General Comments:

This manuscript presents a timely and important study that challenges the conventional view of gullies as purely erosional, degraded features by positioning them as significant zones for groundwater recharge in the semi-arid Loess Plateau. The research employs an integrated multidisciplinary approach, combining stable isotope analysis, chloride concentration measurements, water-level fluctuation analysis, and hydro-statistical modelling to trace moisture flow paths among surface water, pore water, fissure water, and spring water at a high resolution. Based on this evidence, the authors redefine the hydrological role of gullies in arid ecosystems, directly challenging the traditional view of gullies as symbols of land degradation. The findings reveal that reframing gullies are not merely degraded geomorphic units but rather critical groundwater recharge zones and subsurface connectivity hubs. Precipitation primarily replenishes shallow pore water, while deep fissure water is supplemented by slow, top-down percolation. This understanding overturns the long-standing negative perception of gullies on the Loess Plateau, highlighting their capacity to buffer seasonal hydrological variability and enhance ecosystem resilience. Overall, this study addresses a key knowledge gap regarding groundwater dynamics in gully systems and holds significant practical implications for sustainable water resource management on the Loess Plateau. The manuscript is generally well-written and structured. However, some moderate revisions are needed.

Major Concerns:

1. The manuscript sets up a contrast with “piston flow” and “preferential flow” models from tableland studies but does not clearly define what process is dominant in the gullies. The proposed “gully-dominated preferential recharge mechanism” (Line 779) is not well-defined. Is the “preferential” aspect the topographic focusing of runoff into the gully, or are there actual preferential flow paths (macropores, cracks) within the gully soils?
2. In my opinion, the manuscript could benefit from clearer articulation of the broader implications of the key findings. For example, how can this insight change land management practices or ecological restoration strategies in other dryland regions globally?

Specific Comments:

1. It is recommended to simplify long sentences to improve readability. For example, lines 53–56: “In these ‘fragile’ and diverse landscapes, understanding the processes that govern when, where, and how groundwater is replenished—including the countervailing influences of vegetation dynamics, geomorphology, and engineered features—is essential for sustaining ecosystems, securing water resources, and informing land restoration and catchment management.” This sentence is structurally complex and could be simplified by breaking it into shorter clauses or highlighting the core information more clearly.

1. The text categorises groundwater into “pore water, spring water, fissure water”, and further suggested that the criteria for classification be clarified, such as medium type, storage space, and relationship with aquifer structure, to help readers understand the logical framework.The definition of “piston flow”(Line 145-146) is helpful but could be more concise. Consider: “Piston flow describes the displacement of pre-existing water by newly infiltrating water, moving frontally through the pore spaces.”
2. The manuscript lists permeability for Neogene coarse sandstone and conglomerate as 7.5–36.19 m/d (lines 213–214). These magnitudes are unusually high for such lithologies; I suspect a units or conversion mistake and recommend the authors re-examine the original data and report corrected values if necessary.
3. The text indicates that a low ITTP represents a long residence time, but the high ITTP of ponds (1.5±0.7) is interpreted as "rapid turnover + evaporation dominance,”seemingly overlooking the effect of evaporation on increasing variance. Could this be due to the small sample size for ponds/springs affecting the reliability of the analysis? Additionally, what is the reason for the small sample size for ponds/springs?
4. The high recharge rate of gully groundwater, accounting for 43% of precipitation—significantly higher than that in hill areas (<20%)—is a core conclusion of this paper and key evidence supporting the claim that “gullies are critical groundwater recharge zones and subsurface connectivity hubs.”While this conclusion is important, its robustness and uncertainties require further discussion, such as the assumptions underlying the recharge rate estimation method, spatial representativeness, and the impact of extreme events.
5. 9 shows that significant rises in groundwater levels and the main recharge period occur during the drier autumn and winter seasons (October to April), while recharge during the summer monsoon rainfall peak is minimal. The authors explain this as effective infiltration during the "cool, low-evaporation period" (Lines 601-604). Are there other potential reasons? For example, freeze-thaw processes, soil water reservoir effects, antecedent moisture conditions, or the competition between rainfall intensity and infiltration capacity?
6. The conceptual model (Fig. 10) emphasises the “restructuring”role of the gully system but does not discuss the potential risks of associated pollutant transport. Given that related issues are mentioned in the introduction, it is recommended to include a discussion on this aspect to present a more comprehensive perspective.
7. The conclusion section (Section 7) provides a good summary of the study's core findings. However, some statements appear slightly absolute, such as the claim to be “the first to quantitatively identify the unique cascading recharge processes in a thin loess gully catchment” (Lines 781-782). While the research is innovative, caution is advised with phrases like “the first.” It would be preferable to provide supporting literature references or adopt a more measured description.
8. The manuscript is largely well-written, but some sections contain complex or awkward sentence structures that could be improved for readability. For instance, the introductory and results sections sometimes use dense scientific language, which might be simplified without losing technical precision. Additionally, the formatting of the references section could be revisited for consistency.
9. Some important references are missing from the introduction and discussion sections:

De Vries, J. J., & Simmers, I. (2002). Groundwater recharge: an overview of processes and challenges. Hydrogeology Journal, 10(1), 5-17.

Huang L.M., Shao M.A., Advances and perspectives on soil water research in China’s Loess Plateau. Earth-Science Reviews, 2019: 102962.

Huang, L.M., Wang, Z.W., Pei, Y.W., Zhu, X.C., Jia, X.X., Shao, M.A., Adaptive water use strategies of artificially revegetated plants in a water-limited desert: A case study from the Mu Us Sandy Land. Journal of Hydrology, 2024, 644: 132103.

Xiang, W., Si, B. C., Biswas, A., & Li, Z. (2019). Quantifying dual recharge mechanisms in deep unsaturated zone of Chinese Loess Plateau using stable isotopes. Geoderma, 337, 773-781.

The authors have thoroughly addressed all the reviewers' comments they have received, making detailed revisions, clarifications, and additional analyses. These changes notably improve the manuscript’s clarity, rigor, and broader implications, especially in defining key mechanisms, addressing uncertainties, and incorporating relevant literature. I therefore recommend acceptance for publication in HESS with minor revisions.

1. There are too many dashes in the text; it is recommended to revise and simplify.
2. I suggest replacing the keyword "connectivity" with "hydrological connectivity", and if possiple add a keyword "groundwarer recharge".
3. It is unclear why the legend colors in Figure 6 cover the text instead of being right-aligned.
4. Redrawing Figure 10 to include the main findings of this study would be beneficial.