Large-scale datasets like CAMELS and Caravan have spearheaded the development of both process-based and data-driven models around the world. This manuscript introduces a CAMELS dataset for catchments over India (albeit only for catchments in the peninsular region of the country due to data licensing issues)

I generally found the manuscript structured, well-written and easy to follow. The paper lies well within the scope of ESSD. I have a few comments about the dataset, which could be addressed before the paper is finally accepted for publication.

Major Comments:

1. ESSD generally encourages the sharing of all relevant processing steps and code required to replicate the results (Carlson and Oda, 2018). This is particularly important for datasets to build user confidence and to adhere to FAIR principle. A flowchart could be added to the Appendix detailing the different products and steps used in generating the dataset. Another possible suggestion is to create a separate repository to deposit all the relevant codes and link them to the data availability statement. Similar processing pipelines are already established for the CAMELS-DE (Dolich, 2024) and CAMELS-CH (https://camels-ch.github.io/)  
2. I went through the Zenodo entry and found that the dataset was previously named CAMELS-IND rather than CAMELS-INDIA. I feel that the former version better aligns with the naming conventions of other CAMELS products. In spirit of ESSD open discussion and for the benefit of future readers, I would like to raise this point so that the authors can reply with their reasoning here.
3. Attribute file naming: This is again linked to my previous comment. Generally, only small letters are preferred in file names as this would aid in automation of code pipelines and other scripts. Hence for example, I would suggest camels_ind_name or camels_india_name instead of
4. I have a minor concern regarding the different zip files for each folder. In general, this makes it more tedious to download and extract each file individually. The total file size seems to be under 1 GB only in any case, it would be worthwhile to consider having a single zip file with subfolders for the entire dataset (similar to the Caravan file structure).
5. I would also recommend adding the license/disclaimer as a text file within the dataset to ensure this is readily available when a user directly downloads the product.
6. Some vital information is missing in Section 7. Consider adding more details (including dataset access DOI) about the specific GLDAS model (Noah/CLM/VIC) and versioning (with or without GRACE- Data Assimilation) used for the preliminary quality assessment. The same can also be added as a dataset citation.
7. In the catchment and station shapefiles, I could find some minor mismatches between the flow outlets and catchment boundaries (For example, Station – 15028 (Thiruvattar)). I think this has already been mentioned in Goteti (2023) and the manuscript. I would again mention this in Section 8 so that future users are aware of possible mismatch issues.

Minor Comments:

I have left a few minor comments on the annotated version of the manuscript. Some are more subjective and personal than others. Feel free to make changes that you feel fit.

Overall, I feel the manuscript could have a moderate revision before it can finally be accepted in ESSD.

Thanks for this timely and valuable contribution!

Regards

Ashish

Carlson, D., Oda, T., 2018. Editorial: Data publication – ESSD goals, practices and recommendations. Earth Syst. Sci. Data 10, 2275–2278. https://doi.org/10.5194/essd-10-2275-2018

Dolich, A., 2024. CAMELS-DE Processing Pipeline. https://doi.org/10.5281/zenodo.12760336

This paper describes the CAMELS-INDIA dataset, which consists of hydrometeorological time series and catchment attributes for a large sample of catchments in India. This will be a valuable dataset for the hydrological community. Overall the paper is well-written, figures are well-produced and dataset well-described.

My major comment is the lack of observed streamflow data for many of the catchments. A key characteristic of CAMELS datasets is providing observed streamflow data – without this, the datasets are less useful for large-sample hydrological analyses. I was surprised that from the 472 catchments included in the dataset nearly a third (32%) of the catchments have no observed streamflow timeseries, and 45% of the catchments have less than 10% of streamflow for the chosen timeperiod.  Why is this? Why were these catchments included in the dataset? How do the authors envisage their use in large-sample hydrological analyses without streamflow?  I appreciate that there are modelled flow timeseries, but the model performance for the other gauging stations was not particularly convincing that these would provide a good representation of streamflow in catchments with no data. My recommendation would be to only include catchments where you have good streamflow data i.e. where you have calculated hydrological signatures (228 catchments – still a very valuable dataset).  If the authors disagree with this recommendation then much better justification and clarity needs to be added into the paper highlighting this limitation with the dataset (see additional comments below).

Aside from the major comment above, I only have minor/moderate comments for the authors to consider before publication:

1. I suggest to follow similar naming conventions to other CAMELS datasets and change the name of the dataset to CAMELS-IND and all files.
2. L50. ‘to some relevant questions’ is very vague – can you be more specific, or perhaps give one or two examples of ‘relevant’ questions?
3. L60. I would not include CAMELS-FR in this list as it is not published yet (as far as I know) and this reference is an EGU conference abstract. The same for CAMELS-DK which is currently a pre-print and not published.
4. L65-75. There are a lot of acronyms in this section. Are they all needed?
5. L90. Why is it ‘around 211 catchment attributes’? If you are providing 211 catchment attributes then you don’t need the word ‘around’ in this sentence.
6. L127. What do you mean by ‘reliable metadata’ – can you be more specific here? What metadata are you considering?
7. Figure 1b. It is really hard to see the basin codes on this map – can you make them bigger or a different colour?
8. L146. I would create a new section here called ‘Hydrological timeseries’ or something similar.Or I would rename section 3 to make it clear to readers that the description of the hydrological timeseries is located here.
9. L146. Can you add some more detail on how the river flow data are compiled for these catchments? Do they undergo any quality assurance or quality control checks before they are published online? Are there flags on any suspect data? Did you perform any checks on the flow data (i.e. for negative values, outliers, multiple consecutive values).
10. L153-154. You need to be much clearer here of the data availability of observed streamflow for the catchments. I would argue that ‘most’ stations do not have reliable data availability. Why do so many of the stations have no streamflow data?
11. L165. This needs more detail on how the gridded rainfall and temperature datasets were produced- is it from observed data (i.e. from rain gauges or weather stations) that are then interpolated on a grid, or from reanalysis data? After reading further in the text, I realise that a lot of this information is in Section 7 but it needs to come earlier in the paper.
12. Figure 4. It would be helpful to add the map here of the mean precipitation (Fig A3) into this plot to give readers an understanding of how much rainfall falls on average across these catchments.
13. L250. Please quantify and provide numbers for ‘Higher mean daily precipitation’, ‘precipitation decreases’, ‘moderate precipitation’, ‘moderate magnitudes are in the central and eastern’, ‘high values’.
14. L262. How many gauges did you calculate hydrological indices for?
15. L296. Out of interest, what causes the high runoff ratios along the southwest coast?
16. L395. How do you define a large and medium dam?
17. L403. Where do we see the reservoir use?
18. I would encourage the authors to add CAMELS-IND to the Caravan dataset to aid efforts in global catchment datasets.