This study investigating the source, transport, and fate of particulate organic carbon (POC) in the Huanghe, contributes to the comprehensive understanding of the global carbon cycle. The cross-channel sampling scheme used to investigate POC dynamics is novel and the identification of vertical heterogeneity in organic carbon transport, controlled by bathymetry and hydrodynamic sorting, is noteworthy. Indeed, it underlines the need for considering the heterogeneity of POC transport across channel sections while estimating POC fluxes and determining transport modes.

The authors provide an intriguing perspective on POC sources by using carbon isotopes (13C and 14C), suggesting the mobilization of aged and refractory organic carbon from the deeper soil layers of the loess-paleosol sequence in the Chinese Loess Plateau is a significant contributor to fluvial POC. The paper further presents a comparison of the calculated POC fluxes with existing literature, noting a significant reduction in POC flux in 2016 compared to that in the period of 2008-2013. This reduction, including OCbio and OCpetro, has been attributed to anthropogenic activities, mainly dam construction. However, there are some points I wish to raise for the authors' consideration.

First, the isotopic endmember values used for source deciphering are somewhat unclear, especially the topsoil's 13C endmember value of -24.8 ± 1.9‰, as the observed POC 13C range is -25 to -27‰ during the study period, which falls within the uncertainty of the endmember.

Secondly, the methods used to calculate POC fluxes, including those of OCbio and OCpetro, appear to be somewhat vaguely descripted (lines 470-490). I guess the authors used the observed instantaneous fluxes combined with the total suspended solids (TSS) from the gauging station to calculate monthly fluxes, which were then extrapolated to annual fluxes. And then the annual mean flux in per second unit was obtained for comparison. However, the error introduced by converting one snapshot to the annual averaged flux needs to be explicitly addressed to support their statement regarding the POC flux reduction. To my knowledge, there hasn't been a significant reduction in sediment discharge since 2008.

Overall, this manuscript provides valuable insights into POC transport, sources, and instantaneous fluxes, potentially advancing the current understanding in the field. With further elucidation of the points raised, it could be an important contribution to the literature.

This study presents a detailed examination of particulate organic carbon dynamics in the lower Huanghe River, a highly managed river system. Through a comprehensive analysis of sediment and particulate organic carbon concentrations and compositions across a channel transect, distinct patterns are observed both laterally and vertically, which can be attributed to the influence of riverine hydrodynamics. To quantify the contributions from different sources, a dual carbon isotope mixing model is applied, considering inputs from topsoils, the Chinese Loess Plateau, and petrogenic carbon. Additionally, the study utilizes a Rouse model to simulate instantaneous fluxes. These findings contribute to our understanding of POC dynamics in the lower Huanghe River and its complex interactions with hydrodynamics, providing valuable insights for the management and conservation of riverine ecosystems.

The study emphasizes the importance of conducting depth sampling across a river transect to precisely assess sediment and particulate organic carbon concentrations. It acknowledges the significant impact of hydrodynamics on the distribution and composition of both organic and inorganic components within the water column. Moreover, accurate modeling of export fluxes relies on a comprehensive understanding of these factors. The manuscript is written in suitable language and aligns well with the scope of Earth Surface dynamics.

However, the manuscript contains minor disconnections between the presented data and the corresponding interpretations. To enhance the discussion section, it would be valuable to provide a more detailed and thorough analysis of the hydrodynamic mechanisms, particularly in relation to the Rouse model. This detailed analysis would provide valuable insights into the observed heterogeneity in biogeochemical and sedimentary characteristics within the depth samples. Additionally, expanding the literature review to encompass relevant studies and recent publications that utilize river depth sampling would further enhance the manuscript's robustness and overall quality, by placing the findings within a broader scientific context.

The manuscript lacks sufficient provision and explanation of the statistical metrics used in the analysis, which undermines transparency and reproducibility. To address this concern, it is crucial to offer a more thorough explanation of these metrics, ensuring that readers can understand and replicate the analysis with clarity.

A major concern arises from the incomplete reporting of the Bayesian mixing model, which limits reproducibility. The model description lacks essential details, such as parameterization, prior and posterior distributions, and convergence diagnostics. Detailed guidelines are provided below to address these concerns and improve the reporting of the model.

Before recommending the publication of this study, it is crucial to address these shortcomings and fully resolve the concerns that have been raised.

Lines 18-20: This statement does not accurately apply to JB-1-3, as the highest radiocarbon values are observed at maximum depth in this particular case. Rephrase this sentence more carefully.

Lines 142-143: Please provide the coordinates for the sample location.

Lines 143-144: Please provide references to any relevant previous studies to support your statement.

Lines 168-172: Kindly provide the established standards for both stable and radiocarbon measurements. Moreover, was the amount of extraneous carbon taken into consideration during the radiocarbon measurements?

Lines 174- 183: Since you have implemented a custom Bayesian approach instead of utilizing a reported R package, it is essential to provide additional information in either Section 3.3 or Appendix A, expounding on the Bayesian modeling methodology. I recommend to include the following specific details:

1. Data Variables: Clearly specify the variables used in the analysis and their corresponding data sources.
2. Likelihood Function and Parameterization: Describe the likelihood function employed in the Bayesian model and provide details on how the model parameters are parameterized.
3. Prior Distribution: Clearly state whether the prior distribution used for the model is informed or non-informed. Provide a formal specification of the prior distribution.
4. Prior and Posterior Predictive Checks: Explain the procedures followed for conducting prior and posterior predictive checks, which involve comparing model predictions with both observed and simulated data.
5. Model Comparison: Describe the approach used for comparing different models and evaluating their relative performance in terms of predictive accuracy.
6. Model Bias (Geometric Surface Area): Elaborate on how the concept of model bias, specifically related to geometric surface area, was incorporated and assessed in the analysis.
7. Software Used: Specify the software or programming environment employed for implementing the Bayesian modeling approach.

In addition, it is important to assess model convergence using diagnostic tests such as Geweke, Gelman-Rubin, and Heidelberg-Welch. These tests provide valuable insights into the behavior of Markov Chain Monte Carlo (MCMC) chains and ensure reliable results. To gain further guidance on best practices for mixing models, consider referring to Phillips et al. (2014, Canadian Journal of Zoology) and Kruschke (2021, Nature Human Behavior). Furthermore, it is recommended to carefully evaluate the burn-in period in relation to the length of the MCMC chain to address concerns about convergence adequacy. A short burn-in period may raise questions about convergence. Lastly, it is crucial to provide accurate and comprehensive reporting of the posterior analysis in either the Results or Discussion section, enabling readers to reproduce the analysis effectively.

Line 201: When reporting the average, the corresponding standard deviation or standard error should be included, along with the number of samples.

Line 208: Figure S4 illustrates the values of D10 and D90. However, the actual data corresponding to these values has not been provided in the report.

Line 208-2010: Upon visual inspection, the suggested bimodal distribution of these sediment curves is hardly discernible.

Lines 2018-2019: To avoid confusion, it is necessary to clarify whether the reported value represents the standard deviation or the standard error. Prior to utilizing the value, please provide a clear definition of the specific measure being used.

Lines 295-297 and 317-324: To gain insights into the observed variation in the Huanghe River (JB-1-3), it would be beneficial to reference recent studies that have provided evidence for the systematic transfer and export of discrete plant-derived debris above the riverbed in major river systems. Consider examining research papers such as Feng et al. (2016, JGR: Biogeosciences), Lee et al. (2019, PNAS), and Schwab et al. (2023, JGR: Biogeosciences) for a better understanding of this phenomenon.

Lines 320-323: Given that your sampling focuses on high stage conditions, it is important to acknowledge that the inundation of adjacent riparian zones may contribute to the mobilization and entrainment of discrete plant-derived debris. It is advisable to also take into account factors such as surface runoff triggered by storm events and direct litterfall. These additional considerations can provide insights into the dynamics of plant-derived debris mobilization in the study area.

Lines 309-312 and 325-332: To enrich the interpretation and foster a more comprehensive understanding of the chemical composition variations within the transect, it is valuable to reference your Rouse model. This model, characterized by its ability to offer a more continuous representation, can effectively shed light on the impact of hydrologic dynamics, such as gravitational settling and resuspension. By incorporating a discussion based on the Rouse model, a deeper understanding of how these hydrologic processes influence the observed variations in chemical composition can be achieved within the transect.

Lines 349-352: This sentence is ambiguous. Could you please provide further elaboration on how a reduction in sediment load can impact the radiocarbon composition? Additionally, it would be helpful to discuss the primary sources of radiocarbon that were prevalent before 1950.

Lines 360-364: To strengthen your argument, consider including specific numerical values for the erosion rate of the Chinese Loess Plateau. This addition will provide quantitative support to your statement.

Lines 527-537: It is advisable to consider the potential effects of climate change-induced environmental changes. Notably, Xue et al.'s study (2023, Nature Communications) highlights a strengthening of the East Asian monsoon. This intensified monsoon activity is expected to have a positive influence on the erosivity of the Chinese Loess Plateau, consequently impacting the transport of sediment to the Huanghe River.

Appendix A: In order to evaluate the robustness of the endmember composition, the numerical values of the endmember composition should be supplemented with the corresponding sample numbers. Additionally, it may be worth considering weighted means and standard deviations of each source relative to their respective carbon content. To enhance readability, I recommend presenting the endmember compositions in the form of a table. Furthermore, why was vegetation not considered as an endmember composition? Given that the sampling campaign took place during a high river stage, it may be crucial to account for the entrainment of plant-derived debris from the proximal floodplain due to flooding, as well as the direct litterfall.

Figure 1: As sediment retention in dams is a significant aspect of your analysis, incorporating the dam locations on the map will provide vital visual information and enhance the understanding of sediment dynamics within the studied system.

Figures 3, 4, and 5: To bolster the connection between chemical properties and hydrological dynamics, it would be advantageous to incorporate a size parameter in your plots that reflects sampling depth, flow velocity, or discharge. By adding a third dimension to your plots, such as varying marker sizes, you can visually represent the additional hydrological information.

Figure 6: To ensure a thorough analysis, it is important to report further regression information, such as the equation, p-value, number of samples, root mean square error (RMSE), and mean absolute error (MAE). These details will provide a comprehensive understanding of the statistical relationship between the variables. Additionally, considering the incorporation of the particulate organic carbon content as a size parameter in your plots will enhance the visualization and enable the exploration of its potential impact on the observed patterns.