the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 16 Sep 2026
Tides of Norway: insights from a dense coastal tide gauge network
Abstract. Tide gauges provide highly precise observations of coastal sea-level variability, yet their spatial sparsity remains a fundamental limitation in regions characterised by complex coastal geometry. This study utilises an extensive archive of 250 short-duration tide-gauge deployments by the Norwegian Mapping Authority (Kartverket) between 2009 and 2026 to characterise tides along the Norwegian coast. We characterise the major tidal constituents and several tidal datums along the entire coastline, including the maximum tidal range and high-tide durations. Further, these data provide a unique source for global model validation, as shown by comparisons against five state-of-the-art tide models, yielding median errors of 2.4 to 3.3 cm that scale with model spatial resolution. Collectively, these results provide a high-density in situ characterisation of Norwegian tides, reveal fine-scale tidal variability that is unresolved by sparse permanent networks, and establish an observational basis for future spaceborne investigations of tidal dynamics in Norway's fjords and coastal waters.
- Preprint
(8777 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 11 Nov 2026)
- RC1: 'Comment on egusphere-2026-5493', Philip Woodworth, 22 Sep 2026 reply
Data sets
Tidal constituents of Norway (ToN) from temporary gauges M. Hart-Davis et al. https://doi.org/10.17882/119454
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 287 | 108 | 32 | 427 | 31 | 30 |
- HTML: 287
- PDF: 108
- XML: 32
- Total: 427
- BibTeX: 31
- EndNote: 30
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
22 September 2026
Comments on 'Tides of Norway: insights from a dense coastal tide gauge network' by Hart-Davis et al. (OS)
This is a straightforward presentation of harmonic constants derived from a dense tide gauge network in Norway, pointing out that the tides are more complicated along this indented coastline with many islands than one might gather from a global tide model. I don't have major problems with the paper, but I suggest that the wording needs a number of improvements and possibly an extra set of figures (see comment at line 200).
Section 2 - the collection of tide gauge records, even if short, from 250 sites is quite an achievement, and yet there is no mention of why the measurements were made (for a tidal study such as this?), who had the idea (other than simply 'Kartverket'), whether they have been used elsewhere and so can be referenced etc. Also in the 'Data availability' section you say the data are available from Kartverket but it would be good to give a website or email.
line 7-8 - 'scale with model spatial resolution' is confusing as literally it suggests higher accuracy with coarser grid scale. Suggest something like: 'that reduce with higher model spatial resolution'.
18 - these two papers used only a couple of sites in southern Norway, and I wondered if the sentence gives the impression that they used more, and hence was wrong. I think here I would just reference a text book like Pugh and Woodworth or Gerkema maybe.
Fig 1b - the numbers on the phase lines could be larger, but also it would be better to make them all positive as a reader may miss the negative sign. Also, while all the semidiurnals will have a similar pattern, with amplitudes scaled, what about the diurnals? They are smaller but their dynamics is interestingly different.
Line 1 of caption - Map of the M2 tide. Line 2 - Greenwich phase lag (presumably).
Line 3 - Three key regions
35 - Conclusions are ..
Figs 2a and 2c are mentioned explicitly in the text but 2b isn't
Line 1 of caption .. of the coastal tide gauge ..
38 Suggest: The 250 ... coast (Figures 2a,b) provide ... than global [drop 'traditional'] sea level archives (Figures 2c, A1, A2). Each sensor ... and its record is corrected ..
39 .. 3.5 years (Fig.2a) with ..
Make it clear in the text that the number of observations mentioned in 2b must be the 10 minute and not 1-minute values
46 - resolvable given the span of these tide gauge records.
47 - I would merge lines 53-56 into the sentence here to make the writing neater i.e.
In this manuscript, we focus on the eight major tidal constituents (M2, S2, N2, K2, K1, O1, P1 and Q1), as they .. in the full tidal signal. We derive ... and determine confidence intervals for these constituents, which ... (Para) The pairs ..
49 .. to another within these pairs in records shorter than 182.62 days (this ..
59 and lower down - why 20 years and not 19 (approx 18.6)?
64 - these tide gauge records.
67 - Greenwich phase lags
86 - there is indeed a semidiurnal amphidrome in southern Norway but it is very close to the coast and worth mentioning that it is probably a degenerate (virtual) one - that conclusion may vary between models.
69 - why is that final? I would have thought the masking would happen before the interpolation.
71-72 - I think you have to spell out what the RMS and RSS are so it is clear. It is not reasonable to expect the reader to also look at Stammer et al.
81 - .. the amplitude fields for each constituent exhibit .. coherence.
87 - Further from this region, semi-diurnal amplitude ..
89-90 .. northen coastlines where near-neighbour gauges .. differences.
Other semi-diurnals will have similar patterns but what about the diurnals?
Fig 3 caption - ordered by mean amplitude. Ok but this is anyway the usual order, see comment for figs A1,A2
115-120 - this is ok but the text does repeat itself (general coherence and local complexities) a few times
123 - reference for GESLA-4, or say it is a development of GESLA-3 referencing earlier Haigh et al.
Figure 4 - personally I prefer to express phase lag 0-360 as one's eyes cannot miss the minus sign as with -180 to 180. That's personal preference. But the main problem here is having blue for -180 and red for +180 when the two are the same A colour scale should be used that wraps around.
Fig.4 caption - Greenwich phase lags
128 - the shorter record lengths ... constituents on which to base the predictions.
129 - .. LAT), relative to Mean Sea Level, are key ..
132 - tidal range is not a datum so is strictly not one of the three - a datum is a level that one can point to, but range is a difference between datums. I must admit that MTR, for example, is often listed on the same page as MHW and MLW but even so ... Some minor rewording needed.
137 - define NOAA
138 - for all the tide gauges
What does 'mean' mean here? Is this the mean duration for those times when MHHW is execeeded, or does it include the tides when it is not exceeded?
150 - you could mention that the idea for dodginess came from the mixed tides at Adelaide (I forget the references off-hand)
154 - 'utilise all constituents' - but there are only 4 shown in eq (1).
159 - can better contribute
174 - .. lowest along the coastline, averaging 2.4 cm for both models (Fig 6a,b).
178 - errors
183 - I don't understand what percentage difference means. Minor rewording needed.
187-188 - it is not a grid if it is just a circle with a 10km radius
189 - percentages of land below ... being gauges with land between ... gauges with land above
193 - again, you need to explain better what standard deviation ensemble RMS means. The latter is shown in Fig 6 as a percentage, needs spelling out
195-200 - this is a very nice statistical treatment but I wonder also if it would be useful to calculate the AVD (absolute values of the vector-difference between gauge and model constituents) for your best model (or all models) and plot AVD thereby showing the poorer performance (larger AVD) of models in the complicated areas.
205 - two parameters
207 - should this be tidal amplitude error? Ensemble model error again. Please look at the wording of this para
242 - this waa a temporary web address for GESLA. Use the official one now.
Figures A1, A2 - can the constituents be put in the same order as before.
A2 caption - Greenwich