The authors construct a spatially resolved dataset covering five major river basins in Central Europe, integrating dynamic meteorological variables and 46 static physiographic attributes onto a unified grid. The dataset provides a relatively comprehensive foundation for rainfall–runoff modeling and has potential value for distributed hydrological modeling as well as machine learning applications. However, according to ESSD’s standards for data papers, the manuscript still exhibits several important issues regarding novelty, articulation of scientific contribution, and methodological detail that require further improvement.

1. The manuscript aggregates variables from multiple publicly available data sources and harmonizes them onto the ERA5-Land grid. However, it does not clarify whether this integration process involves any methodological innovation, nor does it explain what added value the dataset provides compared with researchers independently processing the original data themselves. Without such clarification, the dataset’s novelty appears limited.
2. The rationale for selecting the 6 spatiotemporal (“dynamic”) meteorological variables and 46 static (“ancillary”) attributes is insufficiently justified. Although the authors state that the variable selection is based on prior studies, the reasoning remains unclear from a reader’s perspective. For a data paper, where the choice of variables should be supported by scientific or functional justification. Providing the variable lists used in previous datasets and clarifying whether the present study includes all commonly used variables or only a subset would help readers better understand the design and scope of the dataset.
3. The manuscript mentions “reprojecting and subsampling at the locations of the nodes in this grid,” but does not provide methodological details. For a dataset paper, the resampling and reprojection procedures should be described, including the specific interpolation or sampling methods used. Without this information, the processing steps are not sufficiently transparent or reproducible.
4. The manuscript does not address several essential aspects related to data quality and reliability. It does not discuss whether resampling introduces information loss, whether the gridding process may generate boundary effects, or whether any variables contain missing values and how such cases are handled. These issues are fundamental for a data paper, as they allow users to assess the reliability of the dataset for their applications.
5. The conclusion section is brief and lacks a comprehensive synthesis expected of a dataset paper. A proper conclusion should summarize the dataset’s scientific contributions, outline the types of research it can support, and explicitly discuss its limitations.
6. The manuscript frequently emphasizes the dataset’s applicability for neural network–based hydrological modeling, yet the Introduction does not sufficiently cite relevant literature or explain the research gap. the authors should include supporting references and more clearly articulate how this dataset connects to and advances existing machine learning hydrology research.

The manuscript presents a multivariate spatiotemporal dataset of weather forcings and catchment characteristics within a large area in Central Europe. It comprises 5 meteorological variables and 45 static catchment features on a regular grid. The manuscript is concise, well-structured and gives a good overview of the dataset, however it requires some clarifications.

The dataset is a compilation of several data sources that are quite well-known and acclaimed in the international hydrological community, however bringing this data together undoubtedly was an effort. At the same time, these sources of data ensure consistency and universality for the resulting dataset under review a priori. It was already mentioned in the above comment, that the regridding procedures are not documented in the manuscript, hence its’ consistency still needs validation.

The authors state that the dataset is more suitable for distributed hydrological models’ testing rather than CARAVAN dataset since it’s not lumped. While this is obviously true, some uncertainty is still possible originating in interpolation from the variables’ sources and eventually in interpolation on the particular model’s grid.

The name of the manuscript refers to rainfall streamflow modelling, however the domain and especially its’ southernmost part of is located in snowmelt runoff area. The authors are advised to address this issue, since no snow-related data is given in the dataset.