5 July 2020

Comments on revised version of "Predicting tidal heights for extreme environments:
From 25 hr observations to accurate predictions at Jang Bogo Antarctic Research Station,
Ross Sea, Antarctica" by Byun and Hart

I did not look again in detail at the responses to the reviewers as I had seen
them already. I just read the new version afresh and below are some remarks.
This draft is certainly much improved on the first one but there are some remaining
things that need attending to, some trivial, including obvious problems with a couple of
figures. I believe when these issues are fixed the next version should be fine.

I must admit that there are aspects of the method used here I don't understand (other
than where I can relate to them as a response method). I guess one has to use it to
understand it properly, although I am sure it will interest other people.

8 - in the regional

12 - using a record from

13 - regime

53-67 - I must say, as I think I said before, that some of this attempted justification
is a bit over the top. But does no harm I guess.

68 - the applicability

80 and 96 - could you please make it clear what the JBARS and CR instruments are
delivering i.e. real sea level or sub-surface pressure? In the JBARS case it must
be SSP as it is a bottom-mounted instrument, unless there is some processing to
remove air pressure that is not mentioned. The CR instrument is a bubbler gauge
I believe, although your text does not say so, so that would be delivering sea
level I guess - you can check with Glen Rowe.

This does not matter much for diurnal and semidiurnal tides but it certainly does
for the longer period ones. See below.

102 - I think you mean you infer P1 from K1 and K2 from S2? It would be best to
express it that way.

105 - it would be worth adding 'at the two stations' again at the end of the
sentence. When I read it first time it looked like you were saying all the
amplitudes were the same at each station.

125 - from anytime --> throughout

143 - between the complete JBARS

157 - 'spring tide' --> 'diurnal tide'. 'Springs' and 'neaps'
are usually reserved for semidiurnals, although some people like to refer to
diurnal springs (not recommended).

163-166 - please can you make this much clearer? What I believe you are doing is
first relating 2017 JBARS to the 2017 CR predictions through the method and
seeing when it works best (Fig 5). Then you use the constants from the
best section of 2017 to produce predictions for 2019. Right? (Fig 6 in principle).
There is no CR data used directly in the latter. Again right? This could
be worded better.

Anyway there is a problem with figures. Fig 6 has 2019 in the header but the
figure itself is the same as Fig 5a. And then the end of this para refers to
Fig 7 but that is for lunar declination and would be better including in the
next section.

Section 4.2 - again the figures do not seem to be assigned to the text as
well as they might.

Also it occurred to me at this point - why is 25 hr so important? It is
obviously the minimum for looking at one tidal cycle, although you don't
actually say that anywhere. But how well would say short records of 50 or 100
hours do?

198 - it would be worth saying somewhere that this negative AT is very unusual.
You do say somewhere that AT is 1 or 2 days in most parts of the world so this
aspect of this area is worth noting.

215 - better to say: Srun excluded all long-period tides (i.e. the 6 listed in Table 2)

216 drop 'the'
217 ditto

and I think you mean 'And the CSTM+TCC repeated in each case'.

But please also see above my question at line 80 about SSP and sea level. If you
have sea level at CR and SSP at JBARS then you cannot assume the long
period tides (especially Sa and Ssa) are the same.

221 - why does an ice sheet respond only to MSf? Surely it responds to them all?

228-238 - surely the main thing is the amplitudes of K1 and O1 are almost the same,
so they will double or cancel over a fortnight, and when they are cancelling then
the small semidiurnals will manifest themselves. Obvious really.

243 - please can you put underflow and overflow arrows on the colour scale of Fig 11? As it stands there
is no area allowed below 0.25 or above 3 in the plot. Also there is a red (?) blob
half way down the east coast of the peninsula that draws the eye and you don't
comment on.

251-256 - these lines are just repeating yourself.

402 - records --> record (three times)

404 - you mean P1 from K1 and K2 from S2. Your use of brackets is confusing.

408 - from a harmonic analysis of one year-long

Please see my question at line 80. Sa and Ssa will be very different if SSP or
sea level are being used, although I believe if CR is a bubbler gauge then it
should be sea level ok - check with Glen.

Fig 2b - the lons and lats are upside down

Fig 4f - degrees is missing from y-title

430 - of the entire 369 ... record ... entire 17 ... record

Fig 5 - drop

440 - record

Fig 6 - is the same as Fig 5a

452/3 - drop the outer set of (). Too many of them.

Fig 9 - this is trivial stuff, isn't it? (a) and (b) are anyway near identical and
when reduced for print you won't see any difference. I think I would drop
this figure but see what you think

Fig 11 - see comments above

Remaining comments on this manuscript, mainly to do with describing the instruments again:

77 - thanks for clarifying the instruments. I think I would add 'absolute' and 'equivalent' to this to make it clear to the reader that this instrument is not measuring sea level as such i.e.

"using a bottom-mounted absolute pressure sensor (WTG-256S AAT, Korea) with the data converted
to equivalent sea level heights using the hydrostatic equation".

That is, if I have understood you correctly.

218 - I would write this:

Note that since the ROBT observation record was derived from a differential (vented) pressure sensor, unlike the absolute sensor at JBARS, and thus it .... affected by atmospheric (air pressure) forcing (i.e. Sa and Ssa).

229-232 - you say now:

While these results elucidate an issue with predicting Ross Sea tides based on the diurnal and
semidiurnal species alone, the aforementioned differences in gauge and record types in themselves
can also result in different harmonic analysis results and, in turn, different prediction results.

I guess you are now trying to accommodate the gauge type issue at line 77. Perhaps reword to be
clearer if you agree but of course use your own words:

These results indicate one particular issue to do with long-period tides when predicting Ross Sea tides
based on the diurnal and semidiurnal species alone. We note that the aforementioned differences
in gauge records (subsurface pressure or real sea level) introduce another, in that, while diurnal and semidiurnal tides can be considered to be measured equivalently accurately, the longer-period components might be expected to instrument-dependent and so have uncertainties for the above experiments.

419 - a one year-long observation

Figure 6 - I think I would drop although up to you.