the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Wave forecast investigations on downscaling, source terms, and tides for Aotearoa New Zealand
Abstract. This study evaluates the effects of downscaling, source terms, and tidal interactions on numerical wave forecasts in Aotearoa New Zealand. We utilised a set of three nested domains (from global to regional scale) to examine significant wave height (Hs), mean period (Tm01), and peak wave direction at two coastal locations, Banks Peninsula and Baring Head. Downscaling markedly improved forecast accuracy at Baring Head, a tidally constricted region, reducing Hs forecast error by 25 %. However, improvements at Banks Peninsula were minimal, likely due to its open coast characteristics which are adequately represented even by lower resolution models. Source term enhancements using default ST6 parameters generally improved Hs predictions on the west coast but worsened them on the east, indicating a geographical dependency in model performance. This variability was also evident in the Tm01 predictions, with notable improvements in bias reduction through model downscaling, particularly at Baring Head. Tidal influences were significant, especially at Baring Head, where they enhanced the forecast accuracy of wave height and direction due to the strong tidal currents characteristic of this location. In contrast, at Banks Peninsula, tidal effects were less pronounced. The study underscores the importance of tailored modelling approaches that consider local geographical and hydrodynamic conditions to optimise wave forecasting.
- Preprint
(5800 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on gmd-2024-110', Anonymous Referee #1, 07 Feb 2025
In the paper ‘Wave forecast investigations on downscaling, source terms, and tides for Aotearoa New Zealand’ the authors show the results of a set of nested wave models. The results are compared with data from two buoys characterized by different tidal conditions; moreover, the authors discuss the difference in the significant wave height pattern against satellite data. The manuscript is well organized and can help in the setup of new wave simulations.
The performances of the models presented are good. The authors show that the increases in the quality of the results with the model resolution depend on the site conditions. However, some conclusions on the improvement obtained by considering tide in the wave model are based on differences in statistical parameters too small. I think that the authors should take this problem into consideration and underline it in the abstract and in the conclusions. For example, buoy data at Baring Head present peak direction changes of about 80° in a cycle while results from the model experiment including tide show very little modulations both in the wave height and direction. Among the hypotheses presented by the authors to explain the limited impact of tides, there is an underestimation of the current computed by RICOM with respect to that from TELEMAC. It should be interesting to perform a short simulation (just for the period shown in Figure 13) to check the impact of using this alternative forcing.
Specific comments:
Line 7-9 The statement: ‘Source term enhancements using default ST6 parameters generally improved Hs predictions on the west coast but worsened them on the east, indicating a geographical dependency in model performance.” should be postponed, as it seems that the following statement is connected to the previous one.
Line 10-11 ‘Tidal influences were significant, especially at Baring Head, where they enhanced the forecast accuracy of wave height and direction’ as I have previously observed I believe this statement cannot be proven by the statistical parameters.
Line 18-20 ‘Wave climate studies are usually derived from longterm hindcasts or reanalyses. Wave variability can also be described in context of extreme (Simmonds and Keay, 2000) and ambient wave climates (Mortlock and Goodwin, 2015).’ I think these statements should be rewritten or eliminated as they are too generic and not connected with the subject of the paper.
Line 45 – In general, I think that it could be better to sign every place cited on the map (i.e. Cook Strait or Hauraki Gulf, Banks Peninsula and Baring Head on the map in Figure 1, Steep Head, and Te Whanganui-a-Tara/Wellington on Figure 2).
Line 76-77 ‘The EcoConnect platform, comprising data ingestion, numerical modelling applications for a variety of natural hazards and forecast data delivery.’ I think this phrase is not complete.
Line 82-83 Could you please specify more clearly if the model domains of the atmospheric and wave models coincide?
I think it could be useful, due to the complexity of the model configuration, add a table with the main details of the wave models, containing for example: atmospheric model forcing, area covered and simulation length, assimilation...
Line 260 – It is not clear why you use the wording ‘historical wave forecasts’.
In Figure 10 it seems that the total percentage computed over all the directions doesn’t reach 100 in the directional histogram for observations. Could you check?
Line 379 – The timeseries of significant wave height and mean period of the NZWAVE-HR simulation show the 12-hour periodicity, however they are very close to those of the NZWAVE-HR-NOTIDES simulation. The authors say: ‘which highlights the importance of including tides as forcing’. I think this is a too strong statement to do just looking at the Figure, at least the correlation with data could be computed.
Line 464-470 I think that the authors should change the statement : ‘This simulation reduced …’ as the difference in the statistical parameters is extremely low. In the same way they should avoid saying: ‘but tides didn’t show large impact on the wave forecast evaluation statistics…’ when discussing results for Bank Peninsula as they have not shown ‘large impact’ even for the other buoy.
Typos and minor:
Admas-Bashforth scheme => Adams-Bashforth
Line 276 – eliminate ‘in these regions’
Line 379 – the number of the Figure is missing
Line 382 - red rectangles in Figure 11 are not ‘Around midnight on the 21st and 22nd of
Oct 2021’ but around midday and midnight on the 21st of Oct 2021.
In the caption of Figure 13 is written ‘peak wave period’ but it should be ‘peak direction’.
Lin 416 – 419 please check the description as it seems that period should be changed in peak.
Line 463 – ‘generates amplitudes I think should be changed in ‘generates changes in amplitudes’
Line 483 – 486 check the order of the statistics named (mean bias, RMSE and correlation’ should ‘mean bias correlation and RMSE’)
Citation: https://doi.org/10.5194/gmd-2024-110-RC1 -
RC2: 'Comment on gmd-2024-110', Anonymous Referee #2, 17 Feb 2025
This paper presents the results of a wave forecast system for New Zealand. The sensitivity to the variation of resolution, forcing and tide is considered. The paper is written in good English, and the results are well presented. I have some minor comments, reported below:
- People outside New Zealand are not aware of Maori's names. Moreover, this is used only in some locations and in different ways (with a space or a bar /). This is a bit misleading, I suggest putting one of the two names in parenthesis and specifying it with a short sentence in the Introduction.
- The presentation of the results is a bit too long, with too many figures, scatter plots in particular. Reduce them, maybe you can add a table with some statistics.
- Since the wind forces the waves, I would like to see some statistics on its quality.
- The tidal interaction is the more physically interesting thing, I suggest expanding this part. Maybe you can add some spectral analysis.
- p1r7: Define ST6.
- p2r43: remove "refer to".
- Section 2.2.1: Add some citations on the models.
- p8r193-198: I don't understand the use of this model, specify better.
- Tm01, Tm02: explain better the meaning of both these quantities the first time you use one of them.
- p21r379: Fig. reference is missing.
- Fig. 11-13: Use different colours and symbols, they are difficult to read.
Citation: https://doi.org/10.5194/gmd-2024-110-RC2
Viewed
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
227 | 46 | 133 | 406 | 14 | 12 |
- HTML: 227
- PDF: 46
- XML: 133
- Total: 406
- BibTeX: 14
- EndNote: 12
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1