This study focuses on the Qinghai Lake basin in northeastern Tibet, employing comprehensive data from stable water isotopes (δ²H, δ¹⁸O, lc-excess), soil moisture content (SWC), and the MixSIAR mixing model to systematically characterize the dynamic changes in recharge sources and pathways of shallow groundwater during different stages of the freeze-thaw cycle (melting phase, post-melting period, and freezing phase). The research aligns closely with current cutting-edge topics in hydrology and permafrost degradation under climate warming in high-altitude regions, featuring well-defined scientific questions, rigorous experimental design, ample data, and appropriate methodological applications. These findings hold significant scientific value for understanding groundwater formation mechanisms and water resource management in high-altitude regions under climate change. However, while fully acknowledging its scientific contributions, I identify certain shortcomings in the manuscript regarding the rigor of its argumentation logic, the depth of interpretation for certain data, and the completeness of methodological descriptions. To further enhance the paper's quality, I recommend a comprehensive revision. Below are the key issues and specific recommendations for improvement, for the author's consideration.

Major Comments：

1. The conclusion that soil water serves as a "transitional reservoir" lacks sufficient support. The author repeatedly emphasizes in the abstract and conclusions that "soil water is a key transitional reservoir linking precipitation/snowmelt water with groundwater recharge" (Lines 38–40,511–513). However, the MixSIAR model only provides contribution ratios for each element and cannot directly demonstrate that soil water fulfills a transitional role of receiving, storing, and releasing water over time. To substantiate this argument, it is recommended to supplement the analysis with data on the lagged variation of soil water δ¹⁸O over time (e.g., the lag time in soil water isotopic responses following precipitation events). If such evidence is unavailable, the author should downplay this aspect in the manuscript. The current discussion (Section 4.1) focuses more on changes in contribution rates than on direct dynamic evidence supporting its "transitional" nature.
2. The accuracy of the MixSIAR model is highly dependent on the representativeness of element values. In this study, precipitation elements are derived from monthly mixing samples, whereas soil water and groundwater values are based on monthly averages. However, during the melting season (May–June), diurnal freeze-thaw cycles occur, causing precipitation to manifest as rain, snow, or wet snow with distinct isotopic signatures. Additionally, snow cover may be influenced by sublimation of pre-melt snow, meltwater runoff over frozen or partially frozen surfaces, and soil surface evaporation following snowmelt. It is recommended that authors explicitly describe how such mixed precipitation samples are processed in the Methods section (2.4.3) and include a discussion (Section 4.4: Limitations) acknowledging the uncertainties introduced by this simplification, along with an analysis of its potential impact on conclusions regarding the contribution of meltwater (7.9–22.0%).
3. The text interprets a "lc-excess value closer to the precipitation end-point" as indicative of enhanced preferential flow, while a "lower LC-excess value" suggests dominance of piston flow (a soil water evaporation indicator) (Lines 43–44,226–228). However, in Figure 3c, the variation in groundwater lc-excess across the three periods is minimal (approximately between –5‰ and –2‰). Moreover, the term "close to" is overly vague. It is recommended that the authors: a) include the difference between groundwater lc-excess and precipitation-end-point lc-excess (Δlc-excess), or provide monthly data on precipitation, soil water content at different depths, and groundwater levels; and use seasonal variations of this metric to support the claim of "gradual enhancement of preferential flow," thereby enhancing the objectivity of the argument.

Minor Comments： 

1. This manuscript uses hydrogen and oxygen isotope methods to investigate the effects of freeze–thaw processes on the sources and pathways of groundwater recharge. It is recommended that the method used be explicitly stated in the title to make the manuscript’s theme clearer and more specific. For example: Freeze-thaw processes influence shallow groundwater recharge sources and pathways in the Qinghai Lake Basin: insights from water isotopes.
2.  Line 31: “exerts a strong influence” It is recommended to change to “exerts a strong influence on hydrological processes” make the expression more complete.
3. Line 34: “takes the Qinghai Lake basin (QLB) as a case example”. It is recommended to change to“takes the Qinghai Lake Basin (QLB) as a case study” more in line with academic conventions.
4. Line 37: The abbreviation "FT periods" appears for the first time; it is recommended to define it as "freeze-thaw (FT) periods" in the abstract when it first occurs (it has been defined in the main text but not in the abstract; consistency should be maintained).
5. Line 70: Are the references "Zhang et al., 2026" and "Biskaborn et al., 2025" unpublished or preprints? Please confirm whether they have been formally published; if in review, please indicate this.
6. Line 83: “midwestern united states”,“United States” first letter should be capitalized.
7. Line 134: For data such as "2.97×10⁴ km²" and "1.23×10⁴ km²," standardize the index format (use superscripts) and ensure no spaces are added between numbers and units throughout the text (e.g., "2.97 × 10⁴ km²").
8. Figure 1: The subgraphs (a)–(e) in the figure caption lack sufficient detail; it is recommended to specify the exact content displayed by each subgraph in the legend (e.g., the relationship between sampling point distribution and permafrost/seasonal frozen soil boundaries).
9. Table 1: The numbering conventions ("P1, P2..." under the "Precipitation" category and "G1, G2..." under "Groundwater") should be specified in the table header or footnote (whether based on elevation or watershed location).
10. Line 186: "Filtered through 0.45 μm membrane filters": Are all water samples (including meltwater) filtered? Is groundwater filtered? Please provide consistent clarification.
11. Line 196: In the analysis accuracy, the expressions "δ²H ± 0.5‰" and "δ¹⁸O ± 0.1‰" should have the symbol "‰" the numerical value; it is recommended to revise them to formats such as "δ²H: ±0.5‰".
12. Line 261: "In the 0–30 cm soil layer... lowest LC excess." Please supplement with statistical differences (and their significance) between the 30–60 cm and 60–90 cm layers; it is recommended to specify the significance test method used in the methods section to enhance the rigor of the analysis.
13. Line 263-264: “precipitation shifted overall from depleted to enriched”, It is recommended to specify the exact change in the average δ¹⁸O value (e.g., from-8‰ to-5‰) to enhance persuasiveness.14. Figure 3: The color differentiation in subplots (a), (b), and (c) of the figure is insufficient; it is recommended to use a color-blind-friendly palette.
14. Line 295: “GWL was closest to the 60-90 cm SWL” , Provide R² or the slope difference to quantify the degree of "proximity."
15. Line 361-362: The author mentions that "the contribution of meltwater is significantly lower than that of rainfall," yet Figure 6 shows that during the melting period, meltwater still accounts for 11.88–16.17%. It is recommended to emphasize in the discussion section that "although low, it cannot be overlooked" to avoid inconsistency in tone throughout the text.
16. Line 459-461: The bidirectional freezing process causes soil water to transition from liquid to solid state. Please elaborate on the potential impact of this process on lc-excess (whether it may lead to isotopic fractionation), as this aspect has not been addressed in the current discussion.
17. References: The literature formats are inconsistent: some entries include volume and issue numbers, while others lack issue numbers (e.g., Alley et al., 2002; Kooi, 2016). It is recommended to strictly adhere to the author guidelines of the journal Hydrology and Earth System Sciences and maintain uniform formatting. Additionally, there are several instances of inconsistent "et al." formatting (some in italics, others not); please review these cases.

This manuscript investigates the effects of freeze–thaw (FT) processes on shallow groundwater recharge sources and transport pathways in the Qinghai Lake Basin, combining stable water isotopes, hydrometeorological data, and the MixSIAR model, given ongoing permafrost degradation across alpine regions. The study provides useful insights: soil water acts as both a dominant recharge source and a transitional reservoir linking precipitation and snowmelt to groundwater, and thawing progressively enhances vertical soil-profile connectivity. Overall, the experimental design is reasonable, the data are abundant, and the analytical methods are appropriate, making this a well-structured manuscript. However, further improvements are needed in the following aspects: (1) specialized terminology should be further standardized and used consistently throughout the text; (2) figures and tables need refinement; (3) the conclusions and innovative points need further refinement and sharpening.

1. Use consistent terminology throughout the manuscript, including “freeze–thaw processes,” “freeze–thaw periods,” “soil water,” “shallow groundwater,” “recharge source,” and “recharge pathway.”
2. Refine the novelty and practical implications. The manuscript should more explicitly identify what is new relative to previous isotope-based studies of groundwater recharge in the Qinghai–Tibet Plateau. The practical implications for groundwater-resource management, ecohydrological protection, or river-baseflow maintenance in the Qinghai Lake Basin should also be stated more specifically.
3. Line 83: “united states” should be capitalized: “United States.”
4. Lines 83-84: Grammatical error. “snowmelts” is not a verb form. Please revise to “when snow melts and then refreezes in mid-winter.”
5. Line 106: This phrasing is informal. Consider replacing with “In the Three-Rivers Headwaters Region, for example,” for a more formal academic register.
6. Line 107: “Younger” refers to mean residence time or isotopic age, but this is not explained. Please briefly clarify what “younger” means in this context (e.g., shorter mean transit time).
7. Lines 121-122: Please revise to past tense: “we established a comprehensive ecohydrological monitoring network within the QLB.”
8. Line 165: Missing comma after “2015” and after “2023”. Please standardize to “Peng et al., 2015; Li et al., 2022; Peng et al., 2023.”
9. Lines 187-190: This sentence appears at the end of the sampling section but describes a spatial classification that affects the entire study. It would be better placed earlier in Section 2.1 when the study area is introduced.
10. Line 290: “its R² decreased” — decreased relative to what baseline or period? Please specify the comparison explicitly.
11. Line 296: Spaces are missing in the legend entries, e.g., “SW 0–30cm” should be written as “SW 0–30 cm”.
12. Lines 300-323: The SWC values are reported to two decimal places (e.g., 26.90%, 28.41%), implying a precision that may not be justified given the spatial heterogeneity of field measurements. Please report values to one decimal place.
13. Line 378: The unit “‰” is incorrect here; SWC is expressed as a percentage (%). Please correct to “26.90%–31.05%.”
14. Line 388: Throughout the manuscript, contribution values are reported to two decimal places (e.g., 13.01%, 9.33%). This value should be given as “10.60%” for consistency.
15. Lines 423-426: “This is consistent with results from modeling analyses indicating that spring hydrological processes in the alpine regions of the QTP are mainly governed by a combination of active layer temperature, thaw depth, and SWC” — no citation is provided for this modeling result. Please add the appropriate reference.
16. Line 469: Decimal places should be kept consistent throughout Figure 8.
17. Line 474: “RWL” has not been previously defined in the manuscript. Please define this abbreviation on first use.
18. Lines 492-495: Consider restructuring as: “Observations in an alpine meadow further confirm that by regulating active layer thickness, FT processes significantly alter the proportions of piston flow and preferential flow (Musa et al., 2016; Li et al., 2026b).”