Please refer to the attached pdf.

General comments：

This manuscript presents an improved Coupled Carbon and Water (CCW) model incorporating dynamic water use efficiency (WUE) to disentangle the effects of climate change, vegetation dynamics, and atmospheric CO₂ on water yield (WY) across China during 1982–2017. The study addresses an important research gap by explicitly accounting for CO₂ effects and providing a robust attribution analysis at both national and regional scales. The integration of scenario-based attribution and elasticity analysis is innovative and valuable for water resource management and climate adaptation strategies.

The study is generally well-structured, with clear objectives, methods, and results. However, there are a few areas where further clarity, elaboration, and enhancement could improve the overall impact and rigor of the paper. I provide specific comments and suggestions below:

Specific Comments

1. Abstract:

* The abstract effectively summarizes the study, but it is rather dense with technical terms and numerical results. Consider slightly rebalancing it by adding one or two sentences that emphasize the practical implications (e.g., relevance for water resource management and ecological restoration) so that non-specialist readers can more easily grasp the significance.

1. Introduction and Background:

* While the introduction clearly outlines the motivation, it would be useful to more explicitly highlight what distinguishes this study from other model applications. For example, a brief comparison with previous studies could emphasize the novelty of incorporating CO₂-induced WUE changes.

* The introduction mainly focuses on China, but since similar issues of climate–vegetation–CO₂ interactions exist globally, it may help to briefly situate this work in a broader international context. For example, mentioning comparable studies in other semi-arid or monsoon-influenced regions would show the wider relevance of the improved CCW model.

1. Methods:

* The improved CCW model assumes interception evaporation factor fi equals zero. Since this simplification is acknowledged in the Discussion, please provide a short justification earlier in the Methods section so that readers can immediately understand this limitation.

* The attribution analysis is based on “trends” in WY under different scenarios, but the exact method of calculating these trends (e.g., linear regression, Mann–Kendall test, or another approach) is not clearly described. Providing details on the trend detection method, as well as the statistical significance criteria, would help readers better assess the robustness of the results.

1. Results:

* Figures 5 and 6 demonstrate spatial heterogeneity in WY drivers. It would help if the authors could provide a more policy-relevant interpretation, e.g., what the findings imply for water resource planning in regions where vegetation dominates versus where climate dominates.

* The elasticity analysis provides valuable insights into the sensitivity of WY to different drivers.  However, the discussion could be enhanced by more explicitly linking the elasticity results to the relative contributions of each driver.  For instance, why does CO₂ have a higher elasticity than NDVI yet a smaller overall contribution?  Clarifying how elasticity and the magnitude of change jointly determine the net impact would strengthen the interpretation.

1. Discussion:

* While the study focuses on climate, vegetation, and CO₂ drivers, other human activities such as reservoir regulation, irrigation, and groundwater extraction can also significantly affect water yield in China. Since these processes are briefly mentioned as limitations, it would strengthen the discussion if the authors could add a short paragraph acknowledging how such anthropogenic factors may interact with the modeled drivers, and whether the improved CCW framework could potentially incorporate them in future work.