Arctic Sea Ice Loss Amplifies Local Evaporation Influence on Water Vapor Isotopes: Insights from Cruise Observations

Zhang, Yuankun; Liu, Zhongfang; Li, Dongsheng; Li, Zhiqing; Shao, Hebin

doi:10.5194/egusphere-2025-5306

Preprints

https://doi.org/10.5194/egusphere-2025-5306

Preprints

03 Dec 2025

| 03 Dec 2025

Status: this preprint is open for discussion and under review for Atmospheric Chemistry and Physics (ACP).

Arctic Sea Ice Loss Amplifies Local Evaporation Influence on Water Vapor Isotopes: Insights from Cruise Observations

Yuankun Zhang, Zhongfang Liu, Dongsheng Li, Zhiqing Li, and Hebin Shao

Abstract. Rapid Arctic warming and sea ice retreat have increased atmospheric humidity, yet the relative contributions of local evaporation and advected lower-latitude moisture remain poorly quantified. Here, we present high-resolution, ship-based in-situ measurements of near-surface water vapor isotopes across diverse Arctic sea ice regimes. By integrating isotope fractionation models with multi-source meteorological data, we show that sea ice changes act as a key modulator of Arctic water vapor isotopic variations. Under ice-covered conditions, water vapor isotopes are controlled by Rayleigh distillation, producing depleted δ¹⁸O with a strong temperature dependence and elevated d-excess from ice-phase processes. As sea ice retreats, kinetic fractionation from local evaporation becomes increasingly important, particularly at temperatures above ~ 5  °C, generating enriched δ¹⁸O, elevated d-excess, and a characteristic "anti-temperature" effect. A Bayesian isotope mixing model quantifies the resulting moisture source shift, showing local evaporation contributions rise from 9.3 % in ice-covered regions to 22.7 % in melt regions, despite advected moisture remaining predominant. These findings establish a process-based isotope framework for the Arctic hydrological cycle, complementing conventional meteorological diagnostics and offering a robust benchmark for interpreting paleo-isotope archives.

Received: 28 Oct 2025 – Discussion started: 03 Dec 2025

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

Yuankun Zhang, Zhongfang Liu, Dongsheng Li, Zhiqing Li, and Hebin Shao

Status: open (extended)

Post a comment Subscribe to comment alert

RC1:
'Comment on egusphere-2025-5306', Anonymous Referee #1, 16 Jan 2026 reply
Review of “Arctic sea ice loss amplifies local evaporation influence on water vapor isotopes: Insights from cruise observations” by Zhang et al., submitted to EGUsphere

The study by Zhang et al 2025 reports ship-based measurements of water stable vapour isotopes from an extensive cruise on the Chinese research vessel Xuelong 2, spanning across the arctic seas, from northern Norway to the Bering Strait. Combined with meteorological data, the authors use a Bayesian mixing model and Lagrangian trajectory analysis to quantify contributions of local (from the Arctic ice-free region) and remote (from lower latitudes) moisture sources in this area. The main conclusion of this manuscript is that sea ice change is a key modulator or Arctic water vapour isotopic variations. The presented dataset is a valuable and forms an important set of observations in an in general under sampled region. However, there are some aspects of this paper concerning the methods, presentation, and interpretation of results that should be addressed before publication in ACP.

Major comments

The use of the MixSIAR model to determine the contributions of local versus remote vapour appears questionable for the problem at hand. There are other suitable tools available for this application that can calculate the moisture budget directly rather than using a statistical approach. This is particularly important since the authors claim to provide a ‘physical consistent basis for the interpretation of Arctic vapor isotopes’ (line 338). In addition, in this manuscript it is not clear how precisely MixSIAR is used. An important concern is that this model might not be the correct tool to use when there is effectively only one tracer available, as δ¹⁸O and δD are highly correlated. Furthermore, the sensitivity of the results on the isotopic composition of the assumed North Atlantic end member should be discussed.

The authors use hourly, single, 5-day air parcel trajectories at 10 m arrival height to determine moisture sources and, in turn, use this as a basis to compute a weighted isotopic source mean. There are several problems with this choice. (i) The atmospheric moisture residence time can be substantially longer and vary with different weather situations at high latitudes (Gimeno et al., 2021). (ii) Trajectory end points, while indicating air parcel origins, do not necessarily reflect the actual moisture source. For this, one would need to take into account the moisture budget along a trajectory as in commonly applied trajectory-based diagnostic (Brunello et al., 2024). If the authors keep their choice of method, such substantial uncertainties need to be considered in the interpretation of the results, and require a more careful phrasing on how reliable the contributions from different regions found here are (see comment #3).

Broad statements and general conclusions are at several places phrased such that they suggest being a result of this study. Wordings and conclusions throughout this manuscript appear often overstated. For example, lines 16-17: “These findings establish a reliable ...”. This overlooks previous work by others in this region, and overstates the usefulness of the results as a benchmark without taking into account the uncertainty of the analysis. Similar criticism applies to the statement in lines 348-349 and at several other locations. Please rephrase and modify these statements by considering what actually is new and added by this study. It is the reviewer’s impression that this mainly is another ship-based dataset from the high Arctic with sea ice and vapour isotope measurements with corresponding interpretation of the respective local and remote influences on the measured signal. More examples for the use of language that appears to overstate findings or appears to be phrased unnecessarily authoritative are pointed out in the detailed comments.

The authors state that there has been a discrepancy in the literature which they resolve through their study, namely that Klein and Welker (2016) report and anti-correlation between sea ice extent and d-excess, whereas Bonne et al (2019) report a correlation between d-excess and sea ice extent (L. 47). In L. 328 onward, the authors claim to have resolved this discrepancy, but other than the claim it is not evident from the text how their study actually reconciled these opposite findings. There is also a citation of Samuels-Crow et al., 2014 in there, which is a study from the Chilean Altiplano at 5000 m a.s.l.. Clearly, the surface conditions there are not comparable to the Arctic. Also, some of the cited studies discuss precipitation d-excess, whereas others discuss vapour, which cannot be lumped together in this argumentation. Currently, the claim of reconciling conflicting interpretations does not appear to be laid out convincingly.

Part of the analysis of the d-excess uses correlations with relative humidity (RH) calculated in relation to air temperature rather than the more relevant quantity RH calculated with respect to SST (Uemura et al., 2008; Pfahl and Sodemann, 2014). The figure panels a-c in Fig. 5 should be updated accordingly, and preferably plotted with consistent ranges on both axes to facilitate comparability.

More details should be given regarding the acquisition of the measurement data. This includes the exact location of the inlet on the ship, the flow rate in the inlet, the material of the inlet tubing, the heating of the inlet line, the time resolution of the data set, and a depiction or mathematical representation of the spectroscopic correction given in Eq. 3.

The dataset of the measurements has not been made accessible to the reviewers. It is therefore not possible to inspect the actual dataset at this point.

Several recent literature with other ship-based vapour isotope measurements in high latitudes should be considered to be referenced in this manuscript: Thurnherr et al., 2020; Thurnherr et al., 2021; Brunello et al., 2024; Sodemann et al., 2024 (see references).

Minor comments:

The description of the use of MixSIAR is insufficient and currently not reproducible.

Figure 7a and b look exactly the same (same sample points and axis) while they should reflect delta18O and d-excess. It is very difficult to separate the ship track/arrival points of the trajectories in Figure 8.

Throughout manuscript: Using the word ‘melt regions’ suggests that the measurements were taken in an area where there was an ongoing melting process. Do you rather mean ice-free and ice-covered regions as there were large ice-free regions where ice was already melted?

Detailed comments:

Line 12: ‘ice-phase processes’. You mean ‘in cloud ice phase processes’?

Line 21: Change ‘These parameters’ to ‘these isotopes fractionate’

Line 31: Add reference to these studies after the words ‘ice-free ocean’

Line 35: Which sources are meant here, local or lower latitude sources?

Line 42: This is not consistent with Thurnherr and Aemisegger (2022) who observed negative d-excess in extra-tropics.

Line 130: What sources were used as input.

Line 130 (paragraph 2.5): But this uses water stable isotopes as passive tracers? Why not use ensemble trajectory analysis HYSPLIT as tracer option rather than statistical mixing method?

Line 133: ‘Figure 1 shows the’, this comes a bit unexpected and is posed as a conclusion while the ‘co-variation’ has not been introduced to the reader and is also not clear from this figure. Rather reword to something like: Figure 1 suggests/shows a that there may be a co-variation…..

Line 179: no more so on relative humidity refer to Pfahl and Sodeman (2014)

Line 204: The word ‘trajectory’ is confusing in this context, you mean ‘curve’?

Line 231: you mean d-excesss in surrounding vapour

Line 236: supersaturation was only mentioned in relation to cloud processes?

Line 278: supersaturation is not a process but a condition? You mean crystal formation under supersaturated conditions?

Line 281: It is usually the opposite, air masses from the south are warm and moist in general, please explain.

Line 301: (single) 10m trajectory arrival does not necessarily represent the moisture source.

References

Brunello, C.F., Gebhardt, F., Rinke, A., Dütsch, M., Bucci, S., Meyer, H., et al. (2024). Moisture transformation in warm air intrusions into the Arctic: Process attribution with stable water isotopes. Geophysical Research Letters, 51, e2024GL111013. https://doi.org/10.1029/2024GL111013

Thurnherr, I., & Aemisegger, F. (2022). Disentangling the impact of air–sea interaction and boundary layer cloud formation on stable water isotope signals in the warm sector of a Southern Ocean cyclone. Atmospheric Chemistry and Physics, 22(15), 10353-10373.

Sodemann, H., Weng, Y., Touzeau, A., Jeansson, E., Thurnherr, I., Barrell, C., et al. (2024). The cumulative effect of wintertime weather systems on the ocean mixed-layer stable isotope composition in the Iceland and Greenland Seas. Journal of Geophysical Research: Atmospheres, 129, e2024JD041138. https://doi.org/10.1029/2024JD041138.

Thurnherr, I., Kozachek, A., Graf, P., Weng, Y., Bolshiyanov, D., Landwehr, S., Pfahl, S., Schmale, J., Sodemann, H., Steen-Larsen, H. C., Toffoli, A., Wernli, H., and Aemisegger, F., 2020: Meridional and vertical variations of the water vapour isotopic composition in the marine boundary layer over the Atlantic and Southern Ocean, Atmos. Chem. Phys., 20 (9), 5811–5835.

Thurnherr, I., Hartmuth, K., Jansing, L., Gehring, J., Boettcher, M., Gorodetskaya, I., Werner, M., Wernli, H., and Aemisegger, F.: The role of air–sea fluxes for the water vapour isotope signals in the cold and warm sectors of extratropical cyclones over the Southern Ocean, Weather Clim. Dynam., 2, 331–357, https://doi.org/10.5194/wcd-2-331-2021, 2021

Reply
Citation: https://doi.org/10.5194/egusphere-2025-5306-RC1
RC2: 'Comment on egusphere-2025-5306', Anonymous Referee #2, 16 Jan 2026 reply

The manuscript by Zhang et al. uses a new set of cruise-based water vapor isotopic observations to examine Arctic evaporation over a broad spatial scale and range of sea ice conditions. The manuscript presents a valuable observational dataset. Pan-Arctic observations from a single cruise campaign are challenging to conduct, where measurements are collected in a relatively short time window, providing important like-for-like comparisons. The dataset on its own is a valuable contribution, and leads this manuscript to having value as a publication. However, there are some potentially questionable analyses and conclusions drawn that warrant further examination before publication.
In general, the observations presented in this study are generally sound and important, but the interpretations and conclusions drawn do not always fit the analyses. I describe some of these below.
Major comments:

1. Motivation of paper vs analyses presented.

The key motivating point of the manuscript is the discrepancy between different studies that show divergence relationships between d-excess and sea ice coverage. This is a good question (the introduction is very well written), and one that can be addressed with this dataset. However, the analyses as presented do not directly address this problem.
a. Defining the local evaporation signal

The authors simply use the Merlivat and Jouzel 1979 (MJ79) model for predicting the local evaporation signal in the MixSIAR model (section 3.6), despite this being the focus of the initial questions of the manuscript. They do examine MJ79 results compared to the observations, but only for air temperatures over 5 C. Why not try this with lower temperatures? I suspect it breaks down against the observations, which then brings into question why it is used in the mixing model. These
b. Use of mixing model

In addition to the local signal question above, there are other questions with how the mixing model is designed to address the key questions. There are very few details presented on how the various end members are defined; this needs to be clearer. As described many times previously in the manuscript, there are isotopic changes from the source to the observation site during advection of the different air masses. If is unclear how, if at all, Rayleigh distillation is dealt with in this model. There will be considerable changes to those lower latitude end member air mass isotopic values by the time they reach the Central Arctic.
c. Back-trajectory analyses.

More details and justification on HYSPLIT methods needed. With the broad goal of defining where the moisture is coming from, the selected approach here does not seem to match the project needs. A simple 5 day back-trajectory from a single height only tells us where the air at that height came from over that period. Finding the source of moisture with back-trajectories requires looking at humidity changes (moisture uptake) and/or other property changes over the trajectory (e.g., Sodemann et al. 2008). 5 days has been shown to not be long enough at times in the Arctic with significant recirculating in the central Arctic basin and/or long transport trajectories from lower latitudes. Also, small differences in initialization height are known to, at times, result in considerably different trajectories depending on the wind distribution above the surface as represented by the reanalysis data product. It would be ideal to initiate the trajectories from at least several heights near the surface to ensure that appropriate air mass transport is captured.
2. Context of analyses

One other area that could benefit the manuscript is putting the observed relationships into context of prior observations. These observed relationships are the strength of this study, and should be emphasized. The suggestions here are not to take away from or question the analyses presented, but could provide important context to these new observations. For example, see how the spatial patterns compare with other cited studies, e.g., Brunello et al. (2023) in the Central Arctic basin. Examine how the slopes of d-excess relationships with various parameters (e.g., RH) compare with those observed in similar regions, e.g., Bonne et al. (2019). How do the fractions of local evaporation contributions compare with other observational or modeling studies (done with or without water isotopes)?
Without having to do any major new or overhauled analyses, my suggestion would be to really focus in on these observed relationships, compare them with other studies, and highlight the consistencies and discrepancies found here. The uniqueness of this large spatial observational assessment gives an important opportunity to do this effort, and, to me, would be a quite valuable contribution from this effort.

Minor comments.

Line 63. More details needed on the sampling setup. For example: where on the ship was the inlet located; was the inlet tube heated?
Line 97. Consider changing the name of ‘Melt Region’ to something like ‘Open Water Region’ as there is not necessarily any melt associated with the given region.
Figure 1 and 3. What is the temporal resolution of the datasets?
Figures 4 and 5. Consider adding a panel showing all of these together, and/or make the axes the same for each panel. It is not straightforward to compare the relationships as plotted other than comparing slopes.
L228 and Figure 6. ‘Steeper than predicted’…with a single Rayleigh curve. But the air masses are a mixture of moisture sourced from many locations (with different temperatures, etc). I would suggest including several Rayleigh curves starting from different initial air masses. This might better show the mixing the authors are trying to show with their mixing model.
Figure 7. Why only look at conditions greater than 5 C (see 1a above for more)? Also, both panels of the plot seem like they are showing the same d-excess data, not d18O and d-excess data.
Line 335. Should this say ‘producing lower d-excess’ instead of higher?

References

Bonne, J. L., Behrens, M., Meyer, H., Kipfstuhl, S., Rabe, B., Schönicke, L., et al. (2019). Resolving the controls of water vapour isotopes in the Atlantic sector. Nature communications, 10(1), 1632.
Brunello, C. F., Meyer, H., Mellat, M., Casado, M., Bucci, S., Dütsch, M., & Werner, M. (2023). Contrasting seasonal isotopic signatures of near‐surface atmospheric water vapor in the central Arctic during the MOSAiC campaign. Journal of Geophysical Research: Atmospheres, 128(24), e2022JD038400.
Sodemann, H., Schwierz, C., & Wernli, H. (2008). Interannual variability of Greenland winter precipitation sources: Lagrangian moisture diagnostic and North Atlantic Oscillation influence. Journal of Geophysical Research: Atmospheres, 113(D3).

Reply

Citation: https://doi.org/10.5194/egusphere-2025-5306-RC2

Yuankun Zhang, Zhongfang Liu, Dongsheng Li, Zhiqing Li, and Hebin Shao

Viewed

Total article views: 280 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
201	69	10	280	32	29

HTML: 201
PDF: 69
XML: 10
Total: 280
BibTeX: 32
EndNote: 29

Views and downloads (calculated since 03 Dec 2025)

Month	HTML	PDF	XML	Total
Dec 2025	136	56	9	201
Jan 2026	65	13	1	79

Cumulative views and downloads (calculated since 03 Dec 2025)

Month	HTML	PDF	XML	Total
Dec 2025	136	56	9	201
Jan 2026	65	13	1	79

Viewed (geographical distribution)

Total article views: 269 (including HTML, PDF, and XML) Thereof 269 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 19 Jan 2026

Short summary

Arctic warming and melting sea ice are changing how moisture forms and moves through the atmosphere. To find out where this moisture comes from, we measured water vapor composition during a research voyage across the Arctic. We discovered that as sea ice melts, local evaporation increases but long-distance transport from lower latitudes still dominates. By understanding this shift, we can better predict future Arctic climate and more accurately interpret clues about the past from ancient ice.


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