This paper presents a large number (n > 150) of MC-ICP-MS U-series dating of stalagmites from a cave in the northeastern Yucatán Peninsula, Mexico. These stalagmites are particularly important because they cover the entire period of Maya cultural evolution and can be used to reconstruct a continuous, high-resolution climate record over the past 2700 years. These samples are characterized by high U and low Th concentrations, which is generally favorable for precise U-series dating. However, dating results reveal significant age inversions. These inversions are attributed to elevated initial 230Th/232Th ratio, ranging from 4 to 68, as estimated through multiple approaches, including U-series dating of modern carbonate and drip water, isochron analyses, stratigraphic constraints, and lamina counting. This paper demonstrates the potential of large variations of initial 230Th/232Th ratio in speleothem, which may pose a substantial challenge for accurate U-series dating. I think this is well-organized study supported by convincing evidence.

My major comments are given below:

• The study focuses on stalagmites from Áaktun Kóopo Cave, located in the northeastern Yucatán Peninsula, and cites several other stalagmites records from elsewhere on the Peninsula. Is the elevated initial 230Th/232Th ratio specific for Áaktun Kóopo Cave, or does it reflect a regional characteristic?
• Samples with (²³⁰Th/²³²Th) activity ratio <200 were excluded from age modeling, and their corrected ages are not presented in Table S3. However, their initial (²³⁰Th/²³²Th) activity ratios are still presented. I suggest either presenting both the corrected ages and the initial values, or removing both sets of values. In the main text, the initial (²³⁰Th/²³²Th) activity ratios are reported between 4 and 68, but the corrected ages corresponding the higher initial values (>50) were discarded. Does this imply that the derived high initial values (>50) are unreliable?

• The sample EO-C-81.5 has the lowest Th concentration (~0.00013 ppb) and the highest (230Th/232Th) activity ratios (213749), yet its corrected age is still not consistent with the stratigraphic order.
• In Table S3, there are two rows of data for sample E6-31.5 and three rows of data for E6-220. Are these replicated analyses? Given the challenging of the Áaktun Kóopo Cave samples, it would be beneficial to present replicate analyses to show the reproducibility.

This manuscript addresses the challenge of uncertainty in U-Th dating due to high, complex initial thorium contamination in stalagmites. By integrating isochron analysis, stratigraphic constraints, and annual layer counting to determine the initial ²³⁰Th/²³²Th ratio successfully, and by establishing multiple age models to construct robust chronologies over the past 2.7 kyr in the Áaktun Kóopo cave in Yucatan, it provides a crucial chronological foundation for studying hydrological and environmental changes during the Mayan social upheaval. The study indicates that assuming a fixed initial thorium concentration may lead to significant dating biases, with important methodological implications for high-precision stalagmite chronology research.

1. Based on the titles and content, this manuscript is a methodological study to constrain initial thorium concentrations through several approaches and to determine a robust speleothem Age-depth model, which is then used to reconstruct hydrological change in the Yucatán Peninsula, Mexico. The emphasis should be placed more strongly on the process of method development, with the hydroclimatic reconstruction serving merely as an illustrative application. Currently, the organization/content of the abstract appears a bit unbalanced, with a relatively greater focus on presenting the paleoclimate results compared to the methodology used. Furthermore, although the introduction highlights the importance of constraining initial thorium for U-Th dating through a progressive line of reasoning, it should more clearly address the specific challenges posed by initial Th uncertainty in U-Th dating, as well as the limitations of the existing approaches to correct for initial thorium.

2.  References about the Isochron method should be added (line 78). This is important. 

3. In the introduction or study site part, there should be more details about the climatic, geological, and regional hydrological conditions of the study area. The initial (²³⁰Th/²³²Th) activity ratio of detrital material can vary with the characteristics of the host rock, soil, and chemical complexing agents during the aqueous transport of Th. Thus, it is better to give a clear description of the climatic, geological, and regional hydrological background.

4. The relative height or elevation information of the cave samples should be discussed because flood events can disrupt the stability of cave environments through clay input and sedimentary discontinuities.

5. The clarity and quality of Figure 1 both need to be significantly improved. Also, Figure 1b seems unnecessary, since this information is already shown in Figure 1a. It is recommended to provide more detailed information on the cave environment itself, and in particular, the places where the walls were modified. see P. 5

6. Methodology: Since this manuscript addresses the challenges of the initial Th, it is recommended to add more details about the U and Th extraction processes used in this study. see P. 8 205-215. Also, the methodology used to measure drip water is not provided; see Table S2.  Specific half-lives should also be provided, rather than just citing Cheng et al. (2013).

Questions:

• This ms focuses on the impact of high detrital Th on U-Th dating errors and emphasizes the importance of initial Th correction. However, it attributes anomalous fluctuations in high detrital Th concentrations in tropical regions as the rule, without delving into the controlling mechanisms. Does this fundamentally undermine the logical basis of the initial Th correction? In other words, if the specific processes and driving factors behind anomalous Th signals cannot be elucidated, will the effectiveness and applicability of isochronous ages calculated based on the assumptions of "uniform initial ratios" or "statistical averages" necessarily face significant uncertainties?
• The validity of the isochron approach fundamentally relies on homogeneous initial thorium isotopic compositions among co-genetic subsamples. However, the manuscript explicitly reports highly variable and largely unsystematic initial (²³⁰Th/²³²Th) activity ratios. How can the authors justify the assumption of isotopic homogeneity required for meaningful isochron construction under such strongly heterogeneous conditions?
• The authors relied on multiple verifications based on the isochronous method to correct the initial Th. It is known that young, low-thorium samples are highly sensitive to scattering caused by blank errors in the ²³⁰Th/²³²Th ratios. Still, the manuscript lacks a good discussion of the contribution of procedural blanks and the propagation of uncertainty. How do the authors handle the blank correction propagating into the uncertainty of the final isochronous line?
• Since not all speleothem samples have clear annual layers, how are the initial ²³⁰Th/²³²Th ratios determined for samples lacking chronological striations?
• The manuscript applies two-endmember mixing regressions while simultaneously arguing for highly variable and unsystematic initial thorium compositions. How do the authors reconcile the use of a binary mixing framework with evidence suggesting multiple temporally and spatially variable thorium sources?
• The manuscript addresses the limitations of fixed initial thorium corrections under highly “detrital-contaminated” conditions. However, the exclusion of samples with ²³⁰Th/²³²Th ratios < 200 raises an important methodological question regarding the practical applicability of the proposed approach. If the most strongly contaminated samples still need to be discarded to achieve robust chronologies, to what extent does the proposed framework truly overcome the limitations associated with high detrital thorium contamination? The current workflow appears to improve uncertainty treatment primarily for moderately contaminated samples rather than resolving the challenges posed by the most problematic end-members.
• The manuscript primarily focuses on the propagation of age uncertainties associated with initial thorium contamination. However, potential variability in uranium concentrations may also significantly affect the reliability of U–Th ages, particularly under prolonged flooding or submergence. Although petrographic observations suggest no obvious recrystallization or diagenetic alteration, preservation of primary fabrics alone may not fully exclude subtle uranium remobilization or cryptic open-system behavior. Given the hydrologically dynamic cave environment and evidence for flooding events, how do the authors demonstrate that uranium concentrations and isotopic compositions were not modified post-depositionally?
• Human activity increases the uncertainty of cave research. Did the Maya people's modification of the cave walls affect the cave's sedimentary environment? How to identify non-climate impacts of human activity changes?
• The manuscript emphasizes that dry conditions have been a major driver of social and environmental change among the Maya. However, much of the subsequent discussion focuses on flood events, inundation, and enhanced clay transport within the cave system. This raises an important question: the relationship between long-term hydroclimatic trends and short-term extreme hydrological events. Can the authors clarify whether the identified flood events are transient, superimposed on an overall drought trend, or reflect a trend of increasing hydroclimatic variability? Which factors contribute to the profound social and political upheaval marked by widespread population decline and the abandonment of many urban centers? A clearer distinction between mean drought conditions and sudden flood dynamics will help strengthen paleoclimate interpretation and improve the coherence of climate narratives.