Spatio-Temporal Variation of Precipitation in the Qinling Mountains from 1970 to 2100 Based on CMIP6 Data
Round 1
Reviewer 1 Report
Dear Authors,
I appreciate the goal of your study.
Unfortunately, I had a very hard time understanding what you did and how, and what the results of your work mean. My hard time started with the Abstract, which is simply unreadable in its present form, and could not be understood if it stood alone on the Web.
Major Comments:
Lines 84-85: Do you realize that you try to predict 86 years of annual precipitation (2015-2100) from 45 years of data/observations (1970-2014). There is already so much uncertainty before us, say, by 2050...
Table 1 (Overview of global climate models): The amount of text in the manuscript is insufficient for the reader to understand the differences among models. The fact that the letters A, B, C, etc. are called "Numbers" and there is no Country and Institution associated with the first three models is not very engaging...
Section 2.2 (Methods): Here too, the text is very difficult to understand, especially the text and notations supposedly describing the equations on Page 5. There must have been problems with your text/equation editor in many places. For example, how can "2" be "the error variance [] constant at all data points, but generally unknown"?
Minor comments:
Several times, this reader could not tell whether "mode"/"modes" were typos for "model"/"models". See, for example, "the all modes multi-model ensemble (MME)" (Line 91).
The caption and content of Figure 1 are very poor.
Author Response
Point 1:Do you realize that you try to predict 86 years of annual precipitation (2015-2100) from 45 years of data/observations (1970-2014). There is already so much uncertainty before us, say, by 2050...（Lines 84-85）
Response 1:The article is ambiguous in expression and has been corrected to ’In order to verify the simulation accuracy of CMIP6 dataset in Qinling Mountains, 1970-2014 was selected as the historical evaluation period. The optimal model dataset was determined by comparing the simulated and measured values. The future simulation period was 2015-2100.’(Lines 89-92).
Point 2:The amount of text in the manuscript is insufficient for the reader to understand the differences among models. The fact that the letters A, B, C, etc. are called "Numbers" and there is no Country and Institution associated with the first three models is not very engaging...[Table 1 (Overview of global climate models)]
Response 2:Thanks for the advice. In the revised manuscript,' Numbers ' in table 1 be replaced with ' ID ' and that ' Country ' be deleted in the revised manuscript.The sources of the first three models are added to the new Table 1' Origin '.
Point 3:Here too, the text is very difficult to understand, especially the text and notations supposedly describing the equations on Page 5. There must have been problems with your text/equation editor in many places. For example, how can "2" be "the error variance [] constant at all data points, but generally unknown"?[Section 2.2 (Methods)]
Response 3:Special fonts in word cannot be displayed in overleaf, σ2 is not displayed in the submission and is now corrected in the new submission.Other functions have been checked and submitted to the revised manuscript.(Lines 135-141)
Point 4:Several times, this reader could not tell whether "mode"/"modes" were typos for "model"/"models". See, for example, "the all modes multi-model ensemble (MME)" (Line 91).
Response 4: In the revised manuscript, the article rechecks all statements about ' model ' a modifies and Modify the description of MME(multi-model ensemble)'.(Line 98).
Point 5:The caption and content of Figure 1 are very poor.
Response 5:Thanks for the advice.In the resubmitted manuscript, Figure 1 adds the geographical location and spatial distribution of precipitation in the study area.In addition, we rewrite the Abstract, Brief description of research data, purpose, method and work mean.(Lines 1-20)
Reviewer 2 Report
In figures give marker line in state of simple line
Overall the article is very good of unique work
Author Response
Thank you for suggesting that we modify the image format, formula symbols and unsmooth statements in the new manuscript.
Reviewer 3 Report
This manuscript investigated precipitation change over Qinling Mountains area in the 21st century under four SSP scenarios using nine CMIP6 models. The main findings are:
1. Precipitation will increase from 2015 to 2100
2. Under higher emission scenario, precipitation increase will have larger rate and east-west contrast.
3. The precipitation anomaly is related to abnormal transport of water vapor from the Northwest Pacific and the Bay of Bengal, corresponding to East Asian and South Asian summer monsoon strength.
I find this manuscript is a good regional study research. The analysis is clear, the conclusion is well-supported, and the manuscript is well-written. I would recommend it for acceptance after minor revision.
Major comment:
One issue I found needs to be further investigated and explained is the different trends and spatial variabilities between simulated precipitation trend from 2015 to 2100 (Figure 4) and precipitation change related to 1995-2014 mean observations (Figure 5). Figure 4 shows the linear trend from 2015-2100 is increasing for the entire region. Figure 5 shows that future precipitation is less than 1995-2014 mean observations in the south. In some area the precipitation change in Figure 5b is smaller than -100mm, which is roughly -10mm/10a. This is largely different with 18~19 mm/10a trend from 2015-2100. The domain mean change of precipitation in SSP5-8.5 is even negative from 1995-2014 to 2041-2060. Also, the east-west contrast of precipitation trend is not seen when comparing with 1995-2014 average. Is it because CMIP6 models have large bias compared with stational observations, or precipitation in 2014/2015 is quite different from 1995-2014 average, or is there any other explanation? The circle region (~108E, 33.3N in Figure 5) looks like related to interpolation effect from stational data. As stational data have small spatial representation, especially over complex mountain regions, I would suggest the authors use satellite data (e.g., IMERG or TRMM) for the regional analysis, and use stational observations to verify satellite measurements.
Minor comments:
Line 10-11: what are the four numbers corresponding to? the four emission scenarios?
Line 12: same as the comment above
Line 31: reference of the number (-11.95 mm/10a) should be put right after this sentence
Line 43-44: The numbers from this reference is confusing comparing with other numbers in this paper because this reference uses daily precipitation while others are annual precipitation. I recommend the authors either convert the unit to annual precipitation, or change the unit to mm/day/10a and the other unit to mm/year/10a.
Line 50: Reference is needed for “previous studies” based on CMIP5 models.
Line 79: is it a range (545-1155 mm) across time or space? IF it's across time, from which year to which year?
Figure 1: it will be good to include a larger domain map to show the location of Qinling in the region of East China/East Asia. It may also be useful to add a plot to show climatological spatial distribution (line 80) of precipitation in this region.
Line 83: Is there any reason why choosing these nine models but not the other CMIP6 models?
Line 132: is there a format problem here? What I see is number '2' is the error variance and is constant at all data points, but generally unknown.
Line 145-146: be careful to attribute model performance to resolution or parameterization schemes without any analysis or reference.
Line 159-162: how about the numbers (range and change rate) from BM dataset?
Figure 3: what do the shading refer to? the maximum and minimum value among the four BM models?
Line 180: even as a native Chinese, I don't know where Jialing River and Danjiang areas are located. it will be good to add a lat/lon information so readers can find the location on the map. Same as Heihe area and Ziwu River area later.
Figure 5: change the colorbar to center the zero line so the positive and negative regions can be better seen.
Line 219-235: good overview of large-scale conditions associated with precipitation anormalies
Line 220: is it 75%? in observation? how about contribution in CMIP6 models?
Line 245: do you mean upper-level divergence and low-level convergence?
Figure 6: It may be more informative to include both climatology state and anomaly in the figure. Since the patterns in different SSP scenarios are similar and the main difference is magnitude, one can choose to show only one scenario.
Line 279: The correlation between precipitation and South Asian summer monsoon is greater than correlation between precipitation and East Asian monsoon, why the authors conclude the region is affected by the EA monsoon? Moreover, how to explain that the sliding correlation between precipitation and SA monsoon in SSP5-85 is relatively small (Figure 7) but the overall correlation is large with high significant (Table 2)?
Figure 7: what is a "sliding correlation curve"? Is it correlation of 11-year timeseries around the analysis year? It would be good to add confidence information to show significant correlation indexes.
Author Response
Please see the attachment
Author Response File: 
Author Response.pdf
Round 2
Reviewer 1 Report
The newly submitted manuscript represents a mere/cosmetic revision compared to the original version, which is absolutely insufficient as none of the deficient fundamental aspects of the work that I had raised was addressed.
Author Response
Point 1:Unfortunately, I had a very hard time understanding what you did and how, and what the results of your work mean. My hard time started with the Abstract, which is simply unreadable in its present form, and could not be understood if it stood alone on the Web.
Response 1:We rewrite the Abstract and explain in the revised manuscript the significance of precipitation estimation in Qinling Mountains and the data, methods and main conclusions used in the study(lines 1-19).
Point 2:Do you realize that you try to predict 86 years of annual precipitation (2015-2100) from 45 years of data/observations (1970-2014). There is already so much uncertainty before us, say, by 2050...（Lines 84-85）
Response 2:In order to avoid ambiguity, we only introduces the historical period (1850-2014) and future period (2015-2100) of CMIP6 data simulation in chapter 2.1(Line 94), and introduces the method of selecting the optimal model using the measured data of Qinling Mountains in chapter 2.2.1(Lines 115-117).In fact, the CMIP data for 2015-2100 are not based on historical observations in the Qinling Mountains. Instead, researchers from different countries and regions use different CMIP models to estimate future global climate change by setting different radiative forcing scenarios and global development models.The historical simulation period of CMIP6 model is 1850-2014. In order to make the simulation more realistic, the model takes more account of observations, stalagmites and isotope records in simulating historical climate change.For compare the precipitation difference between the middle (2040-2060) and the end (2080-2100) of this century and the current period (1995-2014), the future research period is defined as 2015-2100.In the analysis of precipitation, the method of M-K test and confidence interval of correlation commonly used in CMIP6 are introduced to reduce the uncertainty of CMIP6 model in the estimation of precipitation in Qinling Mountains.
Point 3:The amount of text in the manuscript is insufficient for the reader to understand the differences among models. The fact that the letters A, B, C, etc. are called "Numbers" and there is no Country and Institution associated with the first three models is not very engaging...[Table 1 (Overview of global climate models)]
Response 3:In Chapter 2.1, we explain the reasons for the differences among models and why multiple models are selected as research data.(Lines 86-93).The nine CMIP6 models come from seven different institutions. The main difference of inter-agency models lies in the different physical processes, and the main difference of the models released by the same institution lies in the different atmospheric, oceanic and land surface parameterization schemes.In the revised manuscript,' Numbers ' in table 1 be replaced with ' ID ' and that ' Country ' be deleted in the revised manuscript.The sources of the first three models are added to the Table 1' Origin '.
Point 4:Here too, the text is very difficult to understand, especially the text and notations supposedly describing the equations on Page 5. There must have been problems with your text/equation editor in many places. For example, how can "2" be "the error variance [] constant at all data points, but generally unknown"?[Section 2.2 (Methods)]
Response 4:We rewrite the formulas and mathematical symbols in chapter 2.2.2 and modify the corresponding expressions according to the article by Fick ( Fick 1998 ).(Lines 128-145)
Point 5:Several times, this reader could not tell whether "mode"/"modes" were typos for "model"/"models". See, for example, "the all modes multi-model ensemble (MME)" (Line 91).
Response 5: In the revised manuscript, the article rechecks all statements about ' model ' a modifies and modify the description of MME(multi-model ensemble)'.(Line 100).
Point 6:The caption and content of Figure 1 are very poor.
Response 6:Thanks for the advice.In the resubmitted manuscript, Figure 1 adds the geographical location and spatial distribution of precipitation in the study area.
