We are so grateful for the anonymous referee’s comments. They carefully reviewed our manuscript and provided constructive and helpful comments. The followings are in response to their comments in sequence. The blue font indicates newly added or corrected sentences in the revised manuscript.

Main discussion and comments:

1. One scientific issue:

The reviewers are correct that there are some sea-level indicators (SLI) and terrestrial limiting records (TL) reaching over +5 m. Specifically, we have 3 SLI at Wando-Sinji 3 (WALIS LUM ID #464: +6.1 m), and Sacheon-Daepodong (WALIS LUM ID #460: +6 m; #461: +6 m) in the western and southern areas (Table 3 and Figs. 2 and 7). We also have the lowest TL of Haenam-Ijin-ri 3 (WALIS LUM ID #467: +5.4 m) and the lowest SLI of Wando-Sinji 1 (WALIS LUM ID #462: +2.8 m) in the western areas. In the eastern area, we didn’t consider LIG sea-level elevation data because of neotectonic uplift since MIS 5e. To address this comment, we changed the text to suggest MIS 5e sea levels lie between +3 and +6 m in the revised manuscript, strictly based on the SLI and TL data, discarding our previous intuitive height of 2 to 5 m.

1. Revising some figures and figure captions:

We revised all figures and figure captions that they had commented on. The following provides our responses to their specific comments.

Specific comments:

1. Figures 2 and 7:

We used an inverted triangle legend for the terrestrial-limit site. The following sentences were added in the figure caption: Red circles represent sea-level indicator locations based on OSL ages with exception of the Pohang-Masan-ri area which has an additional paleomagnetic-based age. An inverted triangle shows the location of a terrestrial-limiting point. Age data can be found in Table 3. Bathymetric contour inveral is in meters.

In Figure 7, the following sentence was added in the figure caption: A red circle represents the location of sea-level indicator while blue triangles and brown inverted triangles represent marine-limiting and terrestrial-limiting site locations, respectively. Age data can be found in Table 3. Bathymetric contour inveral is in meters.

The subsite names of sample sites were added in the revised Figure 2, which accords to those in Table 3 and the open-access spreadsheet (Ryang and Simms, 2021).

1. Figure 3:

We corrected Geumgok-ri 1105 GG to Geumgok-ri 1 and 2 in Figure 3, which accords in the subsite name in Table 3 and the open-access spreadsheet (Ryang and Simms, 2021).

1. Figure 4:

A sentence was added in the figure caption as follows: Two ages in red indicate LIG ages (WALIS LUM ID #449 and #450 in Table 3).

1. Line 221 and 222:

The sentence was rewritten as it follows: where U_l represents the upper limit of the modern analog landform’s elevation, L_l represents the lower limit of the modern analog landform’s elevation, SWSH represents storm wave swash height, and d_b represents the breaking depth. Each of these values was obtained from local sources (Table 1).

In addition, the sentence in Line 219 was rewritten as it follows: Where E represents the elevation of the sea-level indicator, RWL represents the reference water level, and δ_RSL accounts for the uncertainties in the paleo RSL (Rovere et al., 2016).

1. Line 283:

We put all subsite names in Table 3 on Figures 2 and 7 as a set of the site and subsite names.

1. Line 298, 308, 315, 324-325, 333, 337, 345, 360, 365, 374, 381, 388-389:

Same responses for Line 283.

1. Line 397:

Same responses for Line 283 and we corrected the box name in the western area on Figure 1.

1. Line 409:

Same responses for Line 283.

1. Figure 11:

(1) A sentence was added in the figure caption as follows: Orange shading highlights the correlation of LIG deposits across the 6 coastal cores and Wando outcrop sections..

(2) We changed ‘West’ and ‘East’ to ‘Seaward’ and ‘Landward’, respectively.

(3) We changed ‘WD1 & SJ’ and ‘Wando-Sinji 1 and 3’, which accords in the subsite name in Table 3, Figure 7, and the open-access spreadsheet (Ryang and Simms, 2021).

(4) A sentence was added in the figure caption as follows: Ages in red indicate LIG ages (WALIS LUM ID #470 to #491 in Table 3).

1. Line 621:

Same responses for ‘1. One scientific issue’ in the upper Main discussion and comments.

I’m sorry for some late upload of the revised manuscript. I am not familiar with the ESSD editorial system as a first-time user. Now you can refer to the revised manuscript with colored fonts in where you, other reviewers, and authors rewrote and corrected.

Thank you again.

Woo Hun Ryang

We cordially appreciate the comments by Drs. Colin V. Murray-Wallace and Dan Muhs. They meticulously reviewed our manuscript and provided valuable and helpful comments. The followings are in response to their comments in sequence. The red font indicates newly added or corrected sentences in the revised manuscript.

Commented [MDR1]: The Kopp et al. study contains a complex mix of tectonically active and stable coasts, dated and undated coasts, etc. The first real recognition of the higher-than-present LIG sea level was H.H. Veeh, in his 1966 JGR paper (and actually, not much has changed since that groundbreaking paper).

Reply [1]: Thanks for your comments, we cited Veeh's (1966) JGR paper.

Commented [MDR2]: What do you mean by "compaction" here? Sediment compaction?

Reply [2]: It is rewritten to ‘sediment compaction’.

Commented [MDR3]: Are these marine sands and gravels? Terrestrial? Both?

Reply [3]: It is rewritten to ‘~ unconsolidated marine and aeolian sands and well-rounded beach gravels ~’.

Commented [MDR4]: Very nice map!

Reply [4]: Thank you for your comments. It was modified after Chough et al. (2000; 2004) and Cummings et al. (2016).

Commented [MDR5]: Also a very nice map, but I suggest that here, as in later figures, use WHITE for land (as you did in Figure 1) and use BLUE (whether one shade or more) for the ocean.

Reply [5]: We redrew Figure 2 referring to the drawing style in Figure 1.

Commented [MDR6]: No need for the second sentence (legend is understood). Can you state what the contour interval is? Also, if you make the terrace shades partially transparent, readers can see the contour lines.

Reply [6]: We deleted the second sentence and stated the contour interval of 10 m in the captions of Figures 3, 4, and 5. Unfortunately, we could not make the terrace shades partially transparent because the figure was covered by the terrace shades in the original paper.

Commented [MDR7]: Please give the contour interval on this map.

Reply [7]: We stated the contour interval of 10 m in the figure caption.

Commented [MDR8]: I do not see the "palaeo-beach pebble gravel" unit in either section. For the NQ t1 terrace, what is the yellow unit?

Reply [8]: We found that marks of the paleo-beach pebble gravel were covered by the green and blue shades in the columnar sections when it was modified. We corrected it. For the NQt₁ terrace, a sentence was added in the figure caption as follows: Sites YH09 and YH10 in the NQt₁ terrace (yellow) were dated to 0.09±0.01 ka and 0.11±0.01 ka by OSL, respectively (J.H. Choi et al., 2009).

Commented [MDR9]: say either "aeolian sand" [deposits] or "dunes" [landforms], but not both. Same comment for the next sentence.

Reply [9]: The word ‘dunes’ was deleted in the two sentences.

Commented [MDR10]: How about showing a photograph of these terraces?

Reply [10]: Unfortunately, the original paper of J.H. Choi (2009) in Quaternary International does not show any photograph of these terraces.

Commented [MDR11]: A major concern here (and this applies to equations (1) and (2), which you discuss later. As you show here, the "Paleo Seacliff Onlap Elevation" is indeed a maximum-limiting elevation for the shoreline angle elevation, complemented by the minimum-limiting platform elevation. However, I have never seen a situation in the field that looks like this. What you have shown here is what appears to be a marine deposit of ONE age draped over wave-cut platforms of TWO ages, including the paleo-sea cliff itself. In 40 years of studying marine terraces on three continents and many islands, I have never encountered such a situation. Now, it is very common for NON-marine deposits to be draped over two or more platforms and the intervening paleo-sea cliff, but not this. What is usually found, however, is a paleo-sea cliff back a platform, and the maximum-limiting elevation is the lowest exposed bedrock elevation on that paleo-sea cliff. That measurement, plus the elevation of the lower wave-cut platform will indeed bracket the elevation of the shoreline angle if it is not exposed.

Reply [11]: Thank you so much for your kind and detailed comments. Your comments on the overlying deposits are right. We had an error in drawing the schematic diagram, so Figure 6 was redrawn adding draped aeolian sands and nonmarine deposits over the marine deposits and paleo wave-cut platform.

In our study area, unfortunately, the paleo-sea cliff back a platform was rarely exposed due to a gradual topographic slope. However, seismic profiles on land showed a subsurface paleo-sea cliff, not exposed on the surface (Figure 5; J.W. Kim et al., 2007a). Thus, we had no choice to use the lowest exposed bedrock elevation on the paleo-sea cliff but should use the paleo seacliff onlap elevation (PSOE) in Figure 6 and equations 1 and 2.

Commented [MDR12]: The same comment as before: make the land white and the ocean blue.

Reply [12]: We redrew Figure 7 referring to the drawing style in Figure 1.

Commented [MDR13]: Here you need to state that the tops of the fluvial terrace fills have elevations that correspond to (i.e., can be traced to) the shoreline angles of marine terraces nearer to the coast. Otherwise, there is no confidence that the fluvial terraces are eustatically controlled.

Reply [13]: Although S.G. Choi (1993; 1996) insisted that the fluvial terrace fills with LIG ages correspond to the shoreline angles of nearby marine terraces, he did not present any scientific evidence of the correspondence in his papers. So, we include his contribution only as historically published data. This work presented the first two numerical LIG ages for coastal areas of the Korean Peninsula, but the ages were not included in our dataset of Table 3 and the open-access spreadsheet (Ryang and Simms, 2021).

Commented [MDR14]: Please don't use the phrase "age dating", which is redundant.

Reply [14]: We corrected the phrase.

Commented [MDR15]: Here, and throughout the manuscript, please delete all occurrences of the phrase "absolute age" and instead please use the term "numerical age", which is preferred by geochronologists.

Reply [15]: We changed the phrase ‘absolute age’ to ‘numerical age’ throughout the revised manuscript.

Commented [MDR16]: Are there no marine fossils in the marine terrace deposits of the Korean Peninsula anywhere? Is there no possibility of amino acid geochronology? If that is the case, you need to state this.

Reply [16]: Unfortunately, we are not aware of any geochronological studies of the marine fossils in the marine terrace deposits of the Korean Peninsula.

Commented [MDR17]: For equations (1) and (2), please see the comment made in the caption for Figure 6. This is a MAJOR issue.

Reply [17]: You can refer to Reply [11] for Figure 6.

Commented [MDR18]: For the age uncertainties, are these ONE sigma or TWO sigma? This needs to be stated explicitly.

Reply [18]: We inserted ‘error range’ of ± 1σ SE or ± 2σ SE into the last column of age uncertainty of Table 3, and the corresponding error range was written in each cell of the age uncertainty.

Commented [MDR19]: I believe you said all this earlier.

Reply [19]: We wrote this already as the first sentence in the 2.2 Overview section. This sentence is restated for the 4.1 and 4.2 sections.

Commented [MDR20]: Again, are these ONE sigma or TWO sigma errors? You need to be clear about this before going into all the site-specific details of the ages.

Reply [20]: We inserted ‘error range’ of ± 1σ SE or ± 2σ SE into the last column of age uncertainty of Table 3, and the corresponding error range was written in each cell of the age uncertainty. We also inserted the following sentence at the end of 3.3.1 Optically stimulated luminescence subsection: the error range had ± 1σ SE or ± 2σ SE according to measurement methods.

Commented [MDR21]: Here, and in the previous paragraphs, why do you say this? Obviously, there has been uplift…some readers are going to be confused by this discussion and assume you really think that sea level during the LIG was actually this high.

Reply [21]: We rewrote these sentences in terms of the SA elevation in subsites of Gangneung-Anin, Donghae-Eadal-dong, Pohang-Yonghan-2, Pohang-Masan-ri, and Gyeongju-Jinri. Table 2 and Figure 14 were also revised.

Commented [MDR22]: What sort of shells? Marine? Is this the only place they are found?

Reply [22]: Unfortunately, the few shell fragments are too small in both size and quantity to identify in the core.

Commented [MDR23]: Any geochronology on these shells? Any indications of paleozoogeography that could be useful checks on the OSL ages?

Reply [23]: Unfortunately, no paleontological descriptions and studies are using the shell beds in the original paper.

Commented [MDR24]: "swaley" is not a word

Reply [24]: The phrase ‘faint swaley cross-bedding’ was rewritten to ‘swaley cross strata’. From page 643 of AGI’s 4^th edition of the Glossary of Geology: “swaley cross strata: cross strata resembling hummocky cross strata but dominated by convave-up strata deposited in swales instead of by convex-up strata deposited on hummocks.”

Commented [MDR25]: What species of oysters? Any useful paleozoogeographic information?

Reply [25]: Unfortunately, no paleontological descriptions and studies are using the oysters in the original paper.

Commented [MDR26]: Why not give some specific examples and some possible uplift rate calculations? For example, in the Subsite Yonghan-2 site of the Pohang area, it looks like the NQt3 terrace has a shoreline angle elevation of ~35 m above sea level and dates to MIS 5e. Using a suite of possible LIG paleo-sea levels (+3 m, +6 m, +9 m, etc.), you could present a range of possible uplift rates. Is NQt1 Holocene? If so, is that a storm deposit or a coseismically uplifted terrace?

Reply [26]: At the end of the 5.1 Uplift section, we wrote a paragraph for the uplift rate contents adding Eq. (12) and Table 4. For the NQt₁ terrace, a sentence was added in Figure 4 caption as follows: YH09 and YH10 sampling sites in the yellow NQt₁ terrace were dated to 0.09±0.01 ka and 0.11±0.01 ka by OSL, respectively (J.H. Choi et al., 2009).

Commented [MDR27]: These are not really "relative sea levels," they are present elevations of marine terraces.

Reply [27]: To address this comment, we revised Table 2 and Figure 14 in terms of the SA elevations. We also rewrote the related sentences in subsites of Gangneung-Anin, Donghae-Eadal-dong, Pohang-Yonghan-2, Pohang-Masan-ri, and Gyeongju-Jinri.

Commented [MDR28]: I am puzzled here. The entire paper is on the subject of LIG sea level and your summary of it all is two sentences? Why not say something about the range of ages you have and compare them to U-series-dated and OSL-dated marine deposits elsewhere?

Reply [28]: The section title of LIG sea-level fluctuation means fine-scale sea-level fluctuations during LIG, which shares the common section title in many papers of the WALIS special issue. We think our data in the Korean Peninsula are too coarse to constrain fine-scale fluctuations in LIG sea levels.

Commented [MDR29]: You still haven't said whether these errors are one sigma or two sigma, so this discussion is not going to be easy to interpret, I'm afraid.

Reply [29]: We inserted ‘error range’ of ± 1σ SE or ± 2σ SE into the last column of age uncertainty of Table 3, and the corresponding error range was written in each cell of the age uncertainty.

Commented [MDR30]: Rephrase, removing "distorted" and "contamination", please.

Reply [30]: We removed them in the revised manuscript.

Commented [MDR31]: What do you mean by "coined"?

Reply [31]: The word was rewritten to ‘initiated’.

I’m sorry for some late upload of the revised manuscript. I am not familiar with the ESSD editorial system as a first-time user. Now you can refer to the revised manuscript with colored fonts in where you, other reviewers, and authors rewrote and corrected.

Thank you again.

Woo Hun Ryang

Authors’ response to comments by reviewer Craig Sloss:

We cordially thank Prof. Craig Sloss for reviewing and his comments. We have done our best to revise the manuscript according to his corrections and comments. We have provided a reply by subject to the comments in the attached document. Although we have revised the manuscript according to most of his comments, some suggestions could not be revised as follows:

Commented [Line 33, 61, 65, and many lines]: S.J. Choi, 2019; G.Y. Lee and Park, 2019, S.H. Yoon and Chough, 1995, S.W. Kim, 1973, etc.

Reply [1]: In Korea, many people have the same surnames such as Choi, Lee, Yoon, and Kim. Our paper has about 30 first authors who have the same surname in our references. Korean readers, as well as international readers, will confuse cited authors while reading the manuscript if given name initials are not provided. Therefore, we prefer including given name initials to help distinguish authors (and sources) in the case where multiple references contain the same surname.

Commented [Line 139]: opps there it is, ignore but still on what??; [Line 186]: What about the Amino Racemization?

Reply [2]: We have inserted the following sentence in line 141: “However, due to inadequate descriptions of the sampled materials and age dating techniques within the original papers, the two ages were not included in our dataset shown in Table 3 or in the open-access spreadsheet (Ryang and Simms, 2021).”

Although S.G. Choi (1993; 1996) insisted that the fluvial terrace fills with LIG ages correspond to the shoreline angles of nearby marine terraces, he did not present any scientific evidence for such in his papers. Thus, we include his contribution only as historically published data. This work presented the first two numerical LIG ages for coastal areas of the Korean Peninsula, but the ages were not included in Table 3 or in the open-access spreadsheet (Ryang and Simms, 2021). This is why we did not introduce the amino acid racemization method in the 3.3 Dating techniques section.

Commented [Line 185]: any faunal elements?

Reply [3]: Unfortunately, faunal elements were not described and studied in the original paper.

---------------------------------------------------------------------

You could refer to the following paragraph.

2.2 Overview of previous studies on the Korean Peninsula

In the northern region, the first LIG age (124 ka B.P.) was obtained from a fluvial terrace in the downstream part of the Seomseok river, located in Gangneung (Fig. 2), using amino acid racemization dating of peats (S.G. Choi, 1993). In the southern region, another LIG age (125 ka B.P.) was obtained from a separate fluvial terrace, located near the Pohang shoreline (Fig. 2), using the same method (S.G. Choi, 1996). However, due to inadequate descriptions of the sampled materials and age dating techniques within the original papers, the two ages were not included in our dataset shown in Table 3 or in the open-access spreadsheet (Ryang and Simms, 2021). Since this initial work, most research has focused not on numerical dating, but documenting the elevations, sedimentary characteristics, and stratigraphy of deposits overlying the marine terraces along the eastern shorelines (e.g., S.G. Choi, 1995a, 1995b, 2016a, 2016b, 2018, 2019; S.G. Choi and Chang, 2019; Hwang and Yoon, 1996, 2020; Hwang et al., 2012; S.O. Yoon et al., 1999, 2003, 2014).