I have reviewed the paper "AltiMaP: Altimetry Mapping Procedure for Hydrography Data" by Revel et al. (2023), which introduces a method for allocating virtual stations (VSs) from satellite altimetry data to a river network. This method can improve the accuracy of comparing simulated water surface elevations (WSEs) using satellite altimetry data with those obtained from hydrodynamic models. I find this paper to be well written and recommend it for publication after addressing the following comments and suggestions.

Line 14: The authors have used numbers 10, 20, 30, and 40 to represent different flags, but it would be helpful if they could provide a more detailed explanation of their reasoning. Why not 1, 2,3, and 4.

Line 37: The authors should provide statistical measures such as mean absolute error or any to define what they consider to be reasonable accuracy. Please avoid using qualitative words.

Lines 42-45: To improve the readability of the paper, it would be helpful if the authors provide separate references for each different radar altimetry mission instead of listing all the references together. This will make it easier for readers to identify and access the specific sources of information relevant to each mission.

Table 1: The authors could add an additional column discussing how these different data sources differ from each other in their data generation algorithms.

Line 46: The authors should provide a brief explanation of how the temporal resolution and inter-track distance of satellite altimetry data affect the temporal and spatial resolution of the data.

Lines 70-77: The authors should provide a more detailed discussion of existing studies that have attempted to accurately locate VSs, rather than only discussing the problem. By doing so, they can highlight the research gap that their work aims to address and demonstrate how their method contributes to the current state of research on this topic. Improving the research gap in detail will help readers better understand the significance of the authors' contribution and appreciate the originality of their approach.

Line 117: The authors should provide a more detailed explanation of their data selection criteria, such as period and temporal resolution, and explain why they chose to use satellite altimetry data from HydroWeb instead of other sources.

Lines 118-119: The authors could improve the readability by moving the discussion about identifying and removing biased VSs to Section 2.4 and providing a more detailed explanation of their criteria for identifying VSs.

Line 120: The authors could include a brief discussion of each data source used in the study and how the data were derived.

Lines 144-145: The authors should elaborate on how river bathymetry and river bank height cause deviations.

Section 3.1: The section lacks sufficient discussion.

Figure 4: The authors should improve the overall quality of the figures, and in the density distribution plot, they should change the color code to make the density distribution line for Flag 10 visible.

Figure 5: The authors should improve the overall quality of the figures. The y-axis level is missing, and there is overlap between Figures 5a and 5b.

Line 228: The authors should explain why they compared the evaluated results in terms of RMSE? please consider showing the correlation coefficient and bias as well.

Line 229: The authors should explain why the elevation causes an increase in RMSE.

Line 230: The authors should rewrite the sentence to clarify that there is no change in RMSE before a certain threshold (<200 m/km) and that the medium of RMSE increases from 2 to 4 m, not just RMSE.

Line 265: The authors should explain why?

Figure 7: The authors should discuss the expert method and ordinary method in the text to help readers better understand the differences between them. Alternatively, they could use constant terms to avoid confusion between the traditional method and ordinary method terms.

Figure 8: The authors should address the same comment as Figure 7. Additionally, they should move Figure 8 from outside of Table 3 to improve the organization of the paper.

This study presents a method called an automated altimetry mapping procedure (AltiMaP) that allocates altimetry virtual stations (VS) to the Multi-Error Removed Improved Terrain Hydrography (MERIT Hydro) river network. Although this study shows an improvement over the traditional method of allocating VS to the coarse-resolution river network, I have some questions and suggestions to the authors.

1. First of all, the authors are using the simulated WSEs from the CaMa-Flood model to be compared with the WSEs from altimetry with AltiMaP. I am not sure how this approach (i.e., using simulated WSEs as a reference to evaluate observed WSEs) can be convincing, especially considering the fact that the CaMa-Flood is a global hydrodynamic model calibrated/validated with available in-situ network (I assume so) which are sparse in large river basins, such as Amazon, Congo and Mekong. In addition, how AltiMaP assigned altimetry-observed WSEs can be used in the future for better calibration/validation of the CaMa-Flood model (this could be added in the summary section or in a separate discussion section)?

Can authors elaborate on this?

1. Line 141: how is this assumption valid? It is not certain whether the observed WSE time series available from the HydroWeb is really the one over the floodplain (which could be flag 20 or 30), or over the open river channel. If the HydroWeb time series are indeed for the floodplain, AltiMaP may be erroneously assigning the VS to the open river channel. Can authors elaborate on this?

1. It is mentioned that mean observed WSEs are used to be compared with MERIT DEM elevation. But it is not explained how “mean” has been obtained. Did the authors simply take the mean of the entire WSE time series? Or did the authors consider the water cycle of the basins? If the entire time series has been simply used to compute the means, that will lead to an inherent bias due to the seasonality of WSE changes. Please clarify.

1. Figure 7: Even using AltiMaP, I see majority of the VSs have high RMSEs over the world. This demonstrates that basically HydroWeb WSEs and CaMa-Flood WSEs are not comparable. There are many factors behind this (as authors mentioned them), but I think the authors should not simply use the time series from HydroWeb without quality check. I’m not saying HydroWeb data is inaccurate, but I’m saying some of their time series may be inaccurate because of the inherent limitation of altimetry over land.

 Minor comment:

• Abstract: “much lower (10.6%)” is a bit of exaggeration in my opinion. I would say “a meaningful improvement” or something like that.

General comments

This study develops an altimetry mapping approach (AltiMaP) that aims to mitigate the mismatches between virtual station (VS) locations and actual river locations, which are caused by DEM errors, the use of discrete river grids, and the allocation of VSs to the center of the WSE observation search area. The topic is interesting to the hydrological community. However, there are many issues unsolved with the manuscript in its present form and I recommend rejection (see details below).

1)         I think one of the major limitations of the study is that the allocation was performed based on Hydroweb whose VSs are located away from the actual river. What is the added value of the method for self-defined VSs or other data sets when the VSs are delineated right at the center of the river?

2)         If the authors focus on satellite radar altimetry, the swath interferometric altimetry mission CryoSat-2 with dense spatial coverage should be added as an important data source for validating the method. Further, how applicable is your method to laser altimetry (ICESat-1/2)? I would provide a preliminary result for these data with a short discussion.

3)         Line 140: You are selecting the largest river for further processing. But it is possible that the observation (also termed POCA, point of closest approach) is from the river closest to the satellite (within the beam limited footprint) when there are multiple river channels near the VS location. Therefore, it would be interesting to perform a similar analysis with the abovementioned assumption (i.e., choosing the closest river as opposed to the largest river to derive WSE) to see the difference.

4)         Line 148: Many previous studies have confirmed the reliability of the median value compared with the mean, which is quite sensitive to outliers. I would suggest the authors use the median value as the final WSE and update all the relevant results.

5)         Do you use lat and lon at nadir for the allocation of the altimetry measurements? If so, I guess you may need to use the corrected ones instead (lat_cor and lon_cor), which are better representations of the radar echoes.

6)         While range correction derived from waveform retracking is not within the scope of the manuscript, it is still one of the major sources of error for WSE. The introduction section should at least mention this. It would make more sense to briefly introduce the processing chain of Hydroweb (e.g., what retracker and/or slope correction method it is using), followed by a citation, such that authors without expertise in altimetry could better capture the contribution of the methodology.

Minor comments

1)       Line 40: ‘following troposphere’-> following dry troposphere?

2)       Line 163: what is the timestamp of the MERIT DEM? Because your altimetry data cover a wide range of time periods (1992–2022), how could you confirm the MERIT DEM is representative of the actual topography that is validated against the altimetry missions?

3)       Line 184: ‘then adding 100 to the flag of any VS that is biased’, please explain or reword. Why not add a fifth flag for biased VSs?

4)       Table 2: how to obtain the river widths, manually?

5)       Line 202: ‘river channel river’. A typo here?

6)       Figure 4: please increase the font size of the figure

7)       Figure 5: please add a title for the y-axis in b,c, and d