This is interesting study because it quantitatively verified Japan's historic river diversion project using a high-resolution global hydrological model and historic data developments. Reconstructing hydrological conditions from centuries ago, when data were scarce, is challenging and contributes not only to validating specific historical events but also to advancing hydrological analysis.

One objective of this study is to test the hypothesis from previous studies that the diversion project's purpose was to enhance low flows to maintain the stability of the navigation network. References to other English literature on this project have indicated various interpretations of its purpose, including land reclamation, military defense, and flood protection (e.g., Mushiake, 1988). Although this study tested one of these interpretations, it is questionable whether it is appropriate to draw conclusions about such an important aspect of river history based solely on the results of this study, which focused on a single interpretation. While the quantitative validation is significant, conclusions should be considered with verification for other interpretations of the project.

The bifurcation function is adjusted from 70:30 in the present day to 50:50 in the historical figure, but the validity of this adjustment is unclear. This bifurcation function is crucial in determining the low flow of the divergent rivers. There is a risk that the stability of the low flow/navigation network of past divergent rivers is almost entirely determined by this function. When examining historical events, making such a bold assumption about this critical figure is questionable. At least some cases of this function need to be verified.

The historical river port locations are used as validation data, but the nature, validity, and reliability of this historical data need to be clarified. As the authors indicate, one of the critical contributions and challenges of this study is validating the simulation results in an era without modern river measurements. The reliability of this validation data could determine the significance of this study. Additionally, the process of developing historical data of river channels should be explained in more detail.

This manuscript presents an interesting study attempting to simulate the streamflow in a hydraulic engineering project influenced basin several centuries ago, and illustrate the values of Tone River Eastward Diversion Project in aiding navigation. The reconstructed streamflow is indirectly validated by some proxies, but no datasets of ancient hydrological and meteorological datasets are used to force or validate the model, which forms a significant disadvantage for a streamflow reconstruction study. However, I think this would be a small issue if the research topic can be adjusted slightly. Besides, I find some important methods/results are presented inadequately or unclearly. Consequently, I recommend a major revision before publication, by addressing the following major and specific comments.

1. Expression of the research focus

When talking about historical streamflow reconstruction, we would certainly expect the reconstruction of the climate factors, since climate forcing is one of the most important factors influencing streamflow, leading to strong hydrological nonstationarity. This study mainly explores the streamflow produced by different historical maps. Although authors have acknowledged the limitation in this aspect, I still think the lack of climate reconstruction is an inevitable shortage. However, if we regard this study as an investigation on the influence of TREDP on streamflow, the lack of climate reconstruction would not be so important. Consequently, I suggest the authors to change the expression of the research topic, focusing on the influence of TREDP, rather than “reconstructing” the “historical” streamflow.

2. Model calibration

I am confused about the calibration procedure. The authors only adopt three different values for each procedure, totally 81 combinations. This seems not a normal calibration procedure, in which simulation discharge produced by a large number of parameters need to be calculated. Besides, the T2E bifurcation function is set as 70:30% because the NSE values are highest at this time (L262-264). Is NSE always highest at 70:30% for all the parameter combinations? Several lines afterward, the authors claim that the rationale for this value is the near-equal NSE of Kitasekiyado and Nishisekiyado stations. Is this consistent with the sensitivity analysis described previously? Shouldn’t this bifurcation function be determined by something like project planning?

3. Method description

The reconstruction of the historical maps seems to be a very important part in this study, because the simulation of historical streamflow is produced by simply replace the present map by historical ones. However, the method producing historical maps is only described by several simple sentences and referring to several literatures. I suggest authors to describe this part in details.

4. Figure 8 is an important figure showing the validation of historical simulation, but I find this figure and the interpretation on it extremely difficult to follow. What is the meaning and unit of the colorbar? What do the hollow circles represent? What do the channels with deep and shallow color in each map mean? The units of frame should be added. Also, I suggest the authors to refer to the specific point/line in the figure (e.g., … as shown by the XX point in Figure X) when describing this figure in the main text.

Minor issues:

-Figure 6: Maybe replace “before” and “after” by “without” and “with”. “Before” and “after” seem to describe the time when an event happens, which may lead to misunderstanding.

-L461: I think Figure 9 is just showing the simulated streamflow of different historical years, not a “validation”.

-Figure 9: 1966 should be 1666?