The study investigated the transport and long-term fate of radionuclides released from wastewater in the global oceanic environment, using FDNPS release as a case study. The authors applied a more comprehensive transport and biogeochemical model – MITgcm ocean tracer model to run a short- to medium-term predictions for current status and model validation, and then predicted a longer-term fate until 2100 under an intermediate scenario and the low-end and high-end emission scenarios for the uncertainty range. The work is extremely significant, as 1) the health risks posed by the radionuclides released after the earthquake and the ongoing release with intentional wastewater discharge is of concern globally; and 2) although previous studies have modelled some radionuclides regionally after the disaster, we obviously would like to know the risks of all major radionuclides in the long-term future in the global ocean, considering the continuous release. This study used a model capable to consider multiple radionuclides and their essential biogeochemical processes, which is important and interesting.

The manuscript is well organized and prepared. The model performance is validated reasonably using observational data. The authors clearly showed the temporal and spatial pattern of radionuclides. The detailed questions below should be addressed before acceptance.

1. For the emissions in the method section, is there any reference for the assumption that the discharge will last until 2050? Is the assumption more likely to be for a conservative or aggressive assessment?
2. Why did the authors choose to simulate the short-term concentration till 2016 for validation? Is this relevant to the sampling year of the available observation data?
3. It is not to say that you have to assess the consequent risks quantitatively, but is it possible to add a bit discussion on the last section. For example, what’s the potential short-term and long-term exposure risks. Is there enough toxicological data for assess this for the combined contamination of such many radionuclides. Can we be sure about the risks based on the predications and available toxicological data of radionuclides both in the short-term and long-term exposure?
4. Please check the language again, as there are minor grammar mistakes, for example, lines 94-96

General Comments

This paper employs a state-of-the-art ocean tracer model (MITgcm) to simulate the transport and fate of several radionuclides, including tritium and carbon-14, in the global ocean. The Fukushima Daiichi Nuclear Power Station (FDNPS) accident and the subsequent release of treated water are taken as case studies to evaluate large-scale dispersion by ocean circulation and the accumulation of particle-reactive radionuclides in sediments. The objectives and methodology are appropriate, and I agree with the importance of developing models that account for particle adsorption.

However, there are serious issues with the validation of the Fukushima accident simulation. In addition, regarding the release of ALPS-treated water, since the discharge has already begun, the model should be validated against available observational data before conducting long-term projections. In addition, the current spatial resolution is too coarse, and the assumption of instantaneous dispersion within a release grid cell is inappropriate. In reality, no impact of ALPS-treated water has been detected outside the release cell.

For model application, I recommend using cases where abundant observational data exist, such as atmospheric nuclear weapons tests or discharges from European reprocessing plants. Major revisions are therefore required before this manuscript can be considered for publication.

Specific Comments

Methods – Emission

The emission scenario has been summarized in TEPCO’s Environmental Impact Assessment Report (TEPCO, 2023), which reflects the actual implementation of ALPS-treated water releases. This report should be cited, and the simulations should be based on actual discharge scenarios rather than pre-release assumptions.

The TEPCO report is available here, starting from page 264:

https://www.tepco.co.jp/en/hd/newsroom/press/archives/2023/pdf/230220e0101.pdf

Results and Discussion

Line 232: Tsubono et al. (2016) provide a more detailed comparison for this region. Figure 1 is not clear; the authors should refer to Tsubono et al. (2016) and perform a similar validation. Note that Tsubono et al. (2016) used a high-resolution model with relatively good reproducibility of the Kuroshio, whereas the reproducibility in the present model appears poor. The advantages and limitations of the model should be explicitly discussed. Also, Tsubono et al. (2016) applied wider atmospheric deposition fluxes, which should also be referenced.

Figure 1: Since background concentrations of Cs-137 in the North Pacific after 2011 are above 1.0E-3 Bq/L, the color scale used is inappropriate, with the entire comparison range shown as red. The comparison should be made over a range of 1.0E-3 Bq/L to 1 Bq/L.

Figure 3: The contour plots are based solely on model output. In reality, the background concentration of tritium is about 50 Bq/L, meaning that the modeled signals are entirely undetectable. The correct conclusion is that the ALPS-treated water signal cannot be detected. Given that dispersion within the release grid cell is a critical limitation, applying this model to ALPS releases is inappropriate.

Figure 4: Other radionuclides should also be compared against present concentration levels. The MARIS database should be used as a reference: https://maris.iaea.org/home

Figures 5 and 6: As noted above, these figures show concentration levels that are not realistically observable.

Discussion

Line 401: Please refer to TEPCO (2023). The concentrations of all other radionuclides are below discharge limits.

Line 413: The uncertainty of concentration measurements is conservatively evaluated using detection limits (TEPCO, 2023). Furthermore, the IAEA MARIS (2021) does not mention underestimation issues. The citation used here is misleading and academically inappropriate.