the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 27 Feb 2025
Salinity Trends and Mass Balances in the Mediterranean Sea: The Role of Air-Sea Freshwater Fluxes and Oceanic Exchange
Abstract. Understanding the drivers of salinity and mass variability in the Mediterranean Sea is essential for assessing regional climate impacts and elucidating climate-driven changes in the water cycle. Although it is possible to close the Mediterranean mass and salinity budgets within uncertainty ranges, the relative contributions of key boundary fluxes, namely, surface freshwater fluxes (Evaporation-minus-Precipitation-minus-runoff) and water exchanges through the Strait of Gibraltar, remain unclear. To address this, we analyzed the Mediterranean’s mass and salinity budgets of 2003–2017 using ECCOv4r4. Our findings reveal a delicate balance between boundary fluxes that jointly regulate the Mediterranean’s mass and salinity dynamics. Surface freshwater fluxes play an essential but often understated role in modulating salinity through changes in volume: salinity decreases when precipitation adds volume, and increases when evaporation reduces volume. This insight further clarifies how the inflow of relatively fresh Atlantic water (AW) and the outflow of saltier Mediterranean water (MOW) can, counterintuitively, lead to an overall reduction in salinity. We found that surface fluxes, primarily driven by evaporation, account for most salinity variability, contributing approximately 1.80±0.10 Sv. The Gibraltar exchange is critical for maintaining mass balance, adding 0.30±0.20 Sv of salt. However, due to density differences between AW, MOW, and Mediterranean water, a net salinity reduction of -1.48±0.20 Sv is achieved through the Strait. This results in an overall increase of 0.29±0.09 Sv in salinity over the 15-year period, consistent with the salinification trend reported in previous studies. These findings provide a more comprehensive perspective of regional water cycle dynamics.
- Preprint
(1715 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-857', Anonymous Referee #1, 08 Apr 2025
This is an interesting paper which use the ECCO4 reanalysis to examine the relative contributions of Gibraltar Straits exchanges and surface fluxes to variability in the mass/salinity budgets of the Mediterranean Sea. It’s relatively well written but could do with some ffurther analysis in a few places as noted below.
line 1281-132. ECCO appears to do well despite it’s relatively coarse resolution. Could the authors comment further here on whether they think the 1 deg ECCO resolution is sufficient for their analysis? Also include some discussion of what aspects of their study are likely to become more accurate if a ¼ or 1/12 deg model were available.
Line 158. The authors choose a depth of 150m to separate the AW and MOW without any justification. So, can they include some further ECCO based results to support this choice? Is there any time dependence in the separation depth?
Line 160-165. The authors chose to work at whole basin level rather than sub-regions. However, they could carry out an intermediate analysis by splitting the whole basin into two sub-basins i.e. E & W Med separated by the Strait of Sicily. Could they include some discussion of whether such an approach is likely to yield further insights to those already presented?
Line 218. The Surface and Strait terms are noted to exhibit near exact, opposite variations. However, the process by which this is achieved is not noted here. So, please discuss. Is this a real balance or an artefact of the ECCO model?
Line 243. ‘this is likely because the inflowing North Atlantic water is becoming saltier over time, which is consistent with some recent findings.’ Please include a time series of the 0-150m mean salinity to show whether this statement is supported by more detailed analysis of ECCO.
Line 246. ‘The air-sea freshwater flux, driven primarily by substantial net evaporation, contributes significantly to this trend.’ A further time series needs to be included showing E and P separately to support this statement.
Sec 3.2. The results on the NAO are interesting but other modes of variability, particularly the EAP and EA/WR patterns are known to influence the Mediterranean. So, the authors need to extend their analysis here to include the EAP and EA/WR in order to provide a complete picture (even if these modes turn out not to have strong correlations with the air-sea mass flux and salinity). Indices for the EAP and EA/WR are available from the same site as employed for the NAO: https://www.cpc.ncep.noaa.gov/data/teledoc/telecontents.shtml.
Fig.6 The E & P salinity numbers in the boxes appear to be incorrect (wrong way round) compared to the values in the Table and main text.
Citation: https://doi.org/10.5194/egusphere-2025-857-RC1 -
AC3: 'Reply on RC1', Chao Liu, 28 Apr 2025
We thank the reviewer for their time and for providing positive feedback and constructive comments on our manuscript. We have carefully revised the manuscript accordingly and incorporated all the proposed changes. A detailed response addressing each of the reviewer’s concerns is provided in the attached PDF file.
-
AC3: 'Reply on RC1', Chao Liu, 28 Apr 2025
-
RC2: 'Comment on egusphere-2025-857', Anonymous Referee #2, 09 Apr 2025
Review of “Salinity trends and mass-balances in the Mediterranean Sea: the role of air-sea freshwater fluxes and oceanic exchange” by Liu et al.
Using output from the ECCOv4r4 model for the Mediterranean Sea over a 15-year period (2003–2017), the authors analyze the region's mass and salinity budgets. They conclude that surface freshwater fluxes dominate salinity variability, while the exchange through the Strait of Gibraltar plays a key role in maintaining the overall mass balance. They report a salinity increase of 0.29 ± 0.09 Sv (units unclear) over the study period.
However, both of these findings—that salinity variability is mainly driven by air-sea fluxes and that the mass balance is maintained through the Gibraltar exchange—are well-established in the literature and do not represent novel results. The potential novelty of the study lies in the use of a global circulation model to revisit these questions. Yet, this approach has a significant limitation: the necessarily low spatial resolution of the ECCO model in the Mediterranean, a basin known for its rich mesoscale activity, which the model cannot resolve. As a result, the authors are constrained to a basin-integrated approach that effectively reduces the Mediterranean to a "box," as acknowledged in line 145. Unfortunately, this limits the originality of the analysis.
In addition to this general concern, there are several major and minor issues (detailed below) that prevent me from recommending this manuscript for publication in its current form.
Section 4 – “Discussion of ECCO resolution and uncertainty” This section is not a true discussion. The authors merely point out that the ECCO model’s resolution is insufficient to capture key oceanographic processes in the Mediterranean Sea and that other numerical models—indeed, they do exist—may be better suited for this purpose. They also mention that further efforts are needed to address this gap. However, the question then arises as why did the authors not attempt to use one of these more appropriate models in their own analysis? Moreover, despite the title, there is no attempt to quantify the uncertainties discussed. As it stands, this section reads as a series of general, largely uncritical remarks that add little value to the manuscript. It could be omitted altogether without any loss.
Section 5 – “Conclusions” The statement that “we reveal the complex dynamics that govern Mediterranean’s mass and salinity variations” is misleading, as the manuscript contains no dynamic analysis whatsoever. Similarly, the claim that the findings contribute to a “broader understanding of sea level dynamics in the Mediterranean” through “regional steric (halosteric) adjustments” is not supported by the content—no such approach is developed or discussed in the paper. Most of the conclusions presented in this section do not reflect the actual analyses or results shown in the manuscript, and should be rewritten to more accurately represent the study's scope and findings.
Salinity flux in Sv. In line 182, the authors state: “Since salinity is a unitless measure, we also express salinity flux in Sv in comparing both types of fluxes.” I must admit I do not understand what 1 Sv of salinity flux means. The authors should provide a clearer explanation of how this unit is defined, as well as how it relates to conventional salinity units.
Line 55. What exactly do the authors mean by "volume fluxes"? Are they referring to variations related to thermosteric effects? Since salinity is expressed in g/kg, it is influenced by mass fluxes—not volume changes unless these are accompanied by changes in mass. It’s unclear how pure volume changes, without any mass input or output, would affect salinity. This needs clarification.
Lines 66–67. The sentence reads: “The water mass exchange at the Strait of Gibraltar was estimated at 1 ± 3 mm/yr…” These units are unusual (even, incorrect) for water mass exchange and add confusion to the manuscript, which also use Sv and km3/year for the same variable.
Figure 3b (and 3e). These panels show the cumulative mass over time derived from the integration of the mass flux (Figure 3a and 3d), which is given in units of Sv. The slope of the cumulative curve corresponds to the time derivative, which essentially takes us back to the flux shown in Figure 3a (or 3d). In fact, the slope is simply the time-averaged value of the flux in Figure 3a (or 3d). Why is this point being made in such a roundabout way? Also, why express the slope in km³/year instead of using Sv, as in Figure 3a? For instance, –1390 km³/year corresponds to -0.044 Sv, a value that can be easily inferred from Figure 3d.
Figure 4. Much of the above comment also applies here. In this figure, it is unclear how the questionable units of Sv for salinity flux (Figures 4a, 4d) translate into actual salinity values in Figures 4b and 4e. Regarding the discussion in lines 230–239: if the total salinity flux (black line) is the sum of the surface freshwater flux (green), which varies considerably month to month, and the relatively steady salinity flux through the Strait (red), then the black and green lines must be highly correlated. This should be not a new and relevant finding of the study.
Line 240. The manuscript refers to a trend of “0.02 ± 0.01 Sv per year.” Does this correspond to the trends shown in Figures 4a and 4d? Please clarify this point explicitly.
Line 242. The authors suggest that the significant increase of salinity found in “the Strait” term is a consequence of an increase of the salinity of the Atlantic inflow through the Strait of Gibraltar. However, other studies point to the opposite trend—namely, freshening of the inflow—driven by melting Arctic ice. This apparent contradiction should be addressed or, at least, mentioned.
Lines 281–284. There is a reference to Figures 2d and 2f, but these sub-panels do not appear in Figure 2. It’s unclear what time series or pattern is being referred to in this sentence. The "cosine-like pattern" is not evident. Please clarify what is meant here, and where this pattern is supposed to appear.
Figure 6. Are the labels in this schematic accurate? It is confusing to see evaporation (E) labeled as a mass gain and precipitation (P) as a mass loss—this seems reversed. Also, a "salinity increase of –2.46" is contradictory in terms—it would imply a decrease. Finally, the numbers in Figure 6 and Table 1 suggest a net salinity increase of 1.80 – 1.48 = 0.32, yet the text states 0.29. This discrepancy should be resolved for consistency.
Citation: https://doi.org/10.5194/egusphere-2025-857-RC2 -
AC4: 'Reply on RC2', Chao Liu, 28 Apr 2025
We thank the reviewer for their time and for providing positive feedback and constructive comments on our manuscript. We have carefully revised the manuscript accordingly and incorporated all the proposed changes. A detailed response addressing each of the reviewer’s concerns is provided in the attached PDF file.
-
AC4: 'Reply on RC2', Chao Liu, 28 Apr 2025
-
EC1: 'Comment on egusphere-2025-857', Sjoerd Groeskamp, 09 Apr 2025
Both reviewers provide several points for the authors to consider in their revised manuscript. In particular, referee #2 suggests to clarify or improve the scientific significance of this study. Both reviewers have questions about the model used for the anlayses and the limitation this causes. This has consequences for the conclusions that are not adequatly discussed. I encourage the authors to consider the referee comments and provide a point by point response to authors comments and revised manuscript to Ocean Sciences.
Citation: https://doi.org/10.5194/egusphere-2025-857-EC1 -
AC5: 'Reply on EC1', Chao Liu, 28 Apr 2025
We sincerely thank you for your careful evaluation of our manuscript and for forwarding the insightful comments from the reviewers. We have thoroughly considered all the points raised, particularly regarding clarifications on the scientific significance of our study and the limitations associated with the model used.
In response, we have substantially revised the manuscript to better address these concerns, including a revised discussion of model limitations and their implications for our conclusions. A detailed, point-by-point response to the reviewers’ comments, along with a revised version of the manuscript, has been submitted for your consideration.
We appreciate the opportunity to improve our work and look forward to your further feedback.
Citation: https://doi.org/10.5194/egusphere-2025-857-AC5
-
AC5: 'Reply on EC1', Chao Liu, 28 Apr 2025
-
CC1: 'Article review 2025-04-20', Jari Miglio, 20 Apr 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-857/egusphere-2025-857-CC1-supplement.pdf
-
AC2: 'Reply on CC1', Chao Liu, 28 Apr 2025
We thank the reviewer for their careful reading and helpful feedback. We have corrected the typographical errors identified—specifically, the incorrect figure reference in Section 3.1 and the swapped evaporation and precipitation labels in Figure 6. We also appreciate the comment regarding the dimensional consistency of the mass and salinity budget equations in Section 2.2. To clarify, the unit for pressure here is the same as sea level, for it is used as the equivalent of water thickness; we have revised this section to clarify the formulation and ensure dimensional consistency. We are grateful for these observations, which have helped improve the accuracy and clarity of the manuscript.
Citation: https://doi.org/10.5194/egusphere-2025-857-AC2
-
AC2: 'Reply on CC1', Chao Liu, 28 Apr 2025
-
CC2: 'Comment on egusphere-2025-857', Campin Jean-Michel, 23 Apr 2025
This is a specific comment from Ou Wang (ou.wang@jpl.nasa.gov) and Jean-Michel Campin (jmc@mit.edu) on paragraph starting on line 166, "It is very important to emphasize ..."
A few statements in this paragraph are not correct. All the ECCO-v4 solutions are using the MITgcm "z*" vertical coordinate (Adcroft and Campin, 2004, doi:10.1016/j.ocemod.2003.09.003)
so that the grid-cell thickness is stretched/squeezed according to sea-surface height (SSH). This contradicts the statement (lines 168-169): "This is because the grid-cell thickness
in MITgcm is held fixed with linear free-surface method".
In addition, all the ECCO-v4 solutions use the "real freshwater flux" formulation meaning that freshwater is NOT converted to a "virtual salt flux" but instead is applied directly as source/sink of freshwater at the surface and the model does take care of salt dilution/concentration by itself.
More details about the MITgcm real freshwater flux formulation in z* coordinate can be found in Campin etal, 2008, (doi:10.1016/j.ocemod.2008.05.005).Citation: https://doi.org/10.5194/egusphere-2025-857-CC2 -
AC1: 'Reply on CC2', Chao Liu, 28 Apr 2025
We sincerely thank Dr. Ou Wang and Dr. Jean-Michel Campin for their detailed clarification regarding the use of the MITgcm "z*" vertical coordinate and the real freshwater flux formulation in ECCOv4 solutions. We have carefully reviewed the references provided and have corrected our statements accordingly in the revised manuscript. We are very grateful for your input, which significantly improves the technical accuracy and clarity of our work.
Citation: https://doi.org/10.5194/egusphere-2025-857-AC1
-
AC1: 'Reply on CC2', Chao Liu, 28 Apr 2025
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 327 | 124 | 21 | 472 | 13 | 19 |
- HTML: 327
- PDF: 124
- XML: 21
- Total: 472
- BibTeX: 13
- EndNote: 19
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1