This manuscript presents a major and valuable update to the CAMELS-GB large-sample hydrological dataset. The inclusion of extended daily records, national-scale hourly rainfall and flow data, groundwater level time series, dynamic land-cover attributes, and enhanced hydrometric and human-influence metadata represents a substantial contribution to the hydrological and Earth system science communities. The manuscript is well written and clearly structured. The dataset is openly available, well documented, and likely to see wide uptake. I recommend acceptance subject to minor revisions.

Major comments

1. Visibility of key methodological details

Several important methodological aspects are described in the Supplementary Information, including propagation and prioritisation of quality-control flags; identification and documentation of suspected outliers and datum changes in groundwater level time series; inter-comparisons between meteorological products at daily and hourly scales. Some of this information is essential for correct dataset usage and could be more explicitly signposted in the main manuscript, to ensure users are aware of these important safeguards and limitations without needing to discover them independently.

2. Guidance on choice between alternative data products

The dataset provides multiple alternative products for rainfall and PET (e.g. CEH-GEAR vs HadUK-Grid; CEH-GEAR1hr vs GRaD-GB(1H1K)), which is a clear strength. The manuscript and Supplement provide quantitative comparisons between these products. But the manuscript would benefit from more explicit user guidance on dataset selection. For example:

• Which products are recommended for long-term trend analysis versus event-based or flood-focused studies?
• Are any products preferred for hydrological modelling consistency with derived attributes?
• Are there specific applications for which users should exercise caution due to known biases or structural differences?

3. Interpretation of extrapolated peak flows and discharge uncertainty

The manuscript explains (L354-357) that discharge uncertainties were recalculated using longer streamflow time series but the same methodological framework as Coxon et al. (2015). However, it remains unclear why the uncertainty estimates are described in L369 as being based on an ‘older set of gaugings’, and why newer gauging information was not incorporated. Please clarify this distinction and its implications for peak-flow analyses.

Please clarify how the percentage of time exceeding the maximum gauged flow is calculated (e.g. proportion of valid daily/hourly timesteps) in Fig5.

Minor Comments

L111 " Mean daily averages calculated from 1st October 1970 – 30th September 2022 for the 671 CAMELS-GB catchments." , ‘mean’ and ‘average’ are synonymous. Do the authors mean daily averages that were then averaged again? The sentence also misses a verb.

L323 “increasing land cover over time” should be “increasing urban land cover over time”

L369-371: The sentence is difficult to follow. I suggest clarifying that the uncertainty bounds are derived from an older gauging dataset and may not fully capture uncertainty at extreme flows.

L386-387: The sentence structure seems odd and needs rephrasing.

L444-445: It states that ten catchments have a normalised upstream capacity greater than 0.25. However, based on the released file “camels_gb_v2_humaninfluence_attributes.csv”, only nine catchments >0.25. Please check and correct this number to ensure consistency between the manuscript and the dataset. The threshold of 0.25 used to highlight catchments with large normalised upstream capacity is not explained. Please justify the choice of 0.25.

Thank you for your work in compiling CAMELS-GB v2 and this manuscript. The data is very nicely formatted and the addition of hourly hydro-meteorological timeseries data is extremely valuable to expand the field of large-sample hydrology to sub-daily resolutions. The addition of groundwater data and the extensive quality control of hourly streamflow data is also very valuable and outstanding among other CAMELS datasets. The extension of the daily timeseries length and the expansion of the catchment attributes, especially the changing land cover attributes, are also of great value and show the author's motivation and dedication to maintain and update CAMELS-GB.

The manuscript gives a very detailed and thorough description of the dataset while highlighting its importance and also describing its limitations. I recommend accepting the manuscript after minor revisions.

I think the accessibility of the dataset could be enhanced by offering the option to easily download a compressed file containing the entire dataset. A more detailed explanation of this as well as minor comments and technical corrections can be found in the attached PDF file.

This manuscript documents the second version of the CAMELS-GB dataset, which represents a significant expansion relative to version 1. The addition of extended daily records, sub-daily hydro-meteorological time series, groundwater level observations, and time-varying catchment attributes considerably enhances the dataset’s value for a wide range of hydrological analyses. The manuscript is clearly organized and well written, and the resulting dataset will be a highly useful resource.

I recommend publication subject to minor revisions. My comments relate both to the manuscript text and to the dataset itself, and focus primarily on points of clarification, documentation, and consistency, as well as optional enhancements such as the provision of a ready-to-use pre-filtered hourly streamflow product, that would further improve usability. See comments in the attached PDF file.

Many thanks for creating the new version of the CAMELS-GB v2 dataset. Some suggestions for further improvement are listed in the appendix.