Human-Induced Shifts in Yellow River Flooding: Population Threshold Effects in the Loess Plateau’s Primary Sediment Source Area (934 CE)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript investigates the abrupt increase in flood frequency of the lower Yellow River following 934 CE and quantitatively links it to human‐induced changes in the midstream Loess Plateau’s Primary Sediment Source Area (PSSA). Using multi‐proxy historical records (population, cropland, sediment yields, flood chronologies), GIS‐based spatial standardization, Pettitt change‐point detection, Wilcoxon rank‐sum tests, and Cliff’s δ effect‐size analyses, the authors identify 934 CE as a critical breakpoint. They show that population density in the PSSA surged from 1.3 to 19.8 persons/km² (a 14.2‐fold increase, p = 0.005), driving a 1.4‐fold rise in cultivation rate, a 5.4‐fold increase in sediment yield, a 10‐fold acceleration in downstream sedimentation rate, and ultimately a 5.5‐fold escalation in flooding frequency (from 1.6 to 10.4 events per 20 years) . The study proposes 19.8 persons/km² as an “ecological pressure threshold” and introduces a conversion framework that adjusts population density by soil‐water conservation measures (e.g., terracing, afforestation), recommending enforcement of sediment‐control practices whenever this threshold is exceeded.

Given its combination of long‐term historical data, spatially explicit analysis, and direct policy‐relevant recommendations, this work addresses key questions in hydrological regime shifts and sediment management. It offers a novel, quantitatively grounded framework for linking past human activities to flood risk, which is highly relevant to Hydrology’s focus on integrated water resources and sustainable flood mitigation.

Although the manuscript has strong potential, the following substantive revisions are required before publication:

 

Abstract:

 

The abstract clearly presents the study’s objectives, methodology, and key findings. Its structure and clarity are strengths of the manuscript.

 

Introduction:

 

The Introduction does not sufficiently frame the study within broader sustainability goals. It would be beneficial to reference the UN Sustainable Development Goals (e.g. SDG-6 “Clean Water and Sanitation”; see https://doi.org/10.1016/j.pce.2025.103900 and https://doi.org/10.1002/2017EF000632).

The literature review is too narrow and does not clearly establish the manuscript’s novelty. A major revision is needed to (a) expand recent relevant studies on sediment management and flood risk, (b) position this work in relation to existing hypotheses (e.g. vegetation degradation vs. climatic drivers), and (c) articulate the unique contribution of the ecological pressure threshold concept.

 

2. Methods

 

Figure 1 is missing a scale bar and clear coordinate grid.

I recommend adding a detailed workflow diagram to illustrate the sequence of data processing and analyses.

The descriptions of key methods (Pettitt change-point detection, GIS-based standardization, statistical tests) are overly superficial. Critical steps—such as historical boundary harmonization, area‐weighting algorithms, and sediment‐yield scaling—must be explained in depth.

 

3. Results

 

The Results section lacks sufficient detail. Each figure should be accompanied by explicit explanations of its significance for the research questions.

For example, explain how the change in population density drives sediment yield increases, and discuss the implications of the 5.5-fold rise in flood frequency.

Consider adding summary tables of effect sizes, confidence intervals, and p-values to support your narrative.

 

4. Discussion

 

The Discussion must clearly differentiate this work from previous studies. Detail how the proposed ecological pressure threshold addresses gaps in the literature.

Discuss the study’s limitations (e.g. uncertainties in historical flood chronologies, biases in HYDE cropland reconstructions) and suggest how future research might overcome them.

 

5. Overall Discussion and Conclusions

 

The current Discussion does not stand alone as a representative synthesis of the manuscript. It should be expanded to include:

Importance and advantages of the threshold framework

Limitations and potential sources of error

Practical recommendations for policy and management

Future research directions

A comprehensive rewrite of the Discussion and Conclusions is required to reflect these points.

Author Response

Dear Reviewer,

Thank you for your valuable comments, which have significantly enhanced the reliability, logicality, and standardization of this paper. We have strictly followed your suggestions in revising the manuscript. Point-to-point responses to the reviewers' comments are provided in the attached document.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

This paper analyzes the threshold effects of population in the Loess Plateau's Primary Sediment Source Area (PSSA) through millennial-scale investigations of population, land use, and Yellow River (YR) flooding events, some interesting results were found. The research methodology is credible in total, but some conclusions appear inconsistent with historical consensus:

(1) It is widely recognized that before 16th century, ancient Chinese population mostly followed dynastic cycles, peaking during mid-dynasty periods and collapsing during the fall of dynasties. The Tang Dynasty (618-907 CE), with its capital adjacent to the PSSA, was thriving in economics. Subsequent fragmentation and weak dynasties(post-907 CE) transformed the PSSA into a war frontier with nomadic groups until its eventual conquest. Meanwhile, China's economic center shifted southward to the Yangtze river basin. Therefore, the abrupt population surge around 934 CE contradicts these patterns. The authors should adopt more accurate ancient demographic models incorporating dynastic rise-fall cycles.

(2) The identification of 934 CE as a flooding transition point also requires scrutiny. In this paper, most ancient flood records seem to derive from disaster reports rather than modern quantitative discharge data. Notably, major YR channel avulsions occurred in 11, 1048, 1128, and 1368 CE. The increased flooding frequency post-934 CE might reflect incomplete flood control systems along new channels, compounded by warfare and some artificial breaches. That is, not a direct result of population and land use. Systematic analysis and discussions are required here.

(3)In Figure 1, what does different color map means?

(4)To enhance credibility, it is better to adopt a refined socio-hydro-geomorphic models to analyze parameters and threshold effects.

Author Response

Dear Reviewer,

Thank you for your valuable comments, which have significantly enhanced the reliability, logicality, and standardization of this paper. We have strictly followed your suggestions in revising the manuscript. Point-to-point responses to the reviewers' comments are provided in the attached document.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

General comments:

The paper deals with the human-induced shifts in Yellow River flooding and investigates the quantitative relationship between land use changes and flooding frequency in the midstream Loess Plateau’s Primary Sediment Source Area. Different datasets were used to obtain long series of population density and consequent land use changes in the study area. The authors were able to extend the datasets on population density and cultivation rate back to 400 years BCE and also data on land use dating 2000 years back.

The authors provide interesting insight into the history of the population increase and land use changes in one of the most erosive-prone areas in the world. However, I found several methodological deficiencies that need to be addressed and several points that need to be explained and discussed further (see specific comments).

First of all, according to the paper title, the main topic are human-induced changes in flooding of the Yellow River. However, little information on the flood characteristics/magnitude is provided. The authors analyzed the flooding frequency using a 20-year moving time window. The flood characteristics and the potential human-induced changes should be investigated more systematically. For example, in view of the potential shifts in flood return period or by using any other characteristics of flood events (e.g. the mentioned flood characterization using traditional Chinese hydrological terminologies). In my view, that would provide scientifically important information on the changed flood event dynamics.

Secondly, the authors have interesting data in the temporal changes in sediment yield. However, the impact of increased sediment yields/rates on flood safety is not investigated in detail as one would expect from the title of the paper.

The data analysis relies on the identification of the “turning point year” 934 CE. It remains pretty unclear throughout the paper how was this temporal threshold defined (see specific comments).

The statement about the population threshold (> 19.8 p/km2) is quite strong. I presume that there is for sure some uncertainty associated with the population density calculation through history since the population spatial density data are available for only 9 benchmark years. Therefore I suggest to be careful while report the population density in values with digits.

The discussion section provides general data on soil erosion processes and potential effects of land use changes on flood events occurrence and amplification of soil erosion processes. In the discussion section the authors merely repeat the results presented in the results section. Some more in-depth discussion on the results obtained in this study are missing, also in view of the erosion processes observed elsewhere.

 

Specific comments:

Abstract (line 26): The return period  or some other characteristics of the investigated flood events should be reported here.

Lines 57-58: Why are the “post-934 CE surge in YR” levee breaches important? Why is this considered as a threshold (turning year), based on what criteria? This should be explained more in detail.

Line 104: Were any population-density analysis done previously in scope of the LP-wide demographic analysis for the studied area (PSSA)?

Lines 127-140: Is “cropland” the only land-use type considered? Is the increase in the cropland areas on the account of less forest land use or any other type of land use? Figure 4 shows much more complicated land use patterns changes.

Lines 154-156: Is it possible to associate the traditional Chinese hydrological terminology for flood events to approx. flood return period? That would be of great value for generalizing the interpretation of results.

Lines 199-202: I suggest to move this part to the Discussion section.

Is the area shown in Figures 3 and 4 the same as the area covered by red polygon border in Figure 1?

Are data shown in Figure 5 from reference  [22], [23]? It should be referenced properly.

Figure 4: Are the data for the land use changes from the Historical Database of the Global Environment (HYDE) or also from some Chinese national data? It should be referenced properly.

Lines 228-230: What is here considered as a flood event, what is the definition of flood event in this context? What are the estimated return periods of flood events?

Lines 234-240: If I understand correctly, the annual precipitation data series do not express any changes in precipitation characteristics? Which dataset was used for the annual precipitation dating back to 400 years BCE?

Figure 6: Data show enormous increase in the number of flood events. How is the flood evet defined in this case? Figure 6 doesn’t show the correlation but only data series.

Table 1 and 2: How does the data given in Table 1 and 2 (from other studies) match with the data from this study (e.g. data shown in Figure 5)?

Line 302: My impression is that the statement about the increased flood frequency has not been supported enough throughout the manuscript.

Line 299-300: This statement is quite strong, in my view additional evidences would need to be provided to support it. Was direct effects of precipitation changes investigated in some other way in scope of this study besides the data show in Figure 6?

 

Author Response

Dear Reviewer,

Thank you for your valuable comments, which have significantly enhanced the reliability, logicality, and standardization of this paper. We have strictly followed your suggestions in revising the manuscript. Point-to-point responses to the reviewers' comments are provided in the attached document.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

I have examined the authors’ point-by-point responses and the revised manuscript. The authors have taken the reviewer feedback seriously and substantially improved the manuscript. Key enhancements include a broadened and better-framed literature review that positions the work relative to prior hypotheses (vegetation degradation vs. climatic drivers) and clearly articulates the novelty of the ecological pressure (threshold) concept. The methodological descriptions—such as historical boundary harmonization, area-weighting, and the construction of the long-term sedimentation rate sequence—are now presented with the necessary depth. Figure 1 has been updated with a scale bar and coordinate grid, and the statistical analysis has been strengthened by inclusion of effect sizes, p-values, and clearer narrative linking observed changes to the research questions.

The Discussion and limitations sections have also been expanded to differentiate the contribution of this work from earlier studies, to acknowledge uncertainties (e.g., in historical chronologies and HYDE reconstructions), and to outline future research directions and policy relevance. Overall, the revisions meaningfully address the major concerns, and the manuscript in its current form is suitable for publication. I recommend acceptance.

Author Response

Dear Reviewer,

We sincerely appreciate the reviewer's time and constructive evaluation of our manuscript titled "Human-Induced Shifts in Yellow River Flooding: Population Threshold Effects in the Loess Plateau's Primary Sediment Source Area (934 CE)". We are grateful for their recognition of our revisions and the positive recommendation for acceptance.

Should there be any additional formalities required prior to publication, please do not hesitate to contact us.

Thank you for facilitating this review process.

Best regards,

Tao Huang

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have adequately replied to my general and specific comments in the 1st review. I suggest properly explaining the background of historical breach event records from ancient Chinese documents in the Methods section, as most readers might not be familiar with such flood event datasets.

I congratulate the authors for an interesting study.

 

Author Response

Dear Reviewer,

Thank you for your valuable comments, which have significantly enhanced the reliability, logicality, and standardization of this paper. We have strictly followed your suggestions in revising the manuscript. The point-by-point responses are provided in the attachment.

Should there be any additional formalities required prior to publication, please do not hesitate to contact us.

Thank you for facilitating this review process.

Best regards,

Tao Huang

Author Response File: Author Response.docx