Title: Shelf Bathymetric Roughness Controls Mixing of the Persian/Arabian Gulf Outflow and Arabian Sea OMZ Ventilation Author(s): Iteration: Initial submission

The manuscript is based on a dataset of 50 CTD casts collected recently in July 2025. The major result is presented in Figure 2, which nicely illustrates the depth stabilization and the erosion of the core of the P/AG outflow in terms of salinity, temperature, dissolved oxygen, pH, and turbidity, as it progresses through the Sea of Oman, along the continental shelf. Figure 3 complements Figure 2 by depicting the properties of the bottom waters.

Despite the highly informative nature of these two figures, the study struggles to provide convincing dynamic elements to better understand the origin of the property erosion (as suggested by the article’s title). The study also suffers from the absence of current measurements (LADCP and ship-ADCP), which would provide additional dynamic insights into the flow, such as hydraulic constraints—mentioned but not quantified, for example, with a Froude number—or the intensity of shear, which is also cited but not quantified.

The manuscript title focuses on the control of the outflow by the topographic roughness. One would therefore expect an article explaining the dynamics of the flow along the slope. However, after reading, the only diagnostic addressing this aspect concerns the parameter h_r/H, where h_r is a roughness height and H is the thickness of the outflow. It is unclear how h_r is precisely calculated; I assume that, here, it simply refers to the height of the submarine micro-relief that dots certain parts of the continental slope (if so, based on which bathymetric datasets (multibeam?), at what resolution, what are the limitations, what about the horizontal length scales, …?). As in many other outflows, the gravity-driven flow accelerating along the first significant slopes (here, depths > 170 m) is likely the primary source of erosion (Froude > 1, strong current shear at outflow/ambient fluid, instabilities and mixing). This deep-ward flow is quickly counteracted by the Earth’s rotation on one hand and bottom friction on the other. No element in this article describe this dynamics and we don’t know whether the selected path for Fig. 2 section (white line on Fig. 1) is along the outflow pathway or if it is more or less across its pathway. We don’t even know how rough the bathymetry is (vertical scales, horizontal scales, isolated vs continuous roughness, separation distance between isolated small topographic bumps, …) while such parameters are among the key parameters for inferring the relationship between roughness and mixing (e.g. Özgökmen and Fischer, 2008, as referenced in the text).

The h_r/H parameter is used merely used to suggest that mixing is possible due to the presence, in certain parts of the region, of small relief features where h_r/H > 1, which could potentially generate form drag and wake turbulence. Fair enough, but no quantitative evidence confirms an increase in turbulence in these areas. The article mentions the occurrence of local instabilities in vertical density profiles and portions of profiles that are homogeneous. Perhaps the authors could use this as a starting point to diagnose mixing more precisely and determine whether profiles with strong local instabilities correspond to areas where h_r/H > 1 (or downstream of these obstacles) ? The inclusion of historical datasets with current measurements should have been part of the study to complement and improve the description and quantification of flow dynamics and turbulence, thereby aligning more closely with the title.

Conclusion: The article primarily describes the erosion of the hydrological characteristics of the P/AG outflow plume and its detachment from the topography during its depth stabilization phase. On these descriptive aspects, the article is well-written and easy to read. Some points need clarification. It is also well-supported by appropriate references to past studies. Nevertheless, it fails to address the essential questions implied by the title: If we were to establish a hierarchy of the most important erosion processes referenced in the study, which are the most significant?

• Hydraulic constraints due to bathymetry (Froude > 1, strong shear)?
• The localized presence of small submarine reliefs (h_r/H > 1) associated with turbulence?
• Double-diffusive phenomena?
• Intrusions by mesoscale processes and slope current instability?
• ...

To answer these questions, the study must be taken further by using the CTD profiles more extensively, incorporating historical data, or leveraging numerical model outputs. In its current version the manuscript falls short of the expectations from the attractive title. It is very descriptive but lacks of tangible quantification of key parameters.

Note: Additional comments are provided in the commented manuscript.

This is an observations based paper addressing the detachment of the P/AG water from the continental shelf into the water column as it encounters a change in the bathymetric slope, in the Gulf of Oman. The paper is well written, the figures are of very good quality and clearly show the detachement and the mixing of P/AG. In particular, the PAG water fraction (Fig 2h) clearly summarizes the findings : the P/AG flows as a very shallow bottom current as long as the shelf is no deeper than about 100m. After a rapid drop in the bathymetry, from 100m to about 300m, the P/AG current detaches from the bottom, entrains surrounding water and thickens. The connection with the detachment and the benthic life is very instructive.

My only concern is on the authors claim that the reason of the detachment is the bathymetric roughness. I find this statement insufficiently supported. The detachment is also concomitant with the slope increase, which could be another cause of detachment. At some point the authors seem to suggest it too, l.162 “the abrupt bathymetric shift serves as the initial trigger for flow transformation”. To me, the paper would be as valuable without this claim of roughness as the controller. This would require to change the title. By the way, from Fig2, it seems that the roughness is not due to small hills but to narrow deep cracks. I believe this makes a difference, especially in terms of form drag, a process invoked by the authors, but also in terms of boundary layer stability.

Minor remark.

Fig 3d : the colorbar for the bathymetry seems inconsistent with Fig1 and Fig2 : e.g. mooring #5 seems to go as deep as 400m (Fig 1 and 2) but looks to go only down to ~-150m on Fig. 3.