the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Earth system modelling of mercury using CESM2 – Part 3: Oceanic model POP2/Hg v1.0
Abstract. Mercury (Hg) is a globally distributed toxicant with complex cycling in the ocean, involving redox reactions, air–sea exchange, and microbial methylation. We present POP-Hg v1.0, a new global ocean mercury model developed within the ocean component of CESM2 (POP2) and coupled to the Marine Biogeochemistry Library (MARBL). POP2/Hg v1.0 simulates dissolved elemental Hg (Hg0), oxidized Hg (HgII), particulate-bound Hg (Hgp), and Methylmercury (MeHg = MMHg + DMHg), linking their transformation to ecosystem processes and dynamics of particulate organic carbon (POC). The model captures observed large-scale features of marine Hg. Surface Hg0 concentrations range from 10 to 120 fmol L-1(fM, 1 fM = 10-15 mol L-1) and are elevated in tropical and subtropical oceans due to strong photoreduction and limited evasion. Subsurface Hg0 maxima emerge in upwelling zones, reflecting remineralization-driven HgII reduction. Surface HgII concentrations peak in regions of high atmospheric deposition and productivity and rise at depth in oxygen-deficient zones. Hgp is closely associated with high POC and HgII levels in the surface ocean but attenuates rapidly with depth due to remineralization. MeHg accumulates in the subsurface and deep sea, broadly consistent with observations. Total Hg (HgT) ranges to above 2.0 pmol L-1 (pM, 1 pM = 10-12 mol L-1) in the surface, with open-ocean surface concentrations matching measurements. Deep-ocean HgT accumulation reflects sustained particle flux and remineralization. The biological pump strongly shapes vertical Hg transport. Above 1000 m, soft POC carries 98.6 % of Hgp flux, driving export and remineralization. Below 1000 m, hard POC becomes the dominant carrier, contributing over 99 % of Hgp flux to sediments, although the total flux is small (0.19 t yr-1). Sensitivity tests show that altering redox rates significantly affects the vertical distribution of Hg transformations. Reduced surface reduction and oxidation rates suppress Hg⁰ production in the upper ocean, weaken seasonal variability, and shift net reduction deeper into the water column. By embedding mercury chemistry into a fully coupled biogeochemical and physical ocean model, POP2/Hg v1.0 offers a process-resolving platform for assessing mercury cycling and its response to environmental change. It enables integrated Earth system simulations and improves predictive capacity for marine mercury under future climate and policy scenarios.
- Preprint
(11592 KB) - Metadata XML
-
Supplement
(103 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-3880', Anonymous Referee #1, 22 Apr 2026
-
AC1: 'Reply on RC1', Yanxu Zhang, 08 Aug 2026
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-3880/egusphere-2025-3880-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Yanxu Zhang, 08 Aug 2026
-
RC2: 'Comment on egusphere-2025-3880', Anonymous Referee #2, 11 Jul 2026
This manuscript presents POP2/Hg v1.0, a new global ocean mercury model embedded in the ocean component of CESM2 (POP2) and coupled to MARBL. The work is a worthwhile contribution: coupling mercury chemistry to a prognostic biogeochemistry library with separate soft and hard POC pools is a genuine advance over offline or climatology-driven predecessors, and the model reproduces several large-scale features of marine mercury. However, several issues must be resolved before the paper can be accepted. Most of these require clarification, correction, and modest added analysis rather than new simulations.
- Lines 350 and 450: Section 4.1 reports a global Hg0 net evasion of ~4.43 Mg yr-1 against an HgII deposition of 4.20 × 103 t yr-1, then calls this "near-equilibrium exchange." Taken literally, evasion (4.43 t) is ~1000× smaller than deposition (4200 t), which is the opposite of equilibrium. The Conclusion instead gives a net evasion of ~4.5 × 103 t yr-1, differing from Section 4.1 by three orders of magnitude. Published global ocean Hg0 evasion is on the order of thousands of Mg yr-1, so "4.43 Mg yr-1" appears to be a unit error. Please reconcile these numbers and re-examine the equilibrium statement.
- Lines 263–265 and 334: Figure 11 is described as "year 10 (balanced state)", yet deep-ocean ventilation timescales are centuries to over a millennium. Consistent with this, the authors state that the elevated Hg0 at ~1000 m (up to 0.8 pM) is contributed by "the initial condition of the model". This undermines confidence in all 985/1000 m results (Hg0, HgII, Hgp, MeHg, HgT), which may reflect the spin-up initial field rather than the model's biogeochemistry. Please justify the spin-up duration for the deep ocean, or explicitly qualify that deep-ocean fields are not equilibrated and interpret them accordingly.
- Lines 365, 371, 375 versus 425–430: Section 4.2 reports surface HgII reduction at 1.53 × 105 t yr-1, intermediate net reduction at 1.74 × 102 t yr-1, and deep net oxidation at 3.74 × 103 t yr-1. The sensitivity section then reports surface, mid-depth, and deep net reduction in "Mg yr-1" with values of 8.22, 3.62/1.53, and 1.11/3.69. These cannot be reconciled with the Section 4.2 values (e.g., mid-depth 1.74 × 102 t yr-1 vs. 3.62 Mg yr-1). Please harmonize units and magnitudes throughout, ideally with a single flux-budget table.
- Sections 3.1–3.4: The model–observation evaluation is entirely qualitative; no quantitative skill metrics are reported. The Results overlay pointwise observations on maps (Figs. 7, 10, 11) and describe agreement in words ("reasonably well," "good agreement"), but no correlation, RMSE, bias, or model-vs-observation scatter/statistics are given anywhere. For a model-description-and-evaluation paper, at least basic quantitative skill statistics against the compiled datasets (cited in Sect. 2.4) are expected and are computable from data already in hand. I recommend adding them.
- Table 1 and Figure 6 caption: In Table 1, the symbols "OCRR," "hv," and "T_c/T_k" (reaction 7) are used but never defined in the text. In Figure 6, panel (a) atmospheric Hg0 "concentration" is labelled ng m-2 d-1 (a flux unit; concentration should be ng m-3), and panel (b) HgII deposition is captioned mol m-2 s-1 while the colorbar reads pg m-2 s-1. Please define all symbols and correct the units.
- Lines 187–189, 249, 320 (plus Eqs. 5–7): Duplicated text, typographical errors, and a POC/bioC notation mismatch. Lines 187–189 contain a near-verbatim duplicated sentence about Hgp gravitational settling. Typos include "Hg0remain" and "below 0.009o4 pM". In addition, sorption (Eq. 5) partitions HgII onto POC only, whereas the sinking-velocity formulation (Eqs. 6–7) uses POC + bioC (living biomass); please clarify whether Hgp is carried by living biomass and reconcile the notation. A careful language and consistency edit throughout is advised.
Citation: https://doi.org/10.5194/egusphere-2025-3880-RC2 -
AC2: 'Reply on RC2', Yanxu Zhang, 08 Aug 2026
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-3880/egusphere-2025-3880-AC2-supplement.pdf
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 1,517 | 921 | 123 | 2,561 | 283 | 121 | 124 |
- HTML: 1,517
- PDF: 921
- XML: 123
- Total: 2,561
- Supplement: 283
- BibTeX: 121
- EndNote: 124
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
In this manuscript the authors present a new global ocean mercury model developed within the ocean component of CESM2. There have been a few efforts to develop such ocean mercury numerical models in the past, but this represents a significant step forward because of its coupling to marine ecosystem processes. It also complements existing CESM2 mercury modules focusing on the atmosphere, terrestrial processes, and river transport. The manuscript is well written, clearly organized and presents state-of-the-art description of the processes implemented in this ocean mercury module.
Overall, this work represents a substantial contribution to modeling science within the scope of the journal. The work is of high quality, and I have only a few minor comments and suggestions for the authors to consider.