The Oxygen Minimum Zones (OMZs) in the Indian Ocean

Oboh, Eugene Odion; Lahajnar, Niko; Gaye, Birgit; Shrikumar, Avanti; Rixen, Tim

doi:10.5194/egusphere-2025-4712

Preprints

https://doi.org/10.5194/egusphere-2025-4712

Preprints

12 Nov 2025

| 12 Nov 2025

The Oxygen Minimum Zones (OMZs) in the Indian Ocean

Eugene Odion Oboh, Niko Lahajnar, Birgit Gaye, Avanti Shrikumar, and Tim Rixen

Abstract. The Oxygen Minimum Zones (OMZs) in the northern Indian Ocean (i.e., the Arabian Sea and the Bay of Bengal) are among the most intense OMZs in the world’s oceans. While there is no clear evidence of a significant change in the Bay of Bengal (BoB) OMZ, the Arabian Sea (AS) OMZ followed the global trend and expanded in the last decades until 2013. Since then, however, this trend has reversed, and the AS OMZ seems to have shrunk. The stability of the BoB OMZ as well as the expansion and shrinkage of the AS OMZ in response to global warming is poorly understood. In this study we redefined the water masses and employed an extended Optimum Multiparameter (eOMP) Analysis to investigate changes in the oxygen supply due to mixing and biological oxygen consumption dynamics in these OMZs based on empirical field data from the Global Ocean Data Analysis Project version 2 (GLODAPv2) and a research cruise conducted with a German research vessel Sonne in 2024. Our findings reveal in line with previous studies a reversal in the expansion trend of the AS OMZ but also a shrinkage of the BoB OMZ between 1995 and 2016. In both regions this is due to an increased northward influx of oxygen-rich water masses from southern Indian Ocean, combined with a reduced contribution from relatively oxygen-poor local and equatorial water masses. However, we also observed that increased physical oxygen supply was accompanied by an increased biological oxygen consumption. These changes are likely linked to the slowdown of the global thermohaline circulation in the Indian Ocean. The slowdown is accompanied by a reduced inflow of the Indonesian Throughflow Water into the Indian Ocean and a lower output of Indian Ocean waters via the Agulhas Current/Leakage (at 32° S) into the Atlantic Ocean. A resulting increase in the residence time of water masses in the Indian Ocean is consistent with the detected biological oxygen consumption while the weaker zonal circulation seems have favored the meridional circulation which carried water from the southern Indian Ocean northwards. This implies a coupling between the OMZ in the Indian Ocean and climate change via the effect of the latter on the global thermohaline circulation as also seen in palaeoceanographic archives, whereas the drivers in past and today differs.

Received: 24 Sep 2025 – Discussion started: 12 Nov 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

Preprint (PDF, 1640 KB)

Supplement (538 KB)

Download & links

Eugene Odion Oboh, Niko Lahajnar, Birgit Gaye, Avanti Shrikumar, and Tim Rixen

Status: final response (author comments only)

RC1:
'Comment on egusphere-2025-4712', Anonymous Referee #1, 15 Dec 2025
Review of manuscript egusphere-2025-4712 by Oboh et al., “The Oxygen Minimum Zones (OMZs) in the Indian Ocean”
This is my first review of the manuscript “The Oxygen Minimum Zones (OMZs) in the Indian Ocean”. The manuscript analyses shipboard transects from 1995 and 2018, along with available observation-based products, to assess recent changes in the Northern Indian Ocean OMZs. The study is a topical analysis of available datasets and aims to address gaps in the literature regarding recent variability in OMZ extent. However, I have several concerns about the implementation of the methodology, interpretation of the results, and presentation of the manuscript. The study requires major revisions before being considered for publication. I outline my major points of concern below.

Methodology:
Perhaps my biggest concern is with the implementation of the eOMP method. In this study, the authors define 9 unknowns (8 source water types plus remineralization). This requires at least ten constraints for an “optimum” multiparameter analysis or nine constraints for a simple determined set of equations. What are these constraints? I can only count 6 unique constraints – temperature, salinity, oxygen, phosphate, nitrate, and mass.

The definition used for PGW in this study, which here has an oxygen concentration of 15.64 umol/kg, is misleading. PGW is generally considered to be oxygen saturated, and gets quickly diluted once it enters the Arabian sea (e.g. Font et al., 2024, JGR:Oceans). The definition here is thus of highly diluted PGW, which is essentially just the AS OMZ water (i.e. the water mass being evaluated). This is why PGW is so prevalent in Figure 3a. The definition of PGW here will make it difficult to connect changes in the source water fraction to changes in the actual PG outflow rates or properties.

Interpretation:
A recurring point in the manuscript is that this study differs from past studies because it focuses on “surface-originated water masses”. This is quite misleading. First of all, some of the source water types are not really surface-originated at all, like ITFW or BoBW. But more importantly, for the source waters which do outcrop, the corresponding source water types are not defined at those outcrops. Rather, they are defined in regions where those waters enter the Indian Ocean (Table 3), after they have already experienced significant water mass transformation. So, changes in those defined source water types do not necessarily reflect changes in the formation rates/conditions of those water masses. This is especially evident in the oxygen properties used for the source water types (would be nearly saturated at formation for many of the source water types).

The study is inferring climate-driven changes by comparing transects in 1995 and 2018, each providing a snapshot within a single year. The influence of interannual and decadal variability has to be discussed at some point in the manuscript. Limitations of the methodology in general need to be more clearly addressed.

Presentation:
The manuscript needs to be structured and presented as a research article, starting with the title. The title suggests that the article is a discursive piece, or even a review, which is misleading. Plenty of literature already exists on OMZs in the Indian Ocean and the purpose of this article is not to be a comprehensive resource (like a textbook or an article like Schott and McCreary, 2001). Please choose a title that highlights the new contribution of the article. Confusion about the article’s identity is also evident throughout the introduction. For example, Section 1.1 provides far too much information that is not relevant for the reader to understand the results of this study. Don’t be afraid to rely on existing literature to provide a comprehensive reference so that you can focus on details that the reader needs to understand the present results.

Please separate results and discussion. As it is written, it is unclear what conclusions are drawn from the presented results and what conclusions are drawn from the literature. For example, Section 3.2 discusses connections to the literature but do not present any new results. While Figures 4 and 5 are technically presented in this section, they are not explained or analysed and their connection to the points discussed are unclear.

In the manuscript, only Figures 3 and 6 are explained as results in any meaningful way. For figures 4, 5 and 7, either make them supplementary or actually explain them and how you interpret them.

Other points:
I would suggest using more specific language than “water masses” throughout. In the manuscript, water masses refer to both the source waters and the destination waters and it can get quite confusing. I would recommend using the conventional term of “source water types” or something like that.

Equation 4: How does this compare to the remineralization term calculated within the eOMP which tracks the conversion from oxygen to phosphate? And why is the oxygen residual added here. Should it instead be set aside as an uncertainty bound?

Add reference for Figure 1. Maybe use this figure to highlight your source water types and the pathways that they take to the OMZs. Help guide the reader, they are reading to understand the new result that you are contributing.

Line 668: "...increased physical oxygen supply via the inflow of oxygen-rich water masses from the southern Indian Ocean is able to compensate for the increased biological oxygen consumption." Careful when discussing the opposing roles of physical supply and biological consumption. It's likely that increased biological consumption is simply a response to increased physical supply or vice versa.

Line 680: "The current changes indicate the presence of a longer time-scale cycle in the Indian Ocean OMZs which aligns with paleoceanographic archives, however, this cycle seem to be reversing…” The connections to paleoceanographic signals come off as very speculative and unsubstantiated. Separation of results and discussion is crucial here to make this thought process more clear.
Citation: https://doi.org/10.5194/egusphere-2025-4712-RC1
CC1:
'Comment on egusphere-2025-4712', Yun Qiu, 27 Dec 2025
This study investigates the recent trend reversal of the Arabian Sea (AS) Oxygen Minimum Zone (OMZ) and the shrinkage of the Bay of Bengal (BoB) OMZ. By applying an extended Optimum Multiparameter (eOMP) analysis to historical data and cruise observations, the authors attribute these changes to a basin-wide shift in water mass composition. They found that increased ventilation by oxygen-rich water masses from the southern Indian Ocean (e.g., STSW, mAABW), which outpaced increased biological oxygen consumption, is the primary driver of the trend reversal of AS OMZ. The manuscript addresses a timely and important topic, and the data-driven approach provides valuable insights. However, several issues regarding the clarity of the mechanistic narrative and the presentation of evidence need substantial revisions. In addition, the manuscript still contains some grammatical errors and inappropriate use of transition words throughout (e.g., However, Nevertheless, In contrast), which significantly affect readability and the clarity of the scientific arguments. A thorough language revision is required. I therefore recommend major revision.

Major comments:
The current title is overly broad and does not reflect the actual scope of the study. A more specific, process-oriented title is recommended. For example: Recent changes in oxygen minimum zones of the northern Indian Ocean.

The abstract is overly long. The authors are encouraged to streamline the abstract by focusing on the main findings and their key implications.

The conclusions of this study about the water mass composition are impressive. However, the argumentation could be made more intuitive and compelling. I recommend that the authors include basic dissolved oxygen (DO) distribution maps in the revised manuscript. For instance, providing vertical section plots of DO concentration along key transects, comparing data from 1995 (or earlier) with recent years, would visually demonstrate the changes in oxygen content within the OMZ core and the migration of its boundaries. Furthermore, a direct comparison between the observed DO changes and the shifts in water mass composition identified by the eOMP analysis would significantly strengthen the paper. Correlating these spatial and temporal patterns would provide a clearer,more direct factual foundation for the subsequent complex mechanistic eOMP analysis, directly linking the physical oxygen supply to the documented oxygen changes.This would greatly enhance the clarity and persuasive power of the manuscript.

About Figure 6: It is unclear how the observed dissolved oxygen changes shown in Figure 6 were calculated. If these values represent regional or OMZ-core averages, the corresponding spatial boundaries and averaging method should be clearly specified to ensure consistency with the oxygen supply estimates derived from the water mass analysis. Furthermore, the interpretation that the difference between mixing-derived oxygen supply changes and observed oxygen changes directly reflects variations in biological oxygen consumption requires further justification. In particular, this approach appears to implicitly assume that other processes affecting oxygen, such as local mixing or diffusive transport, remain unchanged or negligible. Please clarify and discuss the validity of this assumption.

Minor comments:
Line 44: does not only affect -> not only affects.

Line 60: are the semi-enclosed basins …-> are semi-enclosed basins …

Lines 59-65: This paragraph suffers from major language issues and an unclear conceptual framework. The definitions of hypoxia and microbial hypoxia are currently confusing and partially incorrect, and the sentence “Hypoxia defines the oxygen threshold…”is grammatically incorrect. Please revise for both language and conceptual clarity.

Line 68: Nevertheless, …-> In particular, …

Line 75: OMZ -> OMZs

Line 89: close link between …to …-> a close link between … with …

In Figure 3, the seven bar charts have very similar visual appearance, while the panel sizes and alignment are inconsistent. A reorganization of the figure layout, along with clear indication of the study region in each subpanel, would substantially improve clarity.
Citation: https://doi.org/10.5194/egusphere-2025-4712-CC1
- AC1: 'Reply on CC1', Eugene Oboh, 07 Jan 2026
  
  Thanks, your suggestions will be considered in our revised manuscript.
  
  Citation: https://doi.org/10.5194/egusphere-2025-4712-AC1
RC2:
'Comment on egusphere-2025-4712', Anonymous Referee #2, 13 Jan 2026
General Assessment
The manuscript by Oboh et al. examines recent changes in dissolved oxygen in the northern Indian Ocean, with a focus on the Arabian Sea and Bay of Bengal Oxygen Minimum Zones (OMZs). Using hydrographic observations from GLODAPv2 (1995–2024), the authors apply an Extended Optimum Multiparameter (eOMP) analysis to investigate the drivers of reported oxygenation trends in the Arabian Sea and the apparent lack of deoxygenation in the Bay of Bengal. They attribute oxygenation at intermediate depths primarily to weakened local ventilation and a reduced contribution from the Indonesian Throughflow (ITF), accompanied by increased inflow of oxygen-rich waters from the Southern Indian Ocean.
While documenting recent changes in Indian Ocean OMZs is scientifically important, and the manuscript is generally well written, I have serious concerns regarding both the methodology and the interpretation of the results. In particular, the application of the eOMP framework relies on assumptions of stationarity in source water properties that are unlikely to hold over decadal timescales. Moreover, attributing differences between two temporal snapshots to long-term climate trends without explicitly accounting for natural decadal variability is problematic. Owing to these fundamental issues, I cannot recommend publication in Biogeosciences in its current form. My major concerns are detailed below.
Major Comment #1: Methodological limitations of eOMP analysis under non-stationary conditions
My primary concern relates to the application of the eOMP analysis and its suitability for diagnosing decadal oxygen changes using the available data. The authors define source water mass properties using World Ocean Atlas (WOA) climatologies, implicitly assuming that the biogeochemical properties (particularly dissolved oxygen) of these source waters remain constant over time. However, ventilation changes can manifest not only through variations in water mass fractions (mixing ratios), but also through temporal changes in the properties of the source waters themselves.
Dissolved oxygen concentrations in source waters can vary substantially on interannual to decadal timescales. If, for example, a source water mass such as the ITF becomes more oxygenated during the study period relative to its climatological definition, the OMP algorithm—constrained by mass conservation—may artificially reduce the estimated fraction of that water mass (or increase the contribution of a high-oxygen end-member such as mAABW) in order to satisfy the oxygen budget. This can lead to spurious attribution of changes in water mass contributions. The authors should explicitly test and demonstrate the robustness of their conclusions to plausible temporal variability in end-member properties.
Major Comment #2. Selection of source water masses and physical consistency
The selection of source water masses appears incomplete and, in some cases, physically inconsistent with the timescales considered. Key water masses known to ventilate the Indian Ocean thermocline and intermediate depths—such as Indian Central Water (ICW) and Subantarctic Mode Water (SAMW)—are not included in the analysis. Omitting these major ventilators may force the OMP solution to compensate by assigning unrealistic fractions to other water masses (e.g., STSW or mAABW), resulting in biased or physically implausible mixing ratios.
Attributing oxygenation of intermediate layers (upper ~1200 m) to Modified Antarctic Bottom Water (mAABW) is difficult to reconcile with physical constraints. Ventilation of OMZ cores by bottom waters originating at depths >3000 m within a ~20-year interval appears implausible.
The analysis compares two snapshots separated by approximately two decades. However, transit times for Southern Ocean-derived waters to reach the northern Indian Ocean via mean circulation are typically longer than this interval. It is therefore unclear how the reported changes between 1995 and 2018 can be attributed to remote Southern Ocean ventilation rather than to local or regional processes.
Major Comment #3. Distinction between decadal variability and long-term trends
The manuscript frequently interprets differences between the two analyzed periods as indicative of long-term climate trends. However, the analysis is based on comparisons between two discrete snapshots rather than continuous time series. Numerous studies (e.g., Long et al., 2016) have shown that internal climate variability can drive large-amplitude oxygen fluctuations that mask or even reverse long-term trends over multi-decadal timescales.
No justification is provided for assuming that the difference between the selected years represents a secular trend rather than a phase of natural variability. The discussion should explicitly address this limitation and more carefully distinguish between internal variability and forced trends.
Major Comment 4. Study objectives, scope, and presentation
The scope and objectives of the study require clearer definition. The title (“The Oxygen Minimum Zones…”) is overly broad and does not reflect the limited spatial and temporal scope of the analysis. The introduction should more clearly state whether the primary objective is to evaluate the eOMP methodology, document a specific observational change, or attribute long-term trends.
In addition, the eOMP methodology requires clearer documentation. The definition of properties A and B, the choice of weights (lines 255–258), and the selection of source water masses should be more thoroughly justified based on existing literature and physical constraints, including transport timescales. The spatial extents of the Arabian Sea (AS) and Bay of Bengal (BoB) OMZs, as well as the portions of the transects used to estimate OMZ oxygen budgets, should be explicitly indicated in Figure 2.

Some more specific comments
P4, line 111:

Section 1.1 is awkwardly placed. Key background information on OMZs appears after the research questions are formulated. This material should be moved earlier in the Introduction and should not appear as a subsection.

P6, line 189:

The steady-state assumption underlying the eOMP method—that source water biogeochemical properties (oxygen, nutrients) do not change over time—should be explicitly discussed. How do the authors address the possibility that source water oxygen concentrations themselves are different between the two snapshots?

P11, line 310:

Water mass properties are derived from WOA climatologies. In several Indian Ocean source regions (e.g., the Red Sea and Persian Gulf), oxygen and nutrient observations are sparse or absent in WOD. Consequently, WOA values may be poorly constrained or extrapolated from nearby regions, which could strongly bias the analysis. This limitation must be discussed.

P12, Table 2:

Persian Gulf Water is known from local observations to be extremely oligotrophic. The nutrient concentrations assigned here are an order of magnitude higher than reported values, suggesting that what is referred to as PGW is mostly Arabian Sea waters.

P13, Table 3:

PGW and RSW properties appear to be estimated downstream within the Arabian Sea rather than at their formation or entry points, implying that these “water masses” are already mixtures. This should be discussed and the implications clarified.

Additionally, some water masses are defined using shallow layers (e.g., ASHSW, BOBW, STSW), while others are defined at depth. Since all water masses originate at the surface, a consistent approach should be adopted and justified.

Lines 373–375:

The phrase “similar climate anomalies” is unclear. The role of interannual and decadal variability between the two periods needs to be more explicitly examined or discussed.

P14, lines 385–387:

While redefining water mass properties for different periods may account for warming, similar temporal changes likely affect oxygen and nutrient content as well.

P18, line 501:

What are the expected ventilation timescales from the Southern Ocean (AAIW, mAABW)? Over a ~20-year period, the observed changes are more likely dominated by local ventilation and local biogeochemical processes.

P20, lines 540–541:

The study does not demonstrate trends over recent decades, but rather differences between two transects separated by two decades. Given strong interannual and decadal variability, conclusions about long-term trends are not supported.
Citation: https://doi.org/10.5194/egusphere-2025-4712-RC2

Eugene Odion Oboh, Niko Lahajnar, Birgit Gaye, Avanti Shrikumar, and Tim Rixen

Supplement

https://doi.org/10.5194/egusphere-2025-4712-supplement

Eugene Odion Oboh, Niko Lahajnar, Birgit Gaye, Avanti Shrikumar, and Tim Rixen

Viewed

Total article views: 562 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
436	97	29	562	37	16	19

HTML: 436
PDF: 97
XML: 29
Total: 562
Supplement: 37
BibTeX: 16
EndNote: 19

Views and downloads (calculated since 12 Nov 2025)

Month	HTML	PDF	XML	Total
Nov 2025	200	20	12	232
Dec 2025	134	55	15	204
Jan 2026	102	22	2	126

Cumulative views and downloads (calculated since 12 Nov 2025)

Month	HTML	PDF	XML	Total
Nov 2025	200	20	12	232
Dec 2025	134	55	15	204
Jan 2026	102	22	2	126

Viewed (geographical distribution)

Total article views: 548 (including HTML, PDF, and XML) Thereof 548 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 17 Jan 2026

Short summary

We studied water mass distribution and oxygen consumption changes in low oxygen regions of the Indian Ocean and found slightly more oxygen present in these regions in 2016/2018 compared to 1995, contrary to the expected decline. This change is linked to a shift in ocean circulation that brought more oxygen-rich waters from the south to the north. Our results show that the development of these low oxygen regions is connected to the relationship between climate change and ocean circulation.


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