General Comments

This manuscript presents the first global high-resolution (10 m) LCC product, derived from Sentinel-2 data using a CSI-based empirical regression approach. It represents a significant advancement in Earth observation and global vegetation monitoring, addressing a critical gap in existing LCC datasets, which have previously been limited in spatial resolution and geographic coverage.. The methodology is clearly described and validated using ground measurements across diverse vegetation types and global locations. I tested the authors’ web application. It is user-friendly and customizable, making it a valuable tool for researchers to access high-resolution LCC data tailored to their needs. However, a few clarifications and minor revisions would further strengthen the manuscript, including more precise descriptions of data/method, explanations for validation limitations, and some language refinements. Overall, the study is well-executed and highly relevant to ESSD, and I recommend minor revision before acceptance.

Specific Comments

Introduction/Data: A summary table comparing existing global LCC products (e.g., spatial/temporal resolution, method, etc.) would provide helpful context for readers.

L48-50: Authors should add a short explanation of how CSI’s insensitivity to LAI and soil reflectance ensures consistent performance across global vegetation types。

L150–152: Please clarify the accuracy of the regression equations (e.g., R², RMSE) and indicate whether they are supported by existing literature or developed in this study.

Table 2: Most CSI-based regression equations are reported to achieve RMSEs below 10 μg/cm2, but the overall RMSE of the product exceeds 15 μg/cm2. Please explain the reasons for this discrepancy.

L233：The link is missing, please add the relevant link.

In Figure 4, the needleleaf forest shows small bias (0.95), but Table 3 reports a clear underestimation across multiple NF sites. Please clarify the discrepancy between these two results.

Figure 9: It would be helpful to include the global MODIS LCC map for the same periods.

Figure 9a, 10a: There is a noticeable lack of data in high-latitude regions of the Northern Hemisphere. Please explain the reason.

Line ~350: The detailed description of RMSE values across different sites is more appropriately placed in the Results section rather than in the Discussion.

Please check the consistency of the term "bias" throughout the manuscript, as the capitalization (e.g., "Bias" vs. "bias") is currently inconsistent.

L365-370: Expand briefly on seasonal or regional limitations of the current cloud masking method, and mention potential improvements or alternatives.

Other Suggestion:

1. I understand that using relatively simple and efficient methods (e.g., empirical regression) is practical for generating global high-resolution products. However, the validation accuracy at certain sites—particularly for needleleaf forests—is relatively low. In future work or within the web APP, would it be feasible to apply region-specific or improved empirical relationships to enhance accuracy?
2. As mentioned in the manuscript, future efforts should focus on collecting more ground validation data from the Southern Hemisphere to improve the global representativeness of the product assessment.

The manuscript presents the first global 10 m-resolution leaf chlorophyll content (LCC) product derived from Sentinel-2 data using the chlorophyll-sensitive index (CSI). While the effort to produce a high-resolution LCC product is commendable and addresses a gap in existing datasets, the study suffers from critical methodological and validation shortcomings that significantly decrease the product's reliability. Below are my major concerns:

1. The validation results reveal severe limitations in accuracy, particularly for non-cropland vegetation. The 10 m-resolution product achieves an R² of 0.27, dropping to R² = 0.11 for the 100 m-resolution. Such low explanatory power—especially for forests (R² = 0.05–0.22)—raises fundamental questions about the parameter retrieval method’s robustness compared to existing approaches.
2. Methodologically, the CSI-based approach is insufficiently described. While CSI is claimed to be insensitive to LAI and soil reflectance, no evidence supports its stability across diverse biomes, seasons, or terrain (e.g., topographic shadows). At 10m resolution, terrain and terrain-induced shadows are significant but entirely unanalyzed. Key steps, such as downscaling 20m Sentinel-2 red-edge bands to 10m, are omitted, leaving unresolved uncertainties about artifacts introduced during resolution enhancement.
3. Looking into the product, Figure 9 exhibits severe striping patterns (adjacent tiles differing by >20 μg/cm²), indicating unresolved processing artifacts. The absence of uncertainty metrics (e.g., per-pixel confidence intervals) leaves users unable to assess data reliability. Implausibly low LCC values in tropical rainforests (<25 μg/cm²) contradict ecological baselines, suggesting algorithmic biases. Expanding validation in tropical evergreen forests is definitely needed.
4. Severe data gaps in tropical regions (Figure 10) and seasonal missing rates (e.g., winter composites) limit completeness. The authors provide no quantitative coverage statistics or strategies to mitigate gaps (e.g., multi-year composites).
5. Placeholders like “link:xxxx” (L235) and undefined terms (e.g., “Ref” in L45) reflect incomplete drafting. Figure 8 lacks a legend to differentiate MuSyQ and MODIS data points. Grammatical errors (e.g., “thanthe” in L234, duplicated text in L307) indicate rushed preparation.

While high-resolution LCC products are needed, considering the issues above, I apologize that I cannot be supportive here. I suggest the authors: (1) expand validation across seasons and at tropical forests; (2) quantify uncertainty sources (e.g., downscaling, terrain effects); (3) resolve technical artifacts and data gaps. Nevertheless, I applaud the authors’ effort to make their data publicly available.