Review of 'A simple weather generator that converts statistical information from downscaled global climate models to 24-hr precipitation input for hydrological models’ by Rasmus E. Benestad

This manuscript is submitted to GMD as a ‘Development and technical paper’. The manuscript describes a simple weather generator (WG) and demonstrates the performance of the WG at a few locations. While the methodology is well documented and the text is straightforward to follow, the following comments must be addressed before this manuscript can be considered for publication.

Major Comments

As per the requirements of the journal - “Development and technical papers usually include a significant amount of evaluation against standard benchmarks, observations, and/or other model output as appropriate”. I am not fully convinced about the performance of the WG. I recommend the author to provide more robust performance evaluation at several other locations.

The text inside the figures, the axis labels, and the legends are too small and often extremely difficult to read. It seems that the figures were prepared in a rush. I recommend the author to improve the figure quality.

Minor Comments

L1: If possible, simplify this sentence.

L3: Mention how many stations are used.

L4, L6: Explain ‘close approximation’ and ‘reasonable skill level’ with a number or error metric.

L6: Consider removing this sentence.

L1-7: The abstract seems more like a plain language summary. I recommend the author to think carefully about the abstract.

L20: Please add a couple of citations.

L23: Please add a couple of relevant citations, to which a new reader can refer to if they want more information about downscaling.

L39: Please provide a sentence describing why previous weather generators were not designed to connect with ESD.

L40: Provide full name of the SPRINGS project. Additionally, the relevance of the SPRINGS project is not clear.

L45: Mentioning names of a couple of relevant and popular hydrological models might be useful.

L55: I recommend adding one sentence better describing the cut-off threshold.

L69: A sentence or two describing the equation would be helpful.

L75: The choice of default smudge factor should be explained.

L93, 94, 102: ‘good agreement’, ‘modest deviations’, ‘comparable to the observations’ - These qualitative descriptions should be avoided and the author must provide quantitative metrics for evaluation of the Weather Generator. 

Figure1: The schematic could be improved. In my opinion, the description provided in the caption is not properly reflected in the figure.

Figure2: It is not clear what the colors represent.

General comments:

The manuscript presents a simple weather generator designed to translate downscaled annual precipitation statistics, mainly wet-day frequency and wet-day mean precipitation, into daily precipitation sequences for hydrological and impact models. The topic is relevant because many impact models require daily inputs. The implementation in the open-source 'esd' R package is useful and improves reproducibility.

However, the current manuscript would benefit from stronger evaluation and clearer framing of the method’s limitations. The proposed weather generator reproduces some marginal precipitation statistics reasonably well, but it does not yet adequately reproduce wet-spell and dry-spell duration, which are important for hydrological applications, flood generation, soil moisture, and drought-related impacts. The manuscript should more clearly explain when this simple generator is appropriate and when a more advanced weather generator may be needed.

Major comments:

First, the evaluation should be strengthened beyond visual comparison. Most results are shown as QQ plots and time-series comparisons, but the manuscript would benefit from quantitative metrics for wet-day frequency, wet-day mean precipitation, annual totals, extremes, and wet/dry spell durations. These metrics would help readers assess performance more objectively.

Second, the underestimation of wet-spell and dry-spell duration is an important limitation. The author notes that the WG tends to generate too many short wet and dry spells, but this issue should be discussed more centrally. Spell persistence strongly affects hydrological model outputs, especially runoff generation, antecedent moisture, flood peaks, and drought duration. The manuscript should either improve the method to better preserve spell statistics or clearly state that the current WG is less suitable for applications that depend strongly on precipitation persistence.

Finally, the temperature generator is presented rather briefly. The precipitation WG is the main focus, but the manuscript also reports reasonable skill for daily temperature. The temperature section should either be expanded with more complete evaluation across sites and seasons or shortened and framed as a preliminary extension.

Specific comments:

1.    Please consider evaluating the WG at more sites or across a broader range of climate regimes to better demonstrate transferability.

2.    Please add quantitative performance metrics to complement the figures.

3.    Please discuss the lack of spatial dependence among multiple stations. This may limit applications to basin-scale hydrological modeling.

4.    Please discuss the lack of dependence between precipitation and temperature. This may matter for snow and evapotranspiration applications.

5.    Some figures are difficult to read because labels and legends are small. Please improve figure readability.

6.    The conclusion should include a concise limitations paragraph, such as spell-duration bias, stationary seasonality, single-site design, lack of spatial correlation, and limited temperature evaluation.