A High-Resolution Air-Sea Synoptic Observation Dataset from Drifting Buoys in the Bay of Bengal

Huang, Wei; Wang, Guihua; Chen, Changlin; Chen, Gengxin; Wang, Weiqiang

doi:10.5194/essd-2026-267

Preprints

https://doi.org/10.5194/essd-2026-267

Preprints

21 Apr 2026

| 21 Apr 2026

Status: this preprint is currently under review for the journal ESSD.

A High-Resolution Air-Sea Synoptic Observation Dataset from Drifting Buoys in the Bay of Bengal

Wei Huang, Guihua Wang, Changlin Chen, Gengxin Chen, and Weiqiang Wang

Abstract. Mass and heat exchanges at the air-sea interface fundamentally drive global weather and climate systems. However, acquiring long-term, high-frequency, synchronous in-situ observations of both atmosphere and oceanic variables remains highly challenging, especially during extreme weather. This paper presents a high-resolution dataset from five air-sea drifting buoys deployed in the Bay of Bengal (BoB) in 2020 and 2022. These buoys captured precise, synchronous measurements of key meteorological parameters (air temperature, sea-level pressure, wind speed and direction, and relative humidity) alongside sea surface temperature. The dataset is typically sampled hourly; however, the sampling was increased to 5-minute intervals during tropical cyclones Nivar, Burevi, Four and Asani. This high-frequency dataset offers invaluable in-situ records for studying diurnal variations and fine-scale processes in the BoB. Furthermore, it provides critical observational data to advance our understanding of air-sea coupling, validate high-frequency satellite products, and improve parameterizations in regional numerical weather prediction models under extreme conditions.

Received: 10 Apr 2026 – Discussion started: 21 Apr 2026

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Wei Huang, Guihua Wang, Changlin Chen, Gengxin Chen, and Weiqiang Wang

Status: final response (author comments only)

RC1:
'Comment on essd-2026-267', Anonymous Referee #1, 23 Apr 2026
The study "A High-Resolution Air-Sea Synoptic Observation Dataset from Drifting Buoys in the Bay of Bengal" by Wei Huang et al. presents a dataset collected from five drifting buoys deployed in the Bay of Bengal between 2020 and 2022. Such a dataset is rare and of great importance for studying high-frequency events in ocean–atmosphere heat exchanges, particularly during short and spontaneous events such as storms. The paper is concise and well written, and fits within the aims and scope of the journal. I recommend it for publication after major revisions.
General comments
The quality of the figures (i.e., their resolution) is somewhat low; it would be appreciated if it could be improved.

The choice of colors in the figures sometimes makes them difficult to read for colorblind readers. I recommend using adapted color palettes (see for example https://jfly.uni-koeln.de/color/).

The journal requires compliance with the International System of Units (SI). This standard requires a non-breaking space between the value and its unit, except when the unit is in degrees (e.g., 5°, but note that for compound units like 5 °C, the space is required). Also, the use of the "/" notation should be avoided; instead, use a space or a dot (e.g., m s⁻¹ or m·s⁻¹ instead of m/s).

Introduction

Line 27: What does "SVP" stand for?

Line 31: "to fully capture meso- and submesoscale air-sea processes over large regions" or something similar would be more appropriate. Indeed, Eulerian moorings can capture these phenomena, but only at fixed locations.

Line 38: "suppresses the vertical mixing": does it completely suppress it? Or would "reduces" be more appropriate?

Line 50: "sampling frequency increased to 5–30 minute intervals": how are these periods chosen, i.e., the periods during which the temporal resolution is increased? Are they selected arbitrarily? This may not be a major issue, but it should be clarified.

Line 51: "ultra-high-resolution" may be too strong in this context. While the dataset is high-resolution in oceanographic standards, this is not necessarily the case compared to other fields. I suggest removing "ultra".

Line 52: "irreplaceable" is also somewhat too strong.

Drifting buoys

Line 65: To obtain purely Lagrangian observations of atmospheric wind velocity, the buoy would need to drift with the atmospheric winds rather than ocean currents. I suggest adding something like "(from the ocean perspective)" after "Lagrangian".

Line 66: "standard height" is too vague. A value or at least an order of magnitude should be provided (e.g., in parentheses).

Line 67: Same comment as above: "below the waterline" should be complemented with an approximate depth.

Lines 70–71: Same comment as before. How are these sampling intervals selected? Arbitrarily? Based on a threshold? This should be clarified at least once in the manuscript.

Table 1: To simplify the table, "RH" could be removed from the relative humidity unit.

Line 81: "two of these buoys drifted in close parallel for an extended period": it would be helpful to specify which ones for easier interpretation of the figure.

Line 81: Are these data really "unique"? Or rather "rare" or "valuable"?

Line 83: "up to 150 days": it is useful to provide the upper bound, but the lower bound should also be given.

Figure 2: "Elevation" in the colorbar title should be replaced by "Bathymetry".

Table 2: Is the "Synoptic Observations" column necessary? It does not seem to vary between buoys and is redundant with Table 1.

Table 2: Why specify "Tropical Cyclone" for Asani and not for the others? This does not seem necessary. Also, why are some names in grey?

Data Processing Method

Lines 101–102: This is one of the most important points for me. I do not think replacing outliers with NaN is a good approach. These data should be preserved in the dataset, and a "QC flag" column should be provided (e.g., with codes for values outside instrument ranges, climatological bounds, or exceeding the 3σ criterion). This last criterion is arbitrary. This is not necessarily a problem, but what may appear as noise for some studies could be informative for others. For instance, wind exhibits turbulent dynamics, and SST behaves as a passive scalar of turbulence. In this context, intermittency (i.e., the presence of extremes across scales, especially small ones) is expected (e.g., Frisch, 1995; Schmitt & Huang, 2016; Robache et al., 2025). Removing spikes may therefore disrupt this intermittent structure without awareness. For this reason, I suggest flagging rather than removing them. These data do not necessarily need to be included in the figures, but they should be preserved in the dataset.

Line 110: An order of magnitude for the underestimation would be useful (e.g., "underestimates by up to XX %").

Figure 3: "Buoy measurements were adjusted to a 10 m reference height": please explain how.

Figure 3: "u- and v-component wind speeds" → "u- and v- wind speed components"?

Lines 118–125: These data are not used or shown later in the manuscript. Why?

Lines 135–138: For each term in the equation, the units should be specified. In addition, it should be clearly stated which variables are directly computed from the observations (with the corresponding formulas) and which are prescribed constants or parameterized quantities (with their values and references).

Line 139: "2.0 m" → "2 m".

Line 140: "empirical profile formulas" is vague: which ones?

Data overview

Line 144: Are there references or data supporting this cyclonic circulation?

Line 147: "mesoscale eddies": are you sure that their displacement is only due to mesoscale processes? Some studies have shown that these scales are not always sufficient to explain Lagrangian buoy divergence (e.g., Ma et al., 2024).

Figure 4: The choice of colors is problematic for colorblind readers (especially green and red).

Figure 4: Are storm names given in chronological order from left to right? If so, this should be specified.

Figure 4: Some curves appear to extend beyond the panel boundaries (e.g., panel c on the far right). Is this normal?

Lines 159–160: The wording is somewhat repetitive. It could be changed, for instance: "Time series of … are shown in Figure 4."

Line 161: "clear diurnal cycle": was a method used to estimate the frequency of these peaks?

Line 162: "high-frequency amplitude of SST": what is meant by "high-frequency"? Also, is the amplitude really smaller, or is it a visual effect due to axis scaling? Orders of magnitude should be provided.

Lines 163–164: "Wind speeds typically ranged from 0 to 10 m/s…": this is not clearly supported by the figures. Values above 10 m s⁻¹ appear outside TC periods. How was this threshold chosen? Were TC and non-TC periods analyzed separately?

Line 166: Figure 4b should not be cited here, since Asani only concerns the right panels. Also, other relevant panels are not referenced (for SST, air temperature, wind speed…).

Line 167: Figure 5c and 5d (right)? Still referring to Asani? If so, avoid breaking the paragraph.

Line 168: "far exceeding normal baseline levels": provide an order of magnitude (e.g., mean value).

Lines 171–175: I recommend moving this paragraph ("In summary…") to the conclusion section.

Data availability

According to journal standards, this section should appear later. I also recommend including the dataset DOI in the abstract.

It would also be useful to provide the source of the bathymetry data used in Figure 2 if possible.

Discussion and conclusion

It may be useful to cite studies demonstrating the importance of high-frequency data for estimating ocean–atmosphere heat flux budgets.

Lines 203–205: I do not understand what the issue is here, or what the authors are trying to convey. Indeed, a Lagrangian approach does not allow for Eulerian time-series analysis. However, does this type of approach not have its own relevance (e.g., this is how passive particles such as phytoplankton would experience heat fluxes)?

Line 206: What is meant by "noise"? It would also be helpful to explain how these processes affect measurements.

Appendix A

For start and end times, do the last numbers correspond to hours? If so, "h" should be added.

Is "Niviar" a typo (instead of "Nivar")?

References
Frisch, U. (1995). Turbulence: The Legacy of A. N. Kolmogorov. Cambridge University Press, ISBN 978-0-521-45713-2.
Ma, Y., Huang, Y., & Hu, J. (2024). Spatiotemporal similarity of relative dispersion in the Gulf of Mexico. Frontiers in Marine Science, 11, 1446297. https://doi.org/10.3389/fmars.2024.1446297
Robache, K., Schmitt, F. G., & Huang, Y. (2025). Scaling and intermittent properties of oceanic and atmospheric pCO₂ time series and their difference in a turbulence framework. Nonlinear Processes in Geophysics, 32(1), 35-49. https://doi.org/10.5194/npg-32-35-2025
Schmitt, F. G. and Huang, Y. (2016). Stochastic Analysis of Scaling Time Series: From Turbulence Theory to Applications. Cambridge University Press, https://doi.org/10.1017/CBO9781107705548.
Citation: https://doi.org/10.5194/essd-2026-267-RC1
RC2: 'Comment on essd-2026-267', Anonymous Referee #2, 10 Jun 2026

Comments on the paper by Huang et al.,
A major source of uncertainty in modern forecasting models stems from an incomplete understanding of air-sea interactions. Consequently, high-frequency observations of relevant physical variables are crucial for capturing and understanding these multiscale, multiphysical processes. In this data description paper, Huang et al. present a valuable dataset compiled from high-frequency Lagrangian drifter measurements. This work represents a genuine and practical contribution to the scientific community. However, there are several areas where the manuscript and data presentation should be improved. My specific comments are outlined below.
1 Line 12: Change “atmosphere” to “atmospheric”.
2 Line 50 (“TCs (Nivar, Burevi, Four and Asani)”): While the timeline of these Tropical Cyclones (TCs) is detailed in Appendix A, I suggest providing the date/year information upon their first mention in the main text for immediate context (e.g., Nivar, Nov. 2020).
3 Figure 1: Please increase the resolution of panel (a) to at least 300 dpi to improve visual quality. In panel (a), please include key structural dimensions directly in the schematic (e.g., the distance between the atmospheric sensor array and the main controller). Additionally, it would be beneficial to add clear labels/annotations identifying each component in both panels (a) and (b).
4 Lines 65–67: It is unfortunate that wave parameters are omitted. If available, I highly encourage the authors to include and share wave information (e.g., significant wave height) with the community.
5 Table 1: Change the column header “Resolution” to “Storage precision” if that more accurately reflects the values shown.
6 Lines 101–102: Please provide the approximate percentage, ratio, or total count of data points identified as outliers during quality control.
7 Lines 122–124: Please assign numbers to all equations and remove the bold formatting from the variables. Ensure that the notation for time bounds ($t_0$ and $t_1$) is consistent throughout. Additionally, please clarify whether the time interval ($\Delta t = t_1 - t_0$) matches the sampling interval of the raw data.
8 Lines 133–136: Please assign numbers to these equations and remove the bold formatting from the variables. Furthermore, please provide a precise physical definition of "near-surface" by stating the exact measurement height above the sea surface.
9 Line 140: Please provide the explicit mathematical form of the empirical profile formulas used to convert wind speed from $U_2$ to $U_{10}$.
10 Figures 3, 4, and 5: The image quality is currently too low. Please provide high-resolution figures (minimum 300 dpi) for the final version.
11 I strongly encourage the authors to make the raw (unprocessed) dataset publicly available alongside the quality-controlled version. This allows other researchers to apply alternative quality control procedures tailored to different applications.

Citation: https://doi.org/10.5194/essd-2026-267-RC2

Wei Huang, Guihua Wang, Changlin Chen, Gengxin Chen, and Weiqiang Wang

Data sets

A High-resolution Air-Sea Synoptic Observation Dataset from Drifting Buoys in the Bay of Bengal Wei Huang et al. https://doi.org/10.5281/zenodo.19469106

Wei Huang, Guihua Wang, Changlin Chen, Gengxin Chen, and Weiqiang Wang

Viewed

Total article views: 395 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
320	58	17	395	16	14

HTML: 320
PDF: 58
XML: 17
Total: 395
BibTeX: 16
EndNote: 14

Views and downloads (calculated since 21 Apr 2026)

Month	HTML	PDF	XML	Total
Apr 2026	115	33	11	159
May 2026	161	17	3	181
Jun 2026	44	8	3	55

Cumulative views and downloads (calculated since 21 Apr 2026)

Month	HTML	PDF	XML	Total
Apr 2026	115	33	11	159
May 2026	161	17	3	181
Jun 2026	44	8	3	55

Viewed (geographical distribution)

Total article views: 395 (including HTML, PDF, and XML) Thereof 395 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 20 Jun 2026

Short summary

Understanding air-sea interactions is essential for global climate and weather systems. However, high-frequency, synchronous observations are extremely rare. We present a unique dataset from five drifting buoys in the Bay of Bengal (2020–2022) capturing both atmospheric and oceanic variables. Sampled at rates up to every 5 minutes during four tropical cyclones, this data provides crucial records to study air-sea coupling, validate satellites, and improve models under extreme conditions.


Total:	0
HTML:	0
PDF:	0
XML:	0