the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 10 Jun 2026
Influence of Sea Surface Temperature on the Columnar Water Vapor Content and Cloud Fraction, Based on Monthly-Averaged MODIS Data at 1° by 1° Resolution
Abstract. Satellite data on global sea surface temperature (SST), water vapor and cloud fraction are analyzed to provide direct relationships on these parameters. Increase in SST elevates the water vapor pressure at the surface following the Clausius-Clayperon (exponential) form, and this effect persists to increase the columnar water vapor up to SST of approximately 300 K, at the 1° by 1°, monthly-averaged scale. Beyond SST of 300 K, a steeper slope for the columnar water vapor is observed. A similar transitional relationship is observed between cloud fraction (CF) and SST, except that a negative slope is found up to SST of 300 K. Then, a reversal occurs at SST of approximately 300 K where CF increases quadratically as a function of SST. Parameterization of water vapor and CF is provided as a function of SST for 1° by 1° spatial resolution and monthly-averaged time scale.
- Preprint
(1289 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 05 Aug 2026)
-
RC1: 'Comment on egusphere-2026-1234', Anonymous Referee #1, 24 Jun 2026
reply
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2026/egusphere-2026-1234/egusphere-2026-1234-RC1-supplement.pdfReplyCitation: https://doi.org/
10.5194/egusphere-2026-1234-RC1 -
AC2: 'Reply on RC1', Taewoo Lee, 02 Jul 2026
reply
Fig. 5 is the combined plot for the entire 17 years data on SST, water vapor and cloud fraction. Other earlier figures have been included to illustrate the repeated trends, and also any monthly fluctuations. As noted in the manuscript, the plots contain global data points (over the sea surfaces) at 1 degree x 1 degree spatial resolution, monthly averaged. In the revision, these points and combined figures will be added.
Cloud fraction is derived from the MODIS Cloud Product (MOD06/MYD06) by analyzing visible and infrared channels to determine the portion of a grid cell covered by clouds, while water vapor is retrieved via the Atmospheric Profile Product (MOD07/MYD07) using near-infrared and infrared absorption bands. For the accuracy and reliability of MODIS data on water vapor and cloud fraction, we refer to the references for the data set: https://atmosphere-imager.gsfc.nasa.gov/products.
The above point will be clarified in the revision.
Citation: https://doi.org/10.5194/egusphere-2026-1234-AC2
-
AC2: 'Reply on RC1', Taewoo Lee, 02 Jul 2026
reply
-
RC2: 'Comment on egusphere-2026-1234', Anonymous Referee #2, 26 Jun 2026
reply
General comments
In this study, the authors investigate the relationships among sea surface temperature (SST), column water vapor (CWV), and cloud fraction (CF) using monthly averaged 1° × 1° MODIS observations. They find a universal relationship between SST and CWV, as well as between SST and CF. In both relationships, the dependence on SST changes around 300 K.
The motivation for this study is understandable, as it addresses a classic problem in atmospheric science, and it is well known that both CWV and CF are primarily dependent on SST. However, it is also well known that they are not solely determined by SST. For example, Kanamaru and Masunaga (2013) introduced the water vapor scale height and found that the relationship between CWV and SST varies systematically with this parameter. Such studies have shown that the dependence of CWV or CF can be further refined by considering additional key parameters in addition to SST.
It is not clear why the authors chose to use the MODIS dataset. CWV is not directly retrieved from MODIS observations, and SST retrieval from MODIS is inherently limited in cloudy regions. Several more reliable and widely used datasets are available for both CWV and SST.
For the above reasons, I find that the present analysis is too simplistic to provide a comprehensive understanding of how CWV or CF is determined. A more thorough investigation should incorporate multiple observational datasets and consider additional controlling factors beyond SST.
Therefore, I do not recommend publication of the manuscript in its present form. The authors should substantially reconsider the approach; employing more appropriate datasets and more comprehensive analyses would be required
Specific comments
L47-48, “The atmospheric optical depth and cloud fraction are close correlated (Engstrom and Ekman, 2010; Gryspeerdt et. al, 2014).”: These literatures focus on the aerosol optical depth. It is not relevant to the present study.
L60-62, “Waliser (1996) investigated the SST-atmosphere interaction by examining thermodynamic and convection dynamic variables, suggesting that there is a feedback mechanism to suppress locally high SST.: “Waliser (1990)” is not listed in the reference. The feedback mechanism is not relevant to the present context.
L66, “vertical gradients”: It is unclear what quantity the authors are referring to. Please specify the variable whose vertical gradient is being discussed.
L79-80, “However, detailed and accurate simulations of entire global circulation are intrinsically difficult and time-consuming”: This statement is no longer appropriate, as detailed and accurate representations of the global atmospheric circulation are now readily available from high-resolution reanalysis datasets, such as ERA5.
L106-108, “Detailed descriptions of the data and processing methods are discussed by various authors of the MODIS data set (https://modis.gsfc.nasa.gov/data/).”: The authors should review scientific papers on the MODIS data set and find their limitations and the rationale for using it in the present analysis.
L120, Figure 1: Which domain on the Earth is analyzed here? The CWV-SST relationship should be analyzed separately in the tropics and extratropics.
L121, “a Clausius-Claperyon (CC) form of exponential dependence on SST is observed up to SST of approximately 300 K.”: To discuss deviations from the Clausius–Clapeyron (CC) relationship, the authors should explicitly compare the observed relationship with that predicted by the CC relationship. For example, they could produce a scatter plot of the theoretical CC scaling against the observational results to quantitatively assess deviations.
L131-133, “During some months, the data are bifurcated for SST < 285 K, with upper branch exhibiting higher WV than the lower CC branch. These months are also associated with larger scatter in the data, in Figure 1.”: The authors should demonstrate this evidence.
L143-145, “Data plotted in Figure 1 indicate that the water vapor content is “over-saturated”, higher than the water vapor pressure at the corresponding SST, SST of of approximately 300 K.”: It is unclear from Figure 1 how the authors reached the conclusions described in the text. The authors should provide additional analyses and supporting figures that directly demonstrate these findings.
L160-163, “The physical mechanism for this reversal and increase in CF above the threshold temperature (Tc) may again be attributed to the increased upward convection at high SST, distributing the water vapor to higher altitude resulting in more vapor available for condensation and cloud formation.”: The reversal in the CF relationship is one of the most interesting findings of the present study and deserves a more in-depth investigation. The authors should explore the underlying physical mechanism by analyzing additional meteorological fields (e.g., reanalysis data), such as convection, large-scale circulation, and humidity.
-
AC3: 'Reply on RC2', Taewoo Lee, 02 Jul 2026
reply
Cloud fraction is derived from the MODIS Cloud Product (MOD06/MYD06) by analyzing visible and infrared channels to determine the portion of a grid cell covered by clouds, while water vapor is retrieved via the Atmospheric Profile Product (MOD07/MYD07) using near-infrared and infrared absorption bands. For the accuracy and reliability of MODIS data on water vapor and cloud fraction, we refer to the references for the data set: https://atmosphere-imager.gsfc.nasa.gov/products. It would be an interesting exercise to compare with data sets, but in a follow-up study.
Current work is a baseline observational model that demonstrates the direct SST to water vapor and SST to cloud fraction, if plotted with 1 degree x 1 degree spatial resolution and monthly averaging. Simplistic or not, one cannot dispute the clear trends shown in the figures. Secondary effects, such as water vapor height variations and spatial location, may add to the scatter in the data. However, the observed trends are clear in the figures. Therefore, one can have opinions as to how a model should be constructed; but even then it should be based on "observational truth", as presented in this work. Considerations of secondary effects and refinements, are subjects for a follow-up study.
We agree with the minor editorial comments, and will be fully adopted in the revision.
Citation: https://doi.org/10.5194/egusphere-2026-1234-AC3
-
AC3: 'Reply on RC2', Taewoo Lee, 02 Jul 2026
reply
-
CEC1: 'Comment on egusphere-2026-1234 - No compliance with the policy of the journal', Juan Antonio Añel, 26 Jun 2026
reply
Dear authors,
Unfortunately, after checking your manuscript, it has come to our attention that it does not comply with our "Code and Data Policy".
https://www.geoscientific-model-development.net/policies/code_and_data_policy.html
In the link that you provide in the "Code and Data Availability" section of your manuscript you do not provide the data, but a code. Also, it is not clear if you provide the code used in your work. Given this, your manuscript should have not been accepted for Discussions in the journal.
The GMD review and publication process depends on reviewers and community commentators being able to access, during the discussion phase, the code and data on which a manuscript depends, and on ensuring the provenance of replicability of the published papers for years after their publication. Please, therefore, publish your code and data in one of the appropriate repositories and reply to this comment with the relevant information (link and a permanent identifier for it (e.g. DOI)) as soon as possible. We cannot have manuscripts under discussion that do not comply with our policy.
Later, if the Topical Editor decides to continue with the review or publication process of your manuscript and you are requested to upload a new version of it, then The 'Code and Data Availability’ section of your manuscript must also be modified to cite the new repository locations, and corresponding references added to the bibliography.
Additionally, your manuscript has been submitted to the journal as a "Model Description Paper". However, your work does not seem to describe a model according to the Manuscript Types of the journal. Please, check them here:
https://www.geoscientific-model-development.net/about/manuscript_types.html#item1
Also, it does not seem to fit any of the categories for manuscript types that we have and accept in the journal. If you disagree, please, reply to this comment with the reasoning for it.
I must note that if you do not fix this problem, we cannot continue with the peer-review process or accept your manuscript for publication in GMD.
Juan A. Añel
Geosci. Model Dev. Executive EditorCitation: https://doi.org/10.5194/egusphere-2026-1234-CEC1 -
AC1: 'Reply on CEC1', Taewoo Lee, 02 Jul 2026
reply
The data statement has been cleared by one of the managing editors, plus as described (and originating website noted) in the manuscript MODIS data sets have been used. We can include that website again as a reference and in the data statement.
This work includes a model of cloud fraction and columnar water vapor, as a function of SST, and therefore that category was selected. In addition, research forums should be open to new approaches that may not fit classical categorization, by having an "Other" or similar category. Categorization is typically used to find appropriate reviewers, but not to stifle new ideas.
I see that you are an executive editor; however, I must state that a late intervention at this point is discouraging and unprofessional, as the reviewers have already completed their reviews.
Citation: https://doi.org/10.5194/egusphere-2026-1234-AC1 -
CEC2: 'Reply on AC1', Juan Antonio Añel, 02 Jul 2026
reply
Dear authors,
Your reply does not solve the issues I pointed out in my previous comment. You have to reply to this comment with a full Code and Data Availability statement that provides repositories for all the code and data necessary to perform your study.
We understand that you could have preferred to address these issues before having the manuscript under review, but when it is not possible, this becomes part of the review process itself. As stated in my previous comment, if you do not address this situation we can not consider your manuscript for additional review or publication in the journal. Please, address the outstanding issues as soon as possible.
Juan A. Añel
Geosci. Model Dev. Executive Editor
Citation: https://doi.org/10.5194/egusphere-2026-1234-CEC2
-
CEC2: 'Reply on AC1', Juan Antonio Añel, 02 Jul 2026
reply
-
AC1: 'Reply on CEC1', Taewoo Lee, 02 Jul 2026
reply
-
RC3: 'Comment on egusphere-2026-1234', Anonymous Referee #3, 26 Jun 2026
reply
General Comments
The manuscript presents an empirical analysis and parameterization of the relationships between Sea Surface Temperature (SST), column Water Vapor (WV), and Cloud Fraction (CF) based on monthly averaged MODIS satellite data with a spatial resolution of 1°×1°. The authors identify a transitional bifurcation point at an SST of ~300 K, proposing a concept of two statistically formalized regimes (exponential and quadratic dependencies). Such a conceptual division represents an interesting idea that could potentially be useful for simplified climate models.
The manuscript is well-structured, easy to read, and follows a logical narrative. The authors provide a good literature review, relying on classic works and clearly distinguishing the current analysis from their previous 2022 research.
However, despite the high-quality technical execution and good data traceability, the manuscript has shortcomings and issues related to its validity and scope of application.
First of all, the paper is purely a visual and statistical analysis. The authors merely fit curves to the existing MODIS data. The manuscript does not address the core issues covered by the journal's scope and falls outside its thematic focus.
Secondly, the physical interpretation and the final conclusions drawn from the data are overly categorical and are not fully supported by the massive scatter present in the observational data.
Specific Comments
1. The study does not address relevant scientific modeling issues within the journal's scope, offering instead a mere statistical approximation of a function using satellite data as arguments. The authors should consider submitting this work to a journal focused on observations or Earth remote sensing.
2. The visual scatter of data in the plots is enormous. Although the authors indicate a variance of 5-25%, the data visually resemble noise from which it is difficult to extract a reliable physical signal, although a general trend is discernible. To better substantiate the findings of the study, the authors are recommended to consider other datasets of the considered parameters (e.g., datasets from other satellite systems such as AVHRR, or reanalyses), which would partially help separate data errors from the actual physical relationship between the parameters. It is also recommended to support their conclusions with statistical criteria; the R-squared value, for instance, would help strengthen the drawn conclusions.
3. Overly categorical conclusions
3.1 Due to such a high level of uncertainty, the conclusion formulated in the final remarks — "SST can be used as the main determining factor for the resulting atmospheric variables (WV and CF)" — is too strong and poorly substantiated. Cloud fraction and water vapor content are highly complex variables influenced by numerous factors: aerosols, atmospheric stability, etc.; they cannot be determined exclusively by SST, although they can be largely driven by it.
3.2 The authors should revise the statement "exact inversion" and clarify that this is merely a visual expert comparison of normalized trends, as the authors do not provide numerical statistical criteria to support this claim.
3.3 The authors also explain the sharp increase in CF above 300 K by "supersaturation" and enhanced ascending convection. While this is physically plausible, the monthly averaging at a 1°×1° resolution used in this study smooths out localized short-term convective events. The authors should elaborate on this point in the discussion, clarifying that monthly large-scale averaging may obscure the actual dynamic convective processes (which are mentioned in the manuscript text) that drive cloud formation at such high SSTs.
The conclusions must be significantly toned down given this level of analysis.
Technical Corrections
Terminology/Spelling: The name of the thermodynamic equation is misspelled several times throughout the text (e.g., lines 23, 51, 121, 249). It should be correctly written as "Clausius-Clapeyron", not "Clausius-Clayperon" or "Clasius-Claperyon".
Line 274-275: Typo in the abbreviation: "...resulting atmospheric variables (WF and CF)..." should be "WV and CF".
Citation: https://doi.org/10.5194/egusphere-2026-1234-RC3 -
AC4: 'Reply on RC3', Taewoo Lee, 02 Jul 2026
reply
Reviewer's comments: First of all, the paper is purely a visual and statistical analysis. The authors merely fit curves to the existing MODIS data. The manuscript does not address the core issues covered by the journal's scope and falls outside its thematic focus. Secondly, the physical interpretation and the final conclusions drawn from the data are overly categorical and are not fully supported by the massive scatter present in the observational data.
A parameterization of the cloud fraction is an important topic, and current discovery and presentation should receive some considerations. We will add more details on the thermodynamics behind the observed relationships in the revision. There is scatted in the data; however, the overall trends in the mean are quite clear. In a follow-up work (if any), we will consider secondary effects that are causing the scatter, for possible corrections.
We agree with the minor comments, which will be adopted in the revision.
Citation: https://doi.org/10.5194/egusphere-2026-1234-AC4
-
AC4: 'Reply on RC3', Taewoo Lee, 02 Jul 2026
reply
-
RC4: 'Comment on egusphere-2026-1234', Anonymous Referee #4, 03 Jul 2026
reply
This manuscript examines the relationships between sea surface temperature (SST), column water vapor (WV), and cloud fraction (CF) using monthly averaged, 1° by 1° MODIS data over 2000 to 2017, and proposes piecewise empirical fits for WV and CF as functions of SST with a transition near 300 K. The paper is clearly written and easy to follow, and the general behavior it describes (a Clausius-Clapeyron-like increase of WV with SST, and a cloud fraction minimum near warm-pool temperatures) is consistent with what is broadly known.
That said, I share the concerns already raised by the other reviewers regarding novelty and scope. The study is essentially a statistical fit to satellite observations rather than a modelling development, and in its present form I do not think it falls within the scope of this journal. The relationships shown are also well established, and it is not always clear what this analysis adds beyond the authors' earlier 2022 study on the same topic.
Beyond scope, I have a more specific and, in my view, more serious concern: the proposed fitting equations, which are the main deliverable of the paper, do not appear to be correct as printed (see specific comment 1). I would ask the authors to check these carefully, because in their current form they cannot be used. I also agree with the other reviewers that the analysis needs a proper statistical treatment (density plots over the full record, goodness-of-fit measures) and that the confounding between SST and latitude should be addressed directly, since the apparent 300 K transition may reflect a change of regime between the extratropics and the tropics rather than a threshold in SST itself.
Please see the specific comments below:
1. I tried to reproduce the fits of Eqs. (1) and (2) and ran into a problem. Evaluating Eq. (1a) as written gives water vapor values of roughly 0.001 to 0.004 cm between 285 and 300 K, which is about three orders of magnitude smaller than any physical column water vapor amount (order 1 to 6 cm), and it is inconsistent with the "+25/-9%" variance the authors themselves quote for this equation at 285 K. The value that makes Eq. (1a) both physically reasonable and consistent with Eq. (2a) is an offset of 561.15 rather than 661.15, so I suspect this might be a typo error. But I saked the authors to double check and confirm this!
2. Still for the Eqs. (1) and (2), even setting aside the point above, the two branches do not appear to join at the transition temperature. For cloud fraction, Eq. (2a) gives about 0.60 at 300 K while Eq. (2b) gives about 0.34, a jump of roughly 0.26 in CF at the very point the paper emphasizes. A parameterization intended for use elsewhere should be continuous at the breakpoint, or the transition should be represented explicitly. Please constrain the fits accordingly.
3. For the 300k threshold at line 122 and line 157, I agree with the other reviewers that this threshold needs justification. It is worth noting that an SST threshold near 300 K (about 27 °C) is well documented in the tropical convection literature, so the value is not unreasonable, but in this manuscript it is selected visually. I would encourage the authors to estimate the breakpoint objectively (for example with a segmented regression that returns Tc and a confidence interval), to do so separately by latitude band, and to connect the result to the existing convective-threshold studies rather than presenting it as a new finding.
4. I found this passage difficult to follow for the "over-saturation" at lines 143 to 145. As written in the manuscript, the authors compares the column water vapor (a depth, in cm) to "the water vapor pressure at the corresponding SST" (a pressure), which are not the same quantity, and it uses the term "over-saturated," which normally refers to relative humidity above 100%. The underlying observation, that the data lie above the Clausius-Clapeyron extrapolation at high SST, may well be valid, but the wording should be made precise and the comparison put on a consistent footing (for example, maybe in terms of column relative humidity), with a supporting figure.
5. In section 2, line 109 versus Figure 1 (lines 182 to 184) and Figure 4 (lines 224 to 226), the descriptions states that the data span January 2000 to March 2017, but the captions of Figures 1 and 4 include "December 2019," which lies outside that period. Please clarify the true temporal coverage of the dataset, ensure Section 2, the figures, and the Zenodo deposit are mutually consistent, and confirm that all figures were produced from the same record that is described in the text.
6. In conclusion section, the statement that SST can serve as the primary determinant of WV and CF is stronger than the analysis supports, given the large scatter and the many other factors involved. I would soften this, and similarly qualify the "exact inversion" description of the WV and CF traces, which at present rests on a visual comparison rather than a quantitative one.
7. The subpanels of Figures 1 and 4 should be labelled (a), (b), (c), and so on, and the figures would benefit from showing the fitted curves and their statistics.
Citation: https://doi.org/10.5194/egusphere-2026-1234-RC4
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 104 | 20 | 14 | 138 | 13 | 9 |
- HTML: 104
- PDF: 20
- XML: 14
- Total: 138
- BibTeX: 13
- EndNote: 9
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1