulation for nonharmonic tides in the California Current, where tides propagate from many generation sites. Perhaps this relevant work should be cited?
Technical Comments
- Typo line 35 immediately before barotropic. Inverted comma (') should be quotation marks (").
- Typo line 91. "mooing site" should be "mooring site".
- Line 215: missing "the" after "We also need".
- Grammar lines 375--6. Could be resolved with the addition of "for" i.e. "searched for numerically."
- Clunky phrasing lines 480--1: "However, this difference does not matter to see that the phase is roughly uniformly distributed". Something like "However, the phase is roughly uniformly distributed despite this difference" would work better.
- Change "would be" on lines 521 and 522 to "is".
References
NIST Digital Library of Mathematical Functions. https://dlmf.nist.gov/, Release 1.2.3 of 2024-12-15. F. W. J. Olver, A. B. Olde Daalhuis, D. W. Lozier, B. I. Schneider, R. F. Boisvert, C. W. Clark, B. R. Miller, B. V. Saunders, H. S. Cohl, and M. A. McClain, eds.
Kelly, S. M., J. D. Nash, K. I. Martini, M. H. Alford, and E. Kunze, 2012: The Cascade of Tidal Energy from Low to High Modes on a Continental Slope. J. Phys. Oceanogr., 42, 1217–1232.
Morrow, R., Fu, L.-L., Ardhuin, F., Benkiran, M., Chapron, B., Cosme, E., et al. (2019). Global observations of fine-scale ocean surface topography with the surface Water and Ocean Topography (SWOT) mission. Frontiers in Marine Science, 6, 232.
Kachelein, L., Gille, S. T., Mazloff, M. R., & Cornuelle, B. D. (2024). Characterizing non-phase-locked tidal currents in the California current system using high-frequency Radar. Journal of Geophysical Research: Oceans, 129, e2023JC020340.
Citation: https://doi.org/10.5194/egusphere-2024-4192-RC2 -
RC3: 'Comment on egusphere-2024-4192', Anonymous Referee #3, 10 Mar 2025
Review of Shimizu, Part I.
This paper derives the probability distribution of a sum of waves,
where the amplitudes and phases of the component waves in the sum are assumed
to be random variables.
The development sets down a clear derivation in one place. It also does a
nice job of contextualizing and motivating the derivation as a continuation of
previous work.
The author provides a clear justification for the term "nonharmonic", in contrast
to other terms, "non-phase-locked", "incoherent", and "non-stationary", to describe
the type of tidal signals which should be described by the proposed probability distribution.
Finally, the author provides a good qualitative discussion of the distribution in the
limit of multiple wave sources with different properties, which emphasizes the rate
at which the asymptotic distribution is approached, and provides insight into how this
tends to obscure information about the underlying properties of the component waves.
For all these reasons, I think the paper makes a nice contribution to the field,
and it ought to be published.
Although I am familiar with the basic methods employed in the article, it would take
me considerable time to verify all the steps of the derivation in detail, and I have not
attempted to do so. In order to increase the pedagogical value of the paper, and also
to instill more confidence in the results, I would encourage the author to include more
detail, possibly in the form of additional short appendices.
Minor comments:
l70: these are all good justifications
l83: Good---it looks like he has familiarized himself with the
literature of other fields.
l113: It is not clear to me what the expected value is
averaging over. Is it assumed that the random realizations represent different time series?
Or will this be used with segments of time series of short duration, or what, exactly?
(Oh, I see later --- the segments of the time series are treated as independent realizations.)
l142-143: Can he put these steps into an appendix?
l149: \phi is the characteristic function of A',\Theta' ?
l166-l171: Sorry, I'm not sure how to check these.
l181-l182: So are these the transformation of the joint Gaussian in Cartesian coordinates to the polar coordinates?
l190: Why is the phase distribution bimodal?
l224-232: good qualitative discussion
l276: "tidal" --> "non-tidal"?
l306: I'm not sure about the units here, given the prior normalizations.
l319: "displaement"
l385: Is this the same as Colosi and Munk?
l390: Not sure I understand. Couldn't it be estimated using the same approach as for fitting the Lorentzians?
l415-l423: good qualitative discussionCitation: https://doi.org/10.5194/egusphere-2024-4192-RC3
Viewed
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
86 | 23 | 6 | 115 | 4 | 5 |
- HTML: 86
- PDF: 23
- XML: 6
- Total: 115
- BibTeX: 4
- EndNote: 5
Viewed (geographical distribution)
Country | # | Views | % |
---|---|---|---|
United States of America | 1 | 57 | 44 |
China | 2 | 11 | 8 |
United Kingdom | 3 | 8 | 6 |
Australia | 4 | 7 | 5 |
France | 5 | 7 | 5 |
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1
- 57