6 November 2019

Dear Igor,

thanks for the responses to the reviewers and for the revised version (V3).
In particular, thanks for adding a section on percentage of variability due to
the tides or meteorology as suggested by R1. And for the modified figures as
suggested by R2.

As you know, one problem I had from that start with this paper was to do
with many small things that needed fixing. You recall that I sent your first submitted
version back to you for that reason, and you then produced the second version
which went to the reviewers.

Nevertheless, although that V2 was much better, and R1 thought that version was presented well,
he was still able to identify many other detailed things. And I sent you privately
an annotated pdf with edits. But there are still things in this new V3 which I list below - it looks
a lot but they will not take you long. Perhaps you could ask your two co-authors to take the
time to read it through the text again properly.

I also asked the two reviewers to read the new V3:

R1 - made the comments which I include below. He provides some other technical edits, which
reinforces my comment that the text is just not as good as it should be. Also
he has a specific question about your comment on page 22 to do with TPX08. I was concerned
about this also so I asked about it and you can see the email from Svetlana Erofeeva below.
I think you have to remove that TPX08 sentence as it stands and replace it with something
more general to do with the sensitivity of the tides to MSL. See my comment below on line 22 also.

My only disagreement with R1 is his first one. I prefer 'subject'! But I am not going to
lose sleep about it.

R2 - pointed out the Figures 7, 8 and 10 do not use the same depth change for the Caspian.
The first two use 25, 27 and 29, the other 26 28 and 29. I am not insisting on it and it
is up to you but I guess he is right that it would be neater if Figure 10 also used the
former set if it is easy to recompute Figure 10 with them.

I hope you can attend to these things ok. I have no intention of asking the reviewers a
third time so any new version (V4) will be looked at only by me and I hope that will be the final one.

All the best again.

Phil

Detailed comments are:

p1, 8-9 you have generate twice in the same sentence. I suggest

.. are produced directly by the tide-generating forces.

11 .. harmonics, together with maps of form factor, tidal range and
tidal current speed.

14 .. Peninsula separates two individual amphidromes with ..

17 .. Kizlyar Bay. As a result, the semidiurnal ..

p2, 3 though --> although

5 tidal range was found to be 7.7 cm

9 maximum tidal range to be 12 cm at Ogurja Ada (Figure 1).

This is the first time you give a place name so I think you should mention the figure 1
here.

10 - did Spidchenko not come to some conclusion from his 7 sites? ok, if not leave it as
it is if he didn't have anything to say, but otherwise say what he concluded.

12-13 center in the area of the Absheron Threshold.

29 In the present paper, we

p3, 7 (twice) I would order it h and k

Table 1 line 1. I would drop 'and the elasticity factors'. You show just the Love
numbers as mentioned in the text.

p4, 6 General Bathymetric Chart of the Oceans (GEBCO)

8 The dataset is based

what are 'and points'? do you mean coastline positions?

Figure 1 - what are the numbers on the contours? Are they depths below the -28 m MSL or
below BHS datum. It needs a sentence at the end of the caption.

p6, 3 - drop In Medvedev et al. (2019) second time.

5 was obtained for zero

Figure 2 caption line 2 - drop 'results of'

and from tide gauge observations

23 .. in 1978 when Caspian MSL was at around this level (see below in Figure 7).

I would drop the following sentence. You don't discuss tide gauge data in 1978 - you have
discussed tide gauge data in the previous section so I don't see what this sentence
is for.

p7, 13-14 The amplitude of the other .. do not exceed

16 - I would put (Fig 3b) at the end of the first sentence. And I might (not sure)
drop the next sentence as you don't demonstrate the two Kelvin waves.

p10, 4 .. constituents, an 18.6 ... has been predicted ..

10 drop 'among other cities'

The maximum tidal range for the Caspian as a whole has been observed at Algada in
Turkmen ..

p11, 1 - in the overall sea level

4 was observed .. which has the greatest range of 21 cm mentioned above.

I was confused here what new results the second paragraph is presenting so
at line 8 I think you have to say 'based on the new modelling results'.

9 the tides alone; and 2) with meteorological forcing alone produced ..

I think is right? i.e. you don't have tides also in 2?

why do you have these 2 bands? You have to say what each one is sampling.

After paragraph 3 I was wondering, so what? I think you need a sentence at the end
simply saying something like the tides are shown to provide a significant
component of the overall variance. Hence important to understand. And
that should go in the Conclusions also.

p12, 7 axes

p13, 7 - use a comma and not semicolon after 'remain'

9 drop comma before too.

p14, 3 - I would put (Fig. 7) at the end of first sentence. Then I would move
the second sentence to after 'Southern Caspian' to be the penultimate
sentence. Then reword last sentence 'As a consequence, the spatial ..
should also be expected to change'.

Figure 7 x-label Time (Years)

The caption needs an extra sentence saying something like 'Numbers on the
contours indicate the depths of the shallowest areas below MSL in each case'.

p16, 2 - shifts to the east

5-7 I just don't see this. You don't show a map for -30 so it is hard to say,
but the map for -29 shows only a little area in red and I can't see 5.5 cm.

19 amphidromes

24 Ogurja Ada Island

Figure 9 - this is a nice figure but could there be a little more space between
(a,b) and (c,d).

p20, 4 - spectra on the western ... which is confirmed here by ...

13 I would move the bracket to the end i.e.

.. the island's height currently does not exceed 3-5 m (MSL of -27.5 m with
respect to the BHS).

18 the minus sign is on a different line to the 25 (there is some way in Word of
fixing that).

However, according to ..

19 of the island was in fact not flooded at all by

20 .. about 3.5 m). Therefore, there is some modelling uncertainty about this
aspect. According to Badyukova (2015), the island ..

32 could you add (energy) after Q-factor? Some people won't know what that is.

34 about 12 hours (i.e around the period of M2)

p21,1 Conversely, the Q-factor ..

Figure 10 - could this also have period on the x-axis as the text refers to periods
and not frequencies?

19 .. account data from long-term ..

20 simulations

22 certain areas

p22, 1 could you maybe replace 'permanent sea currents' with perhaps 'persistent currents due
to the hydrological regime (and reference needed).'

8 In recent decades,

11 don't --> do not

12 Please see the remark of R1 concerning TPX08 and MSL, andd then see below the email from
Svetlana Erofeeva. This sentence should be dropped and replaced with
a general statement on the sensitivity of the tide modelling on depth.

13 reword: ... but the MSL used to compute that model was incorrect (0 m with respect to the BHS). This invalid
.. and as a result will have introduced errors into the computed tide.

16 I would add a sentence such as.

In addition, we suggest that the findings of the present work be taken into account
when future global tide models are constructed.

Comments from Reviewer 1:

Comments on the revised version os-2019-88-manuscript-version3.pdf
Numerical modelling of the tides in the Caspian Sea
by Igor Medvedev et al.

(Anonymous Referee #1)

GENERAL COMMENTS

The revised version of the manuscript has significantly improved compared to the original
submission. All of my comments and suggestions have been accepted. Whether the
comments of the Editor and the other Referee - all justified in my opinion - have been
considered in an adequate way I cannot assess definitively.

I have one principal doubt remaining after the manuscript revision. This refers to the
Authors' statement that the incorrect Caspian MSL adopted by the TPXO8 model leads to
a distorted prediction of the tidal sea-level variation. I do not insist on my point, but I would
like to draw the attention of an expert more qualified than me to this issue: I understand
that the TPXO tide models are empiric models relying exclusively on satellite altimetric
observations. In this case the local tidal signal is retrieved from an analysis of time series
of sea-level VARIATIONS, independent from the absolute MSL height. Thus an incorrect
assumption on the MSL would indeed lead to an incorrect shore line and sea area, but
should not imply any corruption of the tidal parameters within the area of valid altimetry
data. Or are there any basin-dimension dependent computations involved in the TPXO8
generation?

Apart from this doubt, from my point of view the manuscript would be acceptable for
publication under the condition of a thorough proof reading (preferably by a native
speaker) in order to correct remaining minor flaws in style, grammar and sentence
structure; some of them are highlighted in the following list of technical corrections. I also
strongly recommend a homogenization of the writing regarding the application of past
tense vs. present (e.g., 11-4..6; but also elsewhere).

TECHNICAL CORRECTIONS

1-7: ... and a unique OBJECT for THE analysis of tidal dynamics
I prefer "object" over "subject", also in later occurences
2-29: In the present work we use...
4-8: This dataset IS based on...
6-5: which is used in Eq. (3) to determine...
6-6: The COMPARISON of the amplitudes (H) and phase lags of THE tidal
CONSTITUENTS calculated from ... with those based on the observations ...
6-8: of the amplitude of the M2 harmonic at Baku, ...
(also correspondingly at 6-12, 21-5)
6-11: for Ogurchinsky Island it reached 36° and for Krasnovodsk it amounts to 26°.
6-23: maximum coverage of tide gauge data.
7-12: period missing at the end of the sentence
10-10: Remove "other". -> largest tidal range among the cities listed in Table 2.
11-4: At Alagda, where the tidal range amounts to 31 cm,...
11-6: The smallest relative contribution of the tides was found on the western coast...
11-16: eastern
11-18: the tidal contribution TO the total variance
15-3: The figure caption has to be updated: the red line (Baku) has been removed.
16-3: with amplitudes OF up to 6.5 cm
16-19: ... separation of THE amphidrome near the Absheron Peninsula...
16-30: west coast of THE Northern Caspian
20-31: leads to A change
22-11: ... do not include the Caspian Sea.
22-12: Only the TPXO8 model includes the Caspian Sea, but it adopts an incorrect MSL of
0 m with respect to BHS.

An email from Svetlana Erofeeva:

We did not assimilate any data in Caspian Sea for TPXO8, and whatever tide is seen there ( <3cm)
is produced by solving SWE (with astronomic forcing) - like for any large body of water. Caspian
Sea bathymetry and coastal line in TPXO8 was probably too bad to rely on and we should have masked
it, but we did not pay attention, since it always looked like zero compared to what is in the Ocean.
Anyway this issue was corrected later, and Caspian Sea is masked in TPXO9.

I.e. the short answer is tide in TPXO8 Caspian Sea is wrong not because we used wrong MSL (since we
did not use it at all), but wrong bathymetry/coastal lines and also due to low 1/6 resolution of
TPXO8 base solution (no patch for Caspian Sea in TPXO8-atlas).

Igor - please see my email yesterday for a minor typo.

Many thanks again for submitting the paper.

Phil