This paper examines the change in the North Pacific subtropical mode water (NPSTMW) from 2018 to 2021 using Argo float observations. The authors found that the NPSTMW volume decreases, though the Kuroshio Extension (KE) is in a stable state, and then showed that it is attributable to the change in the westerlies associated with PDO and intensified ocean stratification. The results are overall interesting. Below are comments the authors may want to consider in the revision. In addition, I would like to suggest the authors to ask native speakers of English to check and correct the English language usage. It is sometimes difficult to follow correctly what is mentioned. 

1. As cited in the introduction, Oka et al. (2021) investigated NPSTMW using Argo floats observations. They showed that annual-mean volume actually increased in the northeast region (28–40ºN, 140ºE–170ºW) from 2017 to 2019, while that in the other region decreased. The northeast region includes the main formation region of NPSTMW with densities of larger than 1025.2 kg/m3. This is important difference and needs the comparison with Oka et al. (2021).

2. As shown by Qiu et al. (2020) and Qiu (2023), the stable KE and its northward shift is a combined response to the Kuroshio LM south of Japan and the Rossby wave forcing from the interior wind stress across the North Pacific. The KE northward shift can affect the overlying storm track and basin-scale wind field (Qiu et al. 2020). Is the northward shift of the westerlies, shown by the authors, associated with the stable KE? Or is it due to the PDO? Some additional analyses will be necessary. In addition, are the surface warming and intensified stratification also due to the northward shift of the KE?

Qiu et al. (2020), JC, doi: 10.1175/JCLI-D-20-0237.1

Qiu (2023), Oceanography in Japan, doi: 10.5928/kaiyou.32.3-4_67

3. Davis et al. (2011) showed no correlation between NPSTMW and wintertime surface cooling in the RG region, consistent with Qiu and Chen (2006). However, the decline in the NPSTMW volume after 2018 seems to be caused by wintertime surface cooling. Please discuss the correspondence with Davis et al. (2011) and Qiu and Chen (2006).

4. L173-176. It is hard to see the changes from Fig. 4a. In addition, are these changes statistically significant?

5. L219-228. The differences between 2012-2015 and 2018-2021 are not clear from Fig. 7. Please revise the figures and show the significance together. 

Other points

Please check abbreviations and if they are used throughout the text or if they are not defined more than once. For example, the mixed layer is defined as ML at L33, but later it is still used (e.g., L138). PV is defined at L26 and 55. 

RG dataset (L89) and RG region (L113). The same abbreviation is a bit confusing. 

L136. What do the authors cite here by Oka et al. (2011)?

L136. Does ERA5 provide wind stress data? Do the authors calculate wind stress from the surface wind data?

Equation 6. Is M a function of not only sigma theta but also time?

Section 3.1. Please show a stable and unstable KE period together with Fig 4. 

L169-170. What do authors mean by "Except for … "?

Figures 5 and 7. Winter average, not annual average?

L200-201. Do Suga and Hanawa (1990) pointed out these things?

L207-208. This sentence seems to be different from what the authors say at L199-200. 

L242-244. The northward shift is obvious in not only the dense isopycnal but also the light isopycnal. 

L252. What is RGA?

L262. What is the ventilation region here? Is this the RG region?

L264-265. I am not sure if this is supported by Fig. 9. 

L273. "Except for" should be "In addition to"

Fig. 12. Why is the wind stress curl masked along the international dateline?

Fig. 12. What do the authors mean by ensembled? Do the authors use ERA5 ensemble mean data?

L308. It is hard to see the change in the westerlies from Fig. 6. Instead of the wind stress curl, the authors may want to show the magnitude of the wind stress. 

L314-315. The intensification of the surface stratification seems to be different from what we expect from the deepening of the main thermocline in Figure 11. Is the stratification indeed weakened in the depth of the thermocline after 2018?

The authors focused on a recent decline of subtropical mode water (STMW) formation/volume in 2018−2021 when the Kuroshio Extension (KE) is in the stable state which usually results in an increased STMW formation. They concluded from analyses of grided and individual Argo float data and atmospheric reanalysis dataset that shallow winter mixed layer due to reduced wintertime heat loss associated with large-scale atmospheric circulation changes caused the decreased STMW formation/volume in 2018−2021. I think that the topic addressed in the present study is interesting. However, their analyses lack some important points. Additional discussions are needed to support their conclusion and make the present study more meaningful. In addition, descriptions are felt to be insufficient in many aspects. Explanation of results should also be improved to make the readability better. Further, there are many careless errors in the text.

Major comments

1.    It should be emphasized a little more that the switch to the KE stable state in 2017 is not wind-forced but is a response to an occurrence of the Kuroshio large meander, whereas the stable KE path during 2012−2015 results from wind stress forcing (Qiu et al., 2020). In other words, although the KE is in the stable state in both periods of 2012−2015 and 2018−2021, backgrounds are not necessarily the same.

Reference:

1.    Qiu et al. (2020): On the reset of the wind-forced decadal Kuroshio Extension variability in late 2017. Journal of Climate, 3, 10813-10828. doi:10.1175/JCLI-D-20-0237.1

2.    An impact of wind-induced stratification changes on the STMW formation should be carefully considered. Figure 11a shows Pacific decadal oscillation (PDO) is positive during 2014−2017. This means that upper ocean stratification in the central North Pacific is intensified (i.e., shoaling of the main thermocline is caused). Such signals propagate westward as oceanic Rossby waves and affect the STMW formation region during 2018−2021, considering a 4-year lag suggested by a previous study (Sugimoto and Kako, 2016 cited in the present study). According to Sugimoto and Kako (2016), this may also influence the STMW formation decline during 2018−2021 by hindering deep convection and then development of winter Mixed layer.

3.    The authors focused mainly on formation/volume (and density) of STMW. I suggest that STMW temperature should also be intensively examined as well. According to previous studies (and Figure 13a in the present study), temperature of subducted STMW (i.e., summertime STMW) tends to increase (decrease) during the unstable (stable) KE when the winter mixed layer becomes relatively shallow (deep) (e.g., Oka et al., 2019 cited in the text; see their Figure S5). However, Figure 13a of the present study indicated different feature of STMW temperature between 2012−2015 and 2018−2021. This is also remarkable and is related to the interest of this study. Comprehensive analyses not limited to STMW formation/volume would give us much deeper understandings of STMW.

4.    Descriptions are too focused on periods of 2012−2015 and 2018−2021. Although I understand that the authors have an interest in STMW in 2018−2021 and compare it with that in 2012−2015, discussions can be started from overall features. Explanation with a wider view would be better.

5.    I concern that STMW in 2018−2021 is compared with only one period of the KE stable state (2012−2015). Besides 2012−2015, the KE is in the stable state during 2003−2005. If the authors use another gridded dataset using Argo float data (MOAA-GPV; Hosoda et al., 2008) in addition to data by Roemmich and Gilson (2009), it is possible to further compare STMW formation/volume in 2018−2021 with another KE stable period. Although the number of Argo float profiles is not large in early-2000s, a previous study discussed STMW using Argo float profiles since 2003 (Sugimoto et al., 2013).

Reference:

1.    Hosoda et al. (2008): A monthly mean dataset of global oceanic temperature and salinity derived from Argo Float observations. JAMSTEC Report Research and Development, 8, 47-59. doi:10.5918/jamstecr.8.47

2.    Sugimoto et al. (2013): Marked freshening of North Pacific subtropical mode water in 2009 and 2010: influence of freshwater supply in the 2009 warm season. Geophysical Research Letters, 40, 3102-3105. doi:10.1002/grl.50600

6.    The authors pointed out importance of Ekman transport in the text. But Ekman heat advection is not discussed. It can be calculated from wind stress and temperature data, and further its effect may be considered in eq (5).

7.    Timeseries of winter NHF and heat flux due to Ekman transport in the STMW formation region is worth discussing and being compared with late-winter MLD and subduction volume.

8.    How about to add horizontal bars or something to panels of timeseries in Figures 4, 9, 10, 11, and 13 in order to show periods of stable/unstable states of the KE? Further, in those panels, average values of 2012−2015 and 2018−2021 is better to be plotted to be compared easily because their difference is focused throughout the present study.

Other comments

1.    L12 and many places

“−” instead of “-”.

2.    L31.

To explain a role of STMW as the memory of atmospheric conditions, it would be better to mention the reemergence of sea surface temperature anomalies (e.g., Hanawa and Sugimoto, 2004; Sugimoto and Hanawa, 2005).

Reference:

1.    Hanawa and Sugimoto (2004): ‘Reemergence’ areas of winter sea surface temperature anomalies in the world’s oceans. Geophysical Research Letters, 31, L10303. doi:10.1029/2004GL01990

2.    Sugimoto and Hanawa (2005): Remote reemergence areas of winter sea surface temperature anomalies in the North Pacific, Geophysical Research Letters, 32, L01606. doi:10.1029/2004GL021410

3.    L59

“Cerovečki and Giglio (2016)” Instead of “(Cerovečki and Giglio, 2016)”.

4.    L70-73

Information written in the text is not up-to-date. The recent Kuroshio large meander persists for more than six years and is the longest record in the past 70 years (Qiu et al., 2023). Even if the present study would examine STMW before 2022, this should be written correctly.

Reference:

1.    Qiu et al. (2023): Why did the 2017 Kuroshio large meander event become the longest in the past 70 years? Geophysical Research Letters, 50, 32023GL103548. doi:10.1029/2023GL103548

5.    L83 (Figure 1 caption)

“Copernicus Marine Environment Monitoring Service (CMEMS)” instead of “CMEMS”. Figure 1 is referred to at L70, but full name description of the CMEMS in the text is at L99.

6.    L95, L158, L213 (Figure 6 caption), …

Unify the abbreviation of net surface heat flux (HF or NHF).

7.    L103

“RG dataset” and “RG region” are so confusing. I suggest to use other words. Further, it is not appropriate that the abbreviation of RG region is used in the caption of Figure 2 (referred at L94) before it is defined in the main text (at L113).

8.    L141

Declare what is Q in the text explicitly.

9.    L169-170

I cannot understand this sentence.

10.    L199

Does the “the averaged area of the NPSTMW outcrop window” mean the “the averaged formation area of the NPSTMW”? If so, write so somewhere.

11.    L199

I felt that descriptions at L199-200 and at L207-208 contradict a little. The authors may want to improve the explanation.

12.    L213, L240, …

The word of “winter” and ”wintertime” should be defined uniquely in the text at the first reference. In the current manuscript, two definitions (January−March and February−March) exist.

13.    L221-L223

Show the value of transformation rates and/or their difference in the lighter NPSTMW range to support the result.

14.    L230 (Figure 7 caption)

“(a), (c) Annually averaged surface transformation rate and …” instead of “Annually averaged (a) and (b) surface transformation rate and”

15.    L223-

Does it mean that the STMW formation is concentrated on its lighter density range? If so, write so somewhere explicitly. Further, explain why the STMW formation is concentrated on the lighter range.

16.    L227-228

Need more explanation.

17.    L242-245

The KE has shifted northward during 1993−2021 (e.g., Wu et al., 2021 cited in the manuscript; Kawakami et al., 2023). I suggest to discuss a relationship to the KE shift.

Reference:

1.    Kawakami et al. (2023): Northward shift of the Kuroshio Extension during 1993−2021. Scientific Reports, 13, 16223. doi:10.1038/s41598-023-43009-w

18.    L254

Suggest to discuss the mean MLD distributions during winter of 2012−2015 and 2018−2021 rather than showing only their difference.

19.    L264

Subduction volume depends on MLD probably not only 2018−2021 but also throughout the analysis period. Although I recognize that the main target of the present study is STMW formation in 2018−2021, how about start a discussion from an overall relationship between MLD and STMW formation by performing a correlation analysis or something else?

20.    L276

“Figure 4b” instead of “Figure 2b”.

21.    L283

“Figures” instead of “Figure”

22.    L285 (Figure 11)

It would be better to locate the panel (a) at the left-hand (or right-hand) side of panels (b) and (c) with rotation of 90º as done in Qiu et al. (2023) (mentioned above) so that readers can compare the PDO index and sea surface height (SSH) and main thermocline depth (MTD) anomalies easily.

23.    L285

In panel (c), was the global-mean sea level rise due to freshwater flux and thermal expansion removed?

24.    L294

Specify where the central North Pacific is in Figure 12a.

25.    L301 (Figure 12)

The Order of panels are unnatural. Consider to rearrange as (a) 2012−2015, (b) 2018−2021, and (c) their differences. Further, sea level pressure (SLP) can be superimposed by contours. SLP helps the readers to find the Aleutian Low discussed in the text.

26.    L313

Specify what is stratification anomaly. How was it calculated?

In this paper, the authors examine recent interannual variability of North Pacific Subtropical Mode Water (NPSTMW) and its causes.  At least for more than two decades until 2016, the Kuroshio Extension (KE) shows decadal variability between its stable and unstable states, which is a delayed response to the atmospheric forcing in the central North Pacific associated with the Pacific Decadal Oscillation.  This decadal variability of KE has mainly controlled that of NPSTMW.  However, such decadal KE variability was reset in summer 2017 when the Kuroshio large meander started south of Japan and has forced KE to be in its stable state until now.  The authors have analyzed gridded and raw Argo data to demonstrated NPSTMW decline after 2018 in spite of the KE being stable. 

I understood the authors’ story that less NPSTMW was formed in 2018-2021 compared to 2012-2015 due to less atmospheric cooling and the resultant smaller MLD.  However, Figure 13a shows that negative temperature anomalies during 2012-2013 when the PDO index was negative as in 2018-2021.  Isn’t there a possibility that positive temperature anomalies during 2018-2021 are due to horizontal heat advection from the meandering Kuroshio south of Japan?  Moreover, isn’t this heat advection the major cause of the less formation of NPSTMW after 2018?  This is the only major comment from me.  Otherwise, the manuscript is relatively well written, and I recommend its publication in OS after the above issue is addressed.

Eitarou Oka

Other, minor comments:

L33: south of the Kuroshio Extension (KE) -> south of the Kuroshio and the Kuroshio Extension (KE)

L70-72, “In recent years, KE is in a stable state associated with the Kuroshio large-meander (LM) path south of Japan (Figure 1). Although a persisting Kuroshio LM and the resultant stable state of the KE has already exceeded four years (Qiu and Chen, 2021; Usui, 2019), the NPSTMW volume has declined since 2018 (Oka et al., 2021).” I would expect more detailed explanation here for readers’ understanding.  I would write, “In August 2017, KE switched from an unstable state to a stable state in association with the occurrence of the Kuroshio large-meander (LM) path south of Japan (Figure 1), although negative SSH and MTD anomalies associated with the positive PDO phase were arriving from the central North Pacific (Qiu et al. 2020, JC).  Since then, Kuroshio LM and the stable state of the KE have lasted for more than six years (Qiu and Chen, 2021; Qiu et al., 2023; Usui, 2019), while the NPSTMW volume has declined (Oka et al., 2021).”  Note that the current stable KE state seems to have begun with the initiation of LM (Qiu et al., 2020), it has also been supported by basin-wide wind forcing (Qiu et al., 2023, GRL).

L95, “net surface heat flux (HF)”: This notation is somewhat misleading in eq. (5) (L158) because it looks like a product of H and F.  Consider using a single character.

L108, “sigma-theta = 25.0-25.5 kg m-3”: To what temperature range does this density range correspond?  I am just curious if the warmest (lightest) variety of NPSTMW formed south of Japan, especially in the Kuroshio LM period (Nishikawa et al., 2023, JO), is included in the authors’ analysis.

L169-170, “Except for a short time of 2006-2009 when the KE jet is unstable, the NPSTMW volume has a dramatic decrease during 2006-2009.”: I do not understand.  Do the authors mean, “In a short time period of 2006-2009 when the KE jet is unstable, the NPSTMW volume has a dramatic decrease.”?

L221-223, “transformation rates … were greatly reduced.”: Not obvious for me from Fig. 7a,c.

L223-225, “the annually averaged surface formation rates … density range”: Not obvious for me from Fig. 7b,d.

L273, “Except for the KE dynamic state change,”: I do not understand.  Maybe, “In addition to the KE dynamic state change,”, although the KE dynamic state and “oceanic precondition” are not independent from each other.

L311, “is leaded”: “is led”?