14/2/2021

Comments on the revision of this paper.

This paper has been much improved following the comments of the reviewers, and
the figures have also been improved. I simply read the paper again and have a
remaining large number of small edits which should not take the authors long. There
are also a couple of more important issues.

If the authors disagree with any of these comments then I suggest they mail me
privately. Otherwise I assume they will be all attended to in a final revision which
I will not have read in detail again.

General comment - please see lower down but I got confused why you made comparisons
with FES2014a and not FES2014b which is the final product, isn't it? This needs explicitly
explaining.

Detailed comments:

pages 1-2. I ignored these as the title and abstract are slightly different from those
on page 3 which I took to be the official versions.

p3, title - performance

13 with tidal de-alising corrections with the best

15 last --> latest

drop 'and some accuracy assessments'

17 drop (Lyard et al 2006)

19 solutions,

21 .. augmented, and additional ... derived such as loading and self-attraction
parameterisations, energy diagnostics, and Lowest Astronomical Tide maps.

24 Geophysical Data Records

37 As will be

40 (the most commonly used norm is the L2-norm .... with a Gaussian ..

p4, 5 - Bennett

6 covariances

7 solution errors

17 The FES atlas series started

18 included

20 during the production phase

21-22 world-spread --> worldwide

that use the new atlas.

28 on a data assimilation

32 aimed --> intended

FES2014b was the first

35 geodesic --> geodetic

35-36 long period tide constituents

p5, 5 and a link

14 discretizations

20 Bell (1975) and Baines (1982)

21 accounts

24 drop 'Toulouse ,. models' The acronym has already been defined

(T-UGOm) --> T-UGOm

26 open boundary condition

27 elevations

p6, 16 elevations

31 Lynch and Gray (1977), and continuously developed since, the approach has
evolved from application to the deep [?] global ocean, now up to the inclusion of
near-shore ..

33 Although it allows

39 due to the Courant–Friedrichs–Lewy (CFL) stability

p7. 5 centred-in-time

11 according

15 - you refer to the model grids here, but there is no reference in the text to
the Supplementary Figure and its grid. It needs including either here or somewhere.

19 On the other hand, only minor

28 in all of this area

Following the FES2014 release

30 such as around Tierra de Fuego

31 in a future

35 give references to GEBCO, ETOPO and Smith and Sandwell

40 harmonic technique

etc.

41 in the absence

42 include patches derived from

p8, 8 Seas

13 merging

26 Using older bathymetry

33 They can be implicity

38 what does 'non-free' mean here? Please reword

40 version of FEES2014 (FES2014a) ... atlas was computed. As will be mentioned

p9, 9 T-UGOm

10 element vectors

12 Arctic Sea

26 model performance

32 have a clearly

33 insure --> ensure

35 give reference to ERA-interim.

I feel this paragraph is not complete. Or perhaps it needs a pointer to where the
S1 and S2 issue is discussed lower down.

37 TG and altimetry-derived

p10, 5 tidal gauge

10 drop the first 'respectively'

18 since the FES2014

19 ... missions, a satellite's orbit ..

20 move to

25 mean sea surface profiles computed over a 20-year

25 and 26 TOPEX-Interleaved and Jason-1-Interleaved (denoted TPN and J1N)

27 Geosat Follow-On (GFO)

32 with the longest value

p11, 26 However, the ocean

34 within the

35 temperature and salnity profiles

satellite Sea Surface Temperature (SST)

39 harmonic constant

p12,9 TG and altimetry

12 Deep moorings

16 aliased period

18 Because this signal was strongest during the TOPEX-POSEIDON mission - this
needs a reference also.

I think somewhere you have to make it clear what you mean by 'altimetry' i.e that it
has a routine IB correction, unlike tide gauges

lines 18-25 - I found these lines very hard to understand. Please could you look at and
reword?

28 Not only can the contamination by non-tidal signals at the aliased frequency be comparable

31 aliased frequency space.

minimum of seven years

what does (In TP-Jason mean)?

32 frequency space

33 is a mitigation

41 data gap duration

p13, 6 - what does point-by-point clearing mean?

data dismissing --> data rejection?

9 criterion for continuous

10 decide on a minimum

16 In the case when

23 creates

25 .. imperfect .. vulnerable to the baroclinic wave propagation angle with
respect to the ground track

26 baroclinic tide .. gravity wave

29 in the deep ocean

30 have short wavelength

32 The barotropic tide

36 filtering cut-off length scale

p14, 8 define CADOR

15 For this reason

16 error covariances of the state

33 - define IWD

34 - 'error compensation story' sounds odd. Can you reword?

sensitive component in the adjustment

35 strongest currents and dominates

37 in the presence

38 profit all .. the adjustment steps

p15, 21 define OGCMs

22 as shown in Figure 7.

32 displayed in Figure 8

36 with a large increase

40 Despite being considered

44 T-UGOm has

p16, 10 impact on the bottom friction

12 in regions where

23 ... series, and many more points

25 .. space), using data of the entire

29 as feasible of the

32 .. feasible; giving priority to the

33 ; and add (possibly ..)

36 where data could

37 data available during the

40 except in some

p17, 7 estimates ... components

14 presented in

17 ocean dynamic contamination

19 from the assimation

21 shown in Figure 10

24 process was obtained from several tidal data sets

25 the WOCE/GLOSS

26 an open ocean ... an Arctic

27 the International .. (IHO) data set, the research of Gjevik and Straume (1994),

29 Any inevitable redundancy due to neighbouring

30 neighbouring stations

31 For efficiency

32 numerical

34 constraints

38 experiences -> experiments

40 define CTOH acronym and give reference

43 estimates

p18, 16 is compared

18 sub-harmonic solutions, and the

24 getting into

the tidal prediction

28 drop 'development'

31 Stammer et al. (2014)

34 performance

34-35 sentence 'It might'. I don't understand this. I would drop it.

41 - the Doodson numbers of which are

p19, 6 - The MSf tide was

8 dented --> denoted

10 Despite providing the S1

13 reflect --> reflection of the mean of the S1

17 At present

18 .. based on a DAC from which the mean S1 atmospheric components have been
filtered out, and the ..

22 tidal correction for the next generation of

24 hydrodynamic

34 of the S1 tide .. with the DAC correection

35 The addition of

36 an S1

37 non-linear

p20, lines 5 and 13 - Mf etc. has a different order

24 As for the hydrodynamic simulations

30 tidal model should be very low

31 To a lesser extent

32 the amount of quality TG

p21, 6 The TPXO9-atlas

7 by combining the 1/6

10 vevy last --> latest

12 TPXO9-v2 --> TPXO9v2 to be consistent with above

13 taken advantage

14 hydrodynamic

17 drop '(denoted TG)'. already defined

19 gradual progression

23 tidal gauge

28 tidal gauge

29 to the GOT4v10

30 within the

33 TG --> TGs

35 independent TG information

36 10 TGs when using

p22, 13 Jason ones [i.e. drop the apostrophe]

14 FES2014a in particular

14-15 .. region, especially in the Laptev and Kara Seas

22 except for M2

23 the lack of a more comprehensive TG dataset

24 except for a slight rise

28 .. theory as appropriate to

32 the coast compared to that for the global ocean

36 local error

p23, 27 drop 'precise'

30 with regard to the computation .. which is comparable to ..

34 characteristic

p24, 11 (in which the built-in constraint for slope in topography imposes

13 verification

19 budget is a

22 internal tide

33 give references to these models

34 barotropic tide

38 self-attraction

p25, 11 - shown in

14 valuable parameter

18 (MHHW) tide levels

19 as mentioned in Pugh and Woodworth (2014)

26 too limited for

30 selected for the

33 between the

35 in the tidal ... besides

p26, 16 - surface wave

27 storm surge

28 storm surge

31 Desai et al. (1985)

33 far from being a solved problem

p29 - in the Holgate reference the journal should be Journal of Coastal Research and
not GJRAS

Figures 1 and 2 - please can you have more space between the upper and lower panels,
as in Figure 3?

Fig 2 caption line 1 and 3 - there are no red dots - there are black circles.

Also within the circle there is a sort of hand of a clock - say what that means

You say the size of the circles is proportional to the vector difference but you
ddo not say what the units are. Bottom left of each are symbols with some units but these
are not explained in thee caption.

line 4 - cm should be metres

Figure 3 - same comments as for Fig 2

Fig 4 - I noticed here you have US spelling (meters) and UK spelling (metres) elsewhere.
Please standardise.

Please say these model values correspond to IB-corrected sea level in altimetry and tide gauges
(presumably)

Fig 5 - Along-track ... kilometers)

Fig 7 - Divisions

Fig 11 - I don't understand the positions of the 4 Ray shelf points. Use semicolons to space them in
the list. What does GA R mean? Avonmouth and not Avon Mouth. I would drop 'BODC' or define it

Fig 13 - tidal gauge sites from the

Fig 14 tidal gauge

Some of these figures say they are compared to FES2014a and some to FES2014b. Why? I think you
explained that in the text but may be good also to mention also first time in a caption when used

Fig 17 - for the Altika (AL) mission .. to the coast .. of the FES2012 .. or instead of the
GOT4v10

Fig 19 - remove the stray script top right in the figure

why does the big ellipse near Adelaide have a gap? Looks like a plotting error

Fig 22, line 2 - cm2 should be superscript

insure --> ensure

last line - using the FES2014b

Fig 23 - why did you make an 18 and not 19 year run. Say units are metres

Table 3 line 2 - isobath

Supplementary Figure - as mentioned above, this needs to be referred to in the text.
Also the supplementary information needs to have a cover page which explains
the figure, and the reader has to be aware that he has to expand the figure to see
the grid or it is just black.

Florent - thanks for the edits and I am happy to accept the paper now. I spotted in the latest pdf a couple of remaining edits as below. Thanks again for all this. Phil

P12, line 25 – in turn
P19, line 5 – atlas  atlases
P29, line 29 – there is a stray javascript mention here
P30, line 24 – there is a stray Wikipedia reference here