Summer Monsoon Season Streamflow Variations in the Middle Yellow River since 1570 CE Inferred from Tree Rings of Pinus tabulaeformis
Jonathan Friedman
Round 1
Reviewer 1 Report
Thank you for the opportunity to review manuscript atmosphere-799386 “Summer monsoon season streamflow variations in the middle Yellow River since 1570 CE inferred from tree rings of Pinus tabulaeformis” by Chen et al. The subject of this manuscript is appropriate for the journal, data collection and analysis are professional and reasonable although not innovative, the conclusions are based on the data and the literature is sufficiently addressed. The paper is succinct, but the English is rough and will require substantial improvement. The first author of this manuscript was also an author of Li et al. (2019), a paper with a similar flow reconstruction for the Yellow River that included chronologies from the area sampled here and extended further back in time. The authors need to more clearly state how the current manuscript improves upon the older paper. Minor suggestions are below.
Fig. 1 caption. What do the landscape colors mean?
Fig. 2a caption. Is temperature red and precipitation blue?
How many trees were sampled at each site? How many cores?
Line 96. Exactly how were the three site chronologies merged into a regional chronology? Why did you not use the more flexible approach of applying multiple regression to predict streamflow as a function of the three chronologies, with each chronology included as an independent variable? This would allow sites more strongly correlated with flow to make a stronger contribution to the model.
Line 107. “Vice versa” is unclear here.
Figure 3. What method was used to calculate these monthly correlations and their significance?
Line 141. This line says the Liyuo River data was used as validation data, but the Methods Section said Leave-One-Out Cross validation was performed.
Line 150. The chronology in this study decreased precipitously since 1950, while the chronology in Li et al. (2019) did not. Does that suggest the trees cored in this study are more negatively affected by high temperatures than the trees in Li et al. (2019)? How does that affect flow reconstruction in past centuries?
Lines 152 and 154. Give the actual probability values.
Lines 199-202. You found a correlation between your chronology and the EASM index. How much of this correlation is due to the fact that both are decreasing over time? If you remove the trend from both time series is the relation between the two still significant?
The confusing term “streamflow cutoff” is used repeatedly to indicate zero flow in the river. Please replace this with a clearer phrase like “days of zero flow” or “flow cessation”.
Author Response
Thank you for the opportunity to review manuscript atmosphere-799386 “Summer monsoon season streamflow variations in the middle Yellow River since 1570 CE inferred from tree rings of Pinus tabulaeformis” by Chen et al. The subject of this manuscript is appropriate for the journal, data collection and analysis are professional and reasonable although not innovative, the conclusions are based on the data and the literature is sufficiently addressed. The paper is succinct, but the English is rough and will require substantial improvement. The first author of this manuscript was also an author of Li et al. (2019), a paper with a similar flow reconstruction for the Yellow River that included chronologies from the area sampled here and extended further back in time. The authors need to more clearly state how the current manuscript improves upon the older paper. Minor suggestions are below.
Response: First of all, thank you for your advice. I think there's been a mistake. I'm not the author of that article. The author of that article was Fahu Chen. Our article focused on monsoon season runoff, while the article focused on annual hydrological runoff
Fig. 1 caption. What do the landscape colors mean?
Response: Yes, I have added the color bar in fig. 1.
Fig. 2a caption. Is temperature red and precipitation blue?
Response: Yes, I have added
How many trees were sampled at each site? How many cores?
Response: Yes, I have added in table 1
Line 96. Exactly how were the three site chronologies merged into a regional chronology? Why did you not use the more flexible approach of applying multiple regression to predict streamflow as a function of the three chronologies, with each chronology included as an independent variable? This would allow sites more strongly correlated with flow to make a stronger contribution to the model.
Response: Because the three site chronologies have similar dry-wet signals, if the three chronologies are put together, they can not greatly increase the variance interpretation, but risk of over-fitting increased. In order to avoid over-fitting, we first develop a regional chronology, and then use this regional chronology to reconstruct the runoff. I further discussed the development process of tree-ring chronology.
Line 107. “Vice versa” is unclear here.
Response: Yes, I have improved.
Figure 3. What method was used to calculate these monthly correlations and their significance?
Response: Yes, I have improved.
Line 141. This line says the Liyuo River data was used as validation data, but the Methods Section said Leave-One-Out Cross validation was performed.
Response: Yes, since the 1950s, the runoff of the main channel of the Yellow River was seriously disturbed by human activities. Thus, we need a natural runoff to verify whether our reconstruction sequence can effectively represent the natural runoff of the middle reaches of the Yellow River, such as the Liyuo River data.
Line 150. The chronology in this study decreased precipitously since 1950, while the chronology in Li et al. (2019) did not. Does that suggest the trees cored in this study are more negatively affected by high temperatures than the trees in Li et al. (2019)? How does that affect flow reconstruction in past centuries?
Response: In fact, the tree-ring data in Li's paper come from the Tibetan Plateau with warm and humid condition over the past 30 years. However, at the middle reaches of the Yellow River, where runoff continues to decline. This downward trend is not only due to man-made reasons, regional drought is also an important reason. Therefore, there is a big difference between the two series. Our Chinese Pine tree-ring sequence captures this natural downward trend.
Fig.1b in Li et al., 2019 paper.
Lines 152 and 154. Give the actual probability values.
Response: Yes, I have added.
Lines 199-202. You found a correlation between your chronology and the EASM index. How much of this correlation is due to the fact that both are decreasing over time? If you remove the trend from both time series is the relation between the two still significant?
Response: Yes, if take away the downtrend, the correlation is not very high, nor is it significant. However, difference between runoff and precipitation, runoff is affected by the hysteresis of drought change, and has high low-frequency change signals, which is closely related to monsoon. Our reconstruction season corresponds to the monsoon season, and shows many signs of drought associated with changes in monsoon intensity, such as the droughts of 1928 and the late Ming Dynasty. Therefore, the low-frequency variations of streamflow correspond to the low-frequency variations of the summer monsoon
The confusing term “streamflow cutoff” is used repeatedly to indicate zero flow in the river. Please replace this with a clearer phrase like “days of zero flow” or “flow cessation”.
Response: Yes, I have corrected.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Review of “Summer monsoon season streamflow variations in the middle Yellow River since 1570 CE inferred from tree rings of Pinus tabulaeformis” by Chen et al.
I have completed my review of this paper and have noted this paper to be reconsidered after a major revision. The work undertaken is very good and is an important contribution to scientific knowledge. However, the decision to reconstruct streamflow of the Yellow River rather than the tributary Yiluo River needs to be reevaluated and the manuscript updated to reflect this change before I feel it will be suitable for publication. More detailed comments follow below.
Line 39 – provide
Lines 41-43 – This sentence is rather out of place. I would remove all the information about being the origin of Chinese civilization as that does not fit in with the remainder of the manuscript.
Line 52 and onward – once the full Latin name for the species has been given (as done on line 48) you can shorten the Latin name to P. tabulaeformis as long as it is not at the start of a sentence. Additionally, there is no longer a need to give the common and Latin names together (eg on lines 80-81) as this has already done once. In this manuscript there is a lot of switching back and forth between the common and Latin names and generally it is best to stick with one or the other so as not to confuse your readers.
Figure 1 – great figure but very hard to see the sampling sites in the Yellow colour. These are the most important items so I would suggest making them just black triangles – let the legend do the work for you.
Table 1 – please change ‘Tree Number’ to ‘Number of Trees’
Line 73 – should this be ‘dry-cold airflows’ ?
Line 93&97 – Here the word ‘strong’ is used to represent correlations but I think this is incorrect. A correlation of 0.25 is certainly not strong however in Table 2 it is noted that these values are true at the 99% confidence interval so I think the word that should be used here is ‘significant’ not ‘strong’
Line 97-98 – Why did you choose to use the STD chronology? Please provide your reasoning.
Lines 110-113 – It is indicated here in the methods that statistical analysis are being undertaken including sign test, reduction of error, Pearson correlation coefficient, and coefficient of efficiency. However, these values are not consistently reported on in the manuscript – all of this data needs to be included and I would suggest a table is the most appropriate way to do so.
Line 111 – Pearson
Section 3.1 & Figure 3 – It is discussed in this section on lines 128-130 that some seasonal assessments have been undertaken by what I am assuming is combining monthly streamflow data together. This should also be shown in Figure 3 – more columns can just be added after December and put the seasonal periods so that the data that is actually included in the reconstruction is present and the reader can see definitively that this is the strongest correlation and therefore the period to be reconstructed.
I have some issues with the results – specifically how this is not a reconstruction of the Yiluo River but a reconstruction of the Yellow River. The data for the Yiluo River is really good a 41/48 % explanation of variance is great for this type of study. However, translating this to the Yellow River the way that has been done is incorrect. The first issue is the time period – the Yiluo River data is for June-August and the Yellow River is for July-October so making a direct comparison does not work as this is not comparing like with like and only 25% of streamflow variation is explained here. The subsequent comparison between the chronology and the Yellow River streamflow is weaker than that of the Yiluo River but again I feel the way the authors got here is incorrect. It is indicated that there is a belief that these rivers are linked based on their fluvial influences (lines 157-159) but no references are given so this cannot be confirmed. Also while there is a discussion about anthropogenic factors causing variation here especially in the second half of the 20th century (which I’m sure does have an effect) there are quite a few other contrasts that are not discussed between using the tributary river and the main stream of a major river such as the Yellow River. As such I think that while the reconstruction of the Yellow River does not work here the authors could easily frame this paper around the reconstruction of the Yiluo River representing the tributary components from the Qinling Mountains which are important to the downstream availability of water on the Yellow River. The rest of the analysis that was undertaken for the Yellow River reconstruction would then need to be updated to reflect the change to the Yiluo River but it all would still be valid as representing the mountain tributaries into the mid-reaches of the Yellow River.
I would also suggest a change in the link with Indo-Pacific Ocean climate specifically in regards to Figure 5c. The explanation of this link is really good in that the authors are only looking at the area of actual influence on the Qinling Mountain region. Figure 5c though shows the entire globe. It is highly doubtful that any influence from Arctic waters would affect the Qinling Mountains but this is represented in the figure. This can easily be corrected by only including the area of influence which is discussed in the corresponding figure.
Lines 203-206 – the statements on how the EASM is linked to precipitation conditions over the Qinling Mountains makes me believe even more definitively that it is the Yiluo River that should be reconstructed here.
I really like how the authors have included section 3.3 here – the information is important in representing conditions affecting this region currently and demonstrating how they are not a natural long-standing phenomena. This section would still be valid with a focus shift to the Yiluo River as tributary rivers clearly represent an important water source entering the mid-reaches of the Yellow River.
Author Response
Review of “Summer monsoon season streamflow variations in the middle Yellow River since 1570 CE inferred from tree rings of Pinus tabulaeformis” by Chen et al.
I have completed my review of this paper and have noted this paper to be reconsidered after a major revision. The work undertaken is very good and is an important contribution to scientific knowledge. However, the decision to reconstruct streamflow of the Yellow River rather than the tributary Yiluo River needs to be reevaluated and the manuscript updated to reflect this change before I feel it will be suitable for publication. More detailed comments follow below.
Response: Thank you for your opinion. As you know, the Yellow River is a big river in the north of China, and the Yiluo River is just a tributary of middle Yellow River. However, the runoff of the main stream of the Yellow River is seriously influenced by human activities, so the calibration period is relatively short, which may affect the result of reconstruction. But there's no doubt that this reconstruction is significant, and it's also showed a lot of interesting drought signals, and these tree ring data have shown strong monsoon precipitation signals in previous studies.
Line 39 – provide
Response: I have corrected.
Lines 41-43 – This sentence is rather out of place. I would remove all the information about being the origin of Chinese civilization as that does not fit in with the remainder of the manuscript.
Response: OK, I have corrected.
Line 52 and onward – once the full Latin name for the species has been given (as done on line 48) you can shorten the Latin name to P. tabulaeformis as long as it is not at the start of a sentence. Additionally, there is no longer a need to give the common and Latin names together (eg on lines 80-81) as this has already done once. In this manuscript there is a lot of switching back and forth between the common and Latin names and generally it is best to stick with one or the other so as not to confuse your readers.
Response: OK, I have corrected.
Figure 1 – great figure but very hard to see the sampling sites in the Yellow colour. These are the most important items so I would suggest making them just black triangles – let the legend do the work for you.
Response: OK, I have corrected.
Table 1 – please change ‘Tree Number’ to ‘Number of Trees’
Response: OK, I have corrected.
Line 73 – should this be ‘dry-cold airflows’ ?
Response: OK, I have corrected.
Line 93&97 – Here the word ‘strong’ is used to represent correlations but I think this is incorrect. A correlation of 0.25 is certainly not strong however in Table 2 it is noted that these values are true at the 99% confidence interval so I think the word that should be used here is ‘significant’ not ‘strong’
Response: OK, I have corrected.
Line 97-98 – Why did you choose to use the STD chronology? Please provide your reasoning.
Response: OK, I have added.
Lines 110-113 – It is indicated here in the methods that statistical analysis are being undertaken including sign test, reduction of error, Pearson correlation coefficient, and coefficient of efficiency. However, these values are not consistently reported on in the manuscript – all of this data needs to be included and I would suggest a table is the most appropriate way to do so.
Response: OK, I have corrected.
Line 111 – Pearson
Response: OK, I have corrected.
Section 3.1 & Figure 3 – It is discussed in this section on lines 128-130 that some seasonal assessments have been undertaken by what I am assuming is combining monthly streamflow data together. This should also be shown in Figure 3 – more columns can just be added after December and put the seasonal periods so that the data that is actually included in the reconstruction is present and the reader can see definitively that this is the strongest correlation and therefore the period to be reconstructed.
Response: OK, I have added.
I have some issues with the results – specifically how this is not a reconstruction of the Yiluo River but a reconstruction of the Yellow River. The data for the Yiluo River is really good a 41/48 % explanation of variance is great for this type of study. However, translating this to the Yellow River the way that has been done is incorrect. The first issue is the time period – the Yiluo River data is for June-August and the Yellow River is for July-October so making a direct comparison does not work as this is not comparing like with like and only 25% of streamflow variation is explained here. The subsequent comparison between the chronology and the Yellow River streamflow is weaker than that of the Yiluo River but again I feel the way the authors got here is incorrect. It is indicated that there is a belief that these rivers are linked based on their fluvial influences (lines 157-159) but no references are given so this cannot be confirmed. Also while there is a discussion about anthropogenic factors causing variation here especially in the second half of the 20th century (which I’m sure does have an effect) there are quite a few other contrasts that are not discussed between using the tributary river and the main stream of a major river such as the Yellow River. As such I think that while the reconstruction of the Yellow River does not work here the authors could easily frame this paper around the reconstruction of the Yiluo River representing the tributary components from the Qinling Mountains which are important to the downstream availability of water on the Yellow River. The rest of the analysis that was undertaken for the Yellow River reconstruction would then need to be updated to reflect the change to the Yiluo River but it all would still be valid as representing the mountain tributaries into the mid-reaches of the Yellow River.
Response: Yes, I have to admit that you're right. Tree-ring data of pine should represent runoff changes in the Qin Mountains region. But as through the arid regions of northern China, Yellow River loses a lot of runoff, and replenished by runoff from Qin Mountains. so to a certain extent, hydrological changes in the Qin Mountains can also indicate changes in runoff in the middle reaches of the Yellow River. Based on your suggestion, we corrected the seasonal error and found a high correlation (r=0.70) between the runoff of the two stations. We believe that the reconstruction of the Yellow River is feasible. At the same time, Yiluo River is only a tributary of the Yellow River, this would make the article less meaningful
I would also suggest a change in the link with Indo-Pacific Ocean climate specifically in regards to Figure 5c. The explanation of this link is really good in that the authors are only looking at the area of actual influence on the Qinling Mountain region. Figure 5c though shows the entire globe. It is highly doubtful that any influence from Arctic waters would affect the Qinling Mountains but this is represented in the figure. This can easily be corrected by only including the area of influence which is discussed in the corresponding figure.
Response: Yes, I have corrected.
Lines 203-206 – the statements on how the EASM is linked to precipitation conditions over the Qinling Mountains makes me believe even more definitively that it is the Yiluo River that should be reconstructed here.
Response: You are right. As shown in figure 1, our study area includes not only Yiluo river basin, but also other larger rivers, such as Wei River (NWT and XF), although we do not have these river flow data. Therefore, it is feasible for us to use this regional chronology to reconstruct the Middle Yellow River runoff
I really like how the authors have included section 3.3 here – the information is important in representing conditions affecting this region currently and demonstrating how they are not a natural long-standing phenomena. This section would still be valid with a focus shift to the Yiluo River as tributary rivers clearly represent an important water source entering the mid-reaches of the Yellow River.
Response: Yes, this is the highlight of our paper, which examines the human influence far beyond the natural change. However, but there is streamflow cutoff occured in Yiluo river. If we only discussed about Yiluo Reconstruction, it makes this part meaningless
Reviewer 3 Report
This study obtains three tree-ring chronologies, computes a regional chronology, and uses it to reconstruct streamflow over the middle Yellow River. Overall, the authors do a fine job with the reconstruction, have a good argument for the decline in the correlations with the instrumental data after 1950, and have a defensible signal that shows expected relationships with the East Asian Summer Monsoon and Indo-Pacific SSTs. However, I am a bit hung up on the authors sentence below (lines 139-141):
“Considering the Yiluo River is just a tributary of the Yellow River and the observational data are relatively short, we use the streamflow data of this river as the validation data.”
The use of the word validation makes this unclear. The only validation that occurs is the leave one-out cross validation method during the reconstruction process. Instead, it seems that the authors compare the regional chronology’s correlation with Yiluo River streamflow to its correlation with Yellow River streamflow in order to make the argument that the regional chronology can be used to make a successful reconstruction along the Yellow River despite the chronologies residing upstream along a tributary. This is fair, but just not stated clearly. I recommend changing the wording in the quoted sentence above. I would also recommend that after line 169, the authors refer back to the Yiluo River explained variance of 41% again just to caution the reader on the lower explained variance for the middle Yellow River. This decrease in explained variance also seems reasonable given, as they state line 158, the middle Yellow River streamflow is also influenced by areas farther upstream away from the tree-ring sites. I do want to stress that this is just a minor comment for improved clarity and does not impact the results.
Other minor comments:
Line 38: Make resource plural (resources)
Line 39: An e is left off from the word provide
Line 41: Make system plural (systems)
Line 51: Insert an a between developing and regional (developing a regional)
Line 57: Make site plural (sites)
Line 81: Insert and between mountains and are (mountains and are)
Line 85: Insert in between located and the (located in the)
Line 148: Insert structures between engineering and have (engineering structures have)
Line 204: Change is to are (Yellow River are linked)
Line 211: Make sources singular (water vapor source areas)
Line 212: Change suggesting to suggest
Line 249: Delete for
Author Response
This study obtains three tree-ring chronologies, computes a regional chronology, and uses it to reconstruct streamflow over the middle Yellow River. Overall, the authors do a fine job with the reconstruction, have a good argument for the decline in the correlations with the instrumental data after 1950, and have a defensible signal that shows expected relationships with the East Asian Summer Monsoon and Indo-Pacific SSTs. However, I am a bit hung up on the authors sentence below (lines 139-141):
“Considering the Yiluo River is just a tributary of the Yellow River and the observational data are relatively short, we use the streamflow data of this river as the validation data.”
The use of the word validation makes this unclear.The only validation that occurs is the leave one-out cross validation method during the reconstruction process. Instead, it seems that the authors compare the regional chronology’s correlation with Yiluo River streamflow to its correlation with Yellow River streamflow in order to make the argument that the regional chronology can be used to make a successful reconstruction along the Yellow River despite the chronologies residing upstream along a tributary. This is fair, but just not stated clearly. I recommend changing the wording in the quoted sentence above. I would also recommend that after line 169, the authors refer back to the Yiluo River explained variance of 41% again just to caution the reader on the lower explained variance for the middle Yellow River. This decrease in explained variance also seems reasonable given, as they state line 158, the middle Yellow River streamflow is also influenced by areas farther upstream away from the tree-ring sites. I do want to stress that this is just a minor comment for improved clarity and does not impact the results.
Response：Thank you for you commets. Yes, since the 1950s, the runoff of the main channel of the Yellow River was seriously disturbed by human activities. Thus, we need a natural runoff to verify whether our reconstruction sequence can effectively represent the natural runoff of the middle reaches of the Yellow River, such as the Liyuo River data.
Other minor comments:
Line 38: Make resource plural (resources)
Response: I have corrected.
Line 39: An e is left off from the word provide
Response: I have corrected.
Line 41: Make system plural (systems)
Response: I have corrected.
Line 51: Insert an a between developing and regional (developing a regional)
Response: I have corrected.
Line 57: Make site plural (sites)
Response: I have corrected.
Line 81: Insert and between mountains and are (mountains and are)
Response: I have corrected.
Line 85: Insert in between located and the (located in the)
Response: I have corrected.
Line 148: Insert structures between engineering and have (engineering structures have)
Response: I have corrected.
Line 204: Change is to are (Yellow River are linked)
Response: I have corrected.
Line 211: Make sources singular (water vapor source areas)
Response: I have corrected.
Line 212: Change suggesting to suggest
Response: I have corrected.
Line 249: Delete for
Response: I have corrected.
Round 2
Reviewer 1 Report
Thank you for the opportunity to review the revised version of manuscript atmosphere-799386 “Summer monsoon season streamflow variations in the middle Yellow River since 1570 CE inferred from tree rings of Pinus tabulaeformis” by Chen et al. First, let me apologize for mistaking the corresponding author of this paper (Feng Chen) for one of the authors of the earlier paper by Li et al. (2019). The authors have addressed some of my earlier comments, but two issues remain. First the English will need substantial editing to attain the standards of this journal. Second, the authors need to contrast their results with those of Li et al. (2019) and related studies in more detail. The central reason for this research stated in the Introduction is to examine how a chronology from pines in the Qinling Mountains can explain aspects of flow in the Yellow River not already explained by similar studies using trees from the Tibetan Plateau (e.g. Li et al. 2019). The authors need to strengthen the discussion of this issue to demonstrate the importance of their work. One difference is the strong decline in growth in the Qinling pines in the last half century, which is not seen in the Tibetan Plateau chronology (Li et al. 2019). The authors suggest this indicates the Qinling pines are more negatively affected by high temperatures than the trees cored for the Tibetan Plateau chronology. So how does this different temperature sensitivity differentially affect the two reconstructions of flows in the pre-industrial period? I understand the two chronologies differ in that Li et al. 2019 were examining annual flow while the current manuscript addresses only the monsoonal contribution; however, the season reconstructed here, July-October, includes most of the annual flow, including much of the runoff from the Tibetan Plateau, so the overlap is considerable. Furthermore, the reconstructions in both cases are based on annual ring width, so the chronologies are directly comparable. Finally, the authors argue that precipitation downstream of the Tibetan Plateau makes an important contribution to flow in the Middle Yellow River. It would be helpful for the reader to know what percentage of the July-October flow of the Yellow River at Huayuankou comes from monsoonal precipitation in the regions below the Tibetan Plateau.
Author Response
Thank you for the opportunity to review the revised version of manuscript atmosphere-799386 “Summer monsoon season streamflow variations in the middle Yellow River since 1570 CE inferred from tree rings of Pinus tabulaeformis” by Chen et al.
First, let me apologize for mistaking the corresponding author of this paper (Feng Chen) for one of the authors of the earlier paper by Li et al. (2019). The authors have addressed some of my earlier comments, but two issues remain.
Response: Thank you for your suggestions.
First the English will need substantial editing to attain the standards of this journal.
Response: Yes, I have ask the Translation Company helped me polish the article.
Second, the authors need to contrast their results with those of Li et al. (2019) and related studies in more detail. The central reason for this research stated in the Introduction is to examine how a chronology from pines in the Qinling Mountains can explain aspects of flow in the Yellow River not already explained by similar studies using trees from the Tibetan Plateau (e.g. Li et al. 2019).
Response: Yes, I have added.
The authors need to strengthen the discussion of this issue to demonstrate the importance of their work. One difference is the strong decline in growth in the Qinling pines in the last half century, which is not seen in the Tibetan Plateau chronology (Li et al. 2019). The authors suggest this indicates the Qinling pines are more negatively affected by high temperatures than the trees cored for the Tibetan Plateau chronology. So how does this different temperature sensitivity differentially affect the two reconstructions of flows in the pre-industrial period? I understand the two chronologies differ in that Li et al. 2019 were examining annual flow while the current manuscript addresses only the monsoonal contribution; however, the season reconstructed here, July-October, includes most of the annual flow, including much of the runoff from the Tibetan Plateau, so the overlap is considerable. Furthermore, the reconstructions in both cases are based on annual ring width, so the chronologies are directly comparable. Finally, the authors argue that precipitation downstream of the Tibetan Plateau makes an important contribution to flow in the Middle Yellow River. It would be helpful for the reader to know what percentage of the July-October flow of the Yellow River at Huayuankou comes from monsoonal precipitation in the regions below the Tibetan Plateau.
Response: In fact, we've compared the two series, and we found that they actually involve two different water sources, and their mechanisms are very different, even negatively correlated. The reconstruction of annual runoff of the Tibetean Plateau is based on the runoff of ice-snow melt water, while our reconstruction is only a small part of the runoff of the Yellow River during the monsoon season. So, we didn't put this result. At the same time, the effect of climate on the physiological process of tree rings in the two regions is quite different. The change in temperature produces two diametrically opposed results. Of course, we can also add them to our results
Reviewer 2 Report
I have completed my review of the revised version of ‘Summer monsoon season streamflow variations in the middle Yellow River since 1570 CE inferred from tree rings of Pinus tabulaeformis’ by Chen et al. While the authors have made many improvements to this paper and have taken many of the reviewers comments on board I still cannot agree with the authors that the decision to reconstruct the Yellow River is the correct one based on this data.
What the regional chronology developed here represents is the contribution of streamflow from the Qinling Mountains into the Yellow River. The data in Figure 2 show that this contribution is less than 10% of the overall flow in the middle Yellow River. Therefore the data presented cannot represent the entire middle Yellow River streamflow. I do understand that in Figure 4b the authors show a strong connection between the two study river streamflow z-scores. However, this is the part of the record that is indicated as being highly influenced by anthropogenic effects on the Yellow River. This could therefore represent a stronger representation of modern Yellow River streamflow coming from the mountain tributaries as this area is highly linked to precipitation (as per the strong connection to the EASM) and less likely negatively influenced by anthropogenic effects. Prior to this shift towards anthropogenic influence on the Yellow River there may have been a greater difference between these two areas.
The authors also state that if it was the Qinling Mountain tributary rivers that were reconstructed here that the discussion in section 3.3 about Human Impacts would be meaningless. I disagree with this statement. The fact that the tributary rivers have shown variations between high flow (flood conditions) and low flow (drought conditions) is very important considering anthropogenic influences. These mountain areas have not yet been heavily modified by human activity BUT they still show changes in flow based on natural variability AND they seem to be a more important part of the overall flow system now that human modification is occurring throughout the Yellow River basin. Under a warming climate the likelihood that a shift towards drought conditions will occur in this region is high. So without the contributions from areas such as the Qinling Mountains the issues faced by those relying on the Yellow River as a resource become even more dire.
The points that you are making and the importance of your study does not change by presenting your regional chronology from the Qinling Mountain tributary rivers as a contribution to the middle Yellow River. However, I am not convinced by the links that you have attempted to make between the two study rivers used as they represent very different basins and as such I do not feel they are directly comparable. It is for this reason that I cannot accept your decision to present this study as a middle Yellow River streamflow reconstruction.
Author Response
Thank you for your suggestion, and please reveiw the response letter.
Author Response File: 
Author Response.pdf
