1. This paper develops a long sequence dataset, and the research purpose and user target population of this dataset need to be further clarified;
2. “The objective of this study is to develop a high-resolution and accuracy-assessed dataset of daily near-surface meteorological variables across mainland China, suitable for applications in hydrological modeling, environmental monitoring, and climate analysis.” How did the author consider the issue of temporal "homogeneity" in a long series dataset used for climate analysis?
3. What is the quality of the raw observation data used to establish 1km grid data? Has the author conducted data quality evaluation, analysis, quality control, homogenization processing, etc. on the original observation data during its use?
4. Compared to a grid spatial resolution of 1km * 1km, using over 2000 observation data from China is relatively insufficient, especially in the sparse observation areas of western China. How does the author consider this issue?
5. How is the " day boundary issues" handled? The ground meteorological observation in China adopts "20:00 Beijing time" as the boundary point of the day, which means that the observation day is from 20:00 to 20:00 the next day. This standard is applicable to daily value statistics of factors such as precipitation and temperature. Prior to the 1980s, some stations had a phenomenon of inconsistent day boundaries (such as a few stations using 08 or local time), which led to a decrease in comparability between early data and other stations.
6. Overall evaluation: This long sequence dataset did not take into account the quality of the observation data used, day boundary issues, uniformity issues, etc. during the development process. Therefore, the dataset reconstructed in this article also has day boundary and uniformity issues, which will have a serious impact on downstream user research.

This paper constructs a long-term meteorological variable dataset by decoding the nonlinear relationships between six meteorological variables and their spatial covariates. The method is innovative and the dataset is usable, but the paper needs to be revised based on the following points.

1. Abbreviations such as “CC” in the abstract should be spelled out in full.
2. The abstract does not adequately reflect the research objectives and significance of the study and needs to be improved.
3. The first paragraph of the introduction should provide some supporting citations.
4. I noticed that the “Materials” section includes observation sites from different sources, but the verification sites were only selected from CMA. Have you considered selecting verification sites based on the weight of the number of sites from different sources?
5. How to consider the cumulative error caused by progressively inputted meteorological variables during the modeling process, especially in the sunshine duration model.
6. How can authors reduce errors caused by the boundaries of the study area during modeling, given that these areas have fewer observation stations?
7. I noticed that Figure 4 contains a large amount of information, but the poor resolution and color quality make it difficult to see clearly. Please improve this.
8. Please indicate whether adjustments were made when the author encountered situations where the sunshine duration was less than 0 during the verification.
9. The author mentioned the limitations of satellite remote sensing in estimating the meteorological variables in the “introduction”. However, sunshine duration is greatly affected by cloud and aerosol parameters observed by remote sensing. When comparing sunshine duration products, please consider comparing sunshine duration datasets estimated based on remote sensing data (https://doi.org/10.5194/essd-17-1427-2025) and explain the advantages of the research method used in this study.
10. Is the model independent on a daily scale? Did the authors consider modeling based on different days of year (DOYs) to enhance the model's generalization ability in the future?

1. Regarding the selection of only 2 out of 12 field stations from the hourly land–atmosphere interaction dataset (Ma et al., 2024) for validation after quality control, it is unclear why the other stations were excluded. Please clarify the quality control criteria and explain whether the other stations were omitted due to poor data quality or other reasons.
2. The quality of the figures is suboptimal. Several figures lack units, and the x- and y-axis labels are missing or unclear. For instance, Figure 2 has low color contrast, making it difficult to distinguish between different elements. Improvements in figure clarity and completeness are necessary.
3. The representativeness of the station data at the 1 km grid scale needs further discussion. Please elaborate on how the spatial representativeness of point stations affects the validation results, especially in regions with complex topography or sparse station coverage.
4. The reconstructed sunshine duration product does not show significant advantages over existing datasets. Concerns remain regarding data consistency, likely due to instrument changes and automation upgrades in CMA sunshine duration observations over time. This issue should be addressed to ensure reliability.
5. The comparison with the CMFD product may be unfair, as CMFD utilizes a much smaller number of meteorological stations than this study. This discrepancy in input data may bias the comparison results. The authors should acknowledge this limitation and discuss its potential impact.
6. The long-term trends and homogeneity of the reconstructed dataset are not discussed. An analysis of the temporal consistency and homogeneity of the data—especially concerning non-climatic factors such as station relocations or instrument changes—would strengthen the dataset's credibility.