This paper addresses a critically important issue for arid land hydrology and water management and provides an important case study. It is generally very well written and clearly explained. However, I have two important reservations.

1. The conclusions depend on satellite-derived forcing data applied to a hydrological model. The interpretation of missing water is wholly dependent on the validity of the forcing data, but there is limited discussion of the use of local data to improve these products, and crucially no attempt to quantify the potential errors and their impact on the water balance conclusions. It is a statement of the obvious to note that a difference between two large terms (precip and evaporation) is extremely sensitive to the errors in those two terms. Some effort to quantify the effects of errors on the water balance conclusions is essential.
2. The discussion of the seasonal dynamics of groundwater recharge depends on modelling results, but the hydrological model has been calibrated using only surface flows. It seems that the groundwater dynamics are largely unconstrained and therefore the analysis of dynamics is wholly speculative. Some validation with local groundwater data would be invaluable. If not, at least there should be appropriate caveats around the discussion of results.

A final point is that although the English is generally excellent, there are a few points that need improvement – noted below.

Detailed comments:

Line 13 replace ‘insinuate’ by  ‘imply’

Line 37 composed of

Line 91 of the Uribe et al…

Line 94 into

Section 2.2 para 2 specify the model and forcing data time steps. I assume daily??

Line 112 the Uribe….

Line 116 data … are used…

Line 118 please clarify what is meant by evaporation here. It isn’t obvious until line 137.

Line 131 the Uribe...

Line 151 ‘first order approximation’. I note no discussion as yet of the likely error bounds on the satellite estimates of precip and evaporation, but this is crucial for the data interpretation!  The use of local data to improve the products is summarized rather briefly in Appendix B. More information would be helpful here, e.g. the local data available. The plots in App B do indicate quite large residual scatter. Some efforts to quantify likely errors and incorporate them in the analysis are in my view essential to the credibility of the conclusions.

Line 156 the Uribe... please correct throughout – including Fig 3 caption

Fig 3 results. The model that was used to simulate groundwater recharge was calibrated on observed river flows. So this provides only very limited information to define the dynamics of groundwater recharge fluxes. Were there no groundwater observations available to calibrate/validate this important component?

This manuscript presents an investigation of the spatiotemporal variability of groundwater recharge and upwelling in the Salar del Huasco basin, focusing on its interactions with precipitation, evaporation, and overall water balance dynamics in this arid endorheic system located in the Chilean Altiplano. The study addresses an important and timely topic with relevance to hydrological processes in data-scarce, high-altitude regions and provides valuable insights into the interactions among precipitation, evaporation, and groundwater recharge under conditions of extreme water scarcity. The manuscript is well structured, and the results are presented clearly. The work has the potential to make a contribution to the understanding of water balance processes in arid endorheic systems. However, certain aspects would benefit from clarification and further elaboration.

Some specific comments are provided below for consideration :

Lines 9-10: The message of this sentence is not entirely clear and would benefit from rephrasing to improve coherence and accuracy. Groundwater itself cannot “lead to recharge,” since recharge is the process that feeds or replenishes groundwater storage. The intended meaning seems to be that groundwater storage decreases, or that recharge reaches a minimum following the dry period.

Line 11: The term “competition” is not ideal in this hydrological context. Please replace “trade-off” or similar words to describe the contrasting relationship between groundwater recharge and evaporation.

lines 84-87 It would be helpful to express the total annual discharge in units of mm yr⁻¹, in addition to the volumetric flow rates (m³ s⁻¹), to facilitate direct comparison with precipitation and evaporation values reported elsewhere in mm yr^-1 in the manuscript.

Lines 131-132: The authors state that the Uribe et al. (2015) rainfall–runoff model is driven solely by precipitation, and in the model setup, evaporation occurs only in response to rainfall events. Please clarify the rationale for selecting this model despite this limitation, and discuss how this assumption may affect the results and the interpretation of the water balance.

Figure 3: It is unclear how “monthly mean variability” can be represented in mm yr⁻¹, since those units denote annual totals or rates. It appears that Figure 3 presents monthly mean values on the x-axis; therefore, the corresponding y-axis units should likely be expressed in mm (or mm month⁻¹), rather than mm yr⁻¹. Please clarify the unit definition. If the intention is to show monthly distributions, the values should be expressed in mm month⁻¹ (or simply mm), or the caption should explicitly state that the data have been annualized (e.g., monthly averages scaled to their annual equivalents). Please clarify the units and calculation method to ensure the figure is interpreted correctly. Otherwise, the comparison between rainfall and evaporation may be misleading, as expressing monthly values in mm yr⁻¹ artificially inflates their magnitude by a factor of 12.

Line 196 -206: In this section, the results are presented without a corresponding reference to any figure or table. Please specify which figure (or sub-figure) illustrates these findings so that readers can easily locate and interpret the results. Clear cross-referencing between text and figures would greatly improve readability and traceability of the analysis for the Results and Discussion section

The results described between lines 207–225 also seem to be based on monthly values presented in Figure 3. If this is the case, the current labeling in mm yr⁻¹ is inconsistent and may cause confusion. Conversely, if the data were converted to annual equivalents, it would be preferable to present Figure 3 using an annual scale rather than monthly intervals, to maintain conceptual consistency.

A similar applies to Figure 4, particularly the right panels (a–c). The values are presented as a function of months, suggesting that the data represent monthly means. If the intention is to illustrate the mean variation across available months (e.g., showing typical January, February, etc., values), then the units should correspond to the monthly time step. Alternatively, if the values have been annualized, please clarify this in the caption and consider adjusting the x-axis to represent annual rather than monthly time intervals for conceptual consistency.

Line 244: The text refers to “S1-type HRUs with an annual mean of ∼600 mm yr⁻¹ (see Fig. 2),” but Figure 2 does not display any numerical values or spatial distribution of annual means; it only shows the HRU classification. Please clarify whether these mean values are derived from another figure, dataset, or analysis step, and adjust the figure reference accordingly.

The Discussion section would benefit from a more detailed consideration of the limitations associated with the satellite-derived datasets and the assumptions of the applied model. Currently, these aspects are only briefly mentioned in lines 383–386. Given that the presented data are derived rather than directly measured, it is important to discuss the potential uncertainties arising from data correction methods, model parameterization, and structural assumptions.