Feasibility of Calculating Standardized Precipitation Index with Short-Term Precipitation Data in China
Round 1
Reviewer 1 Report
Dear authors,
Please find attached my review report. I hope you find them useful.
Regards.
Comments for author File: 
Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
The manuscript entitled ‘Feasibility of Calculating Standardized Precipitation Index with Short-Term Precipitation Data in China’ concerns quite an important methodological topic of precipitation totals measurement using both types of data series, i.e. the long one (traditional) and the short (automatic). In fact, the issue has been examined on the basis of the well-known Standardized Precipitation Index (SPI) calculated for the large territory i.e. China. In my opinion the work as well as the topic is quite important and well presented. In a broad sense, the topic is relevant as the knowledge about changes in climate conditions and important to mitigate and adapt to risks. However, there are some points which should be clarified and added. There are as follows:
The Authors use the impressive large number of measuring stations (over 50 000 precipitation posts!) for the period of 2015-2019. The dimension of the network composes a unique and solid basis for further analyses. But the Authors should present the information about the type of range gauges and the estimation of the measurement quality. Having an experience with such type of measurements I know that some data could be suspicious and simply not homogenous due to technical problems. Please add some comments as well as general information (e.g. add a table) about the geographic location of the stations with the relevance to the altitude (simply: number of stations in the particular vertical zones). It is quite important when you analyze and compare of the E and W territory of China. In the line 108 you must add what kind of monthly precipitation data (which variable) has been taken into account (precip. totals, number of days?).
Line 153: add citation of the author of the table.
Chapter 2.3. Please explain and comments on why you have chosen such 3 calculating methods.
Line 211-212: It is not clear/visible; please reformat
Line 230 and further: There is a clear information (statistical) about SPI station differences including some figures and instructive maps. But there is no broader comment of the causes. Please try to inform what are the reasons of these differences, especially between E and W which you mention many times. It will be very valuable to give their physical explanation.
Figs. 9-11. Please explain why you used Sept. 2019 for comparison. It is hard to see it from the text.
Summary and discussion is well formulated. However, it is mainly devoted to the statistical aspects. Meteorological/climatological explanation is really missing here. Please add the special paragraph concerning that issue.
There is missing some quite important and relevant recent articles devoted to the regional drought aspects e.g. for Europe (SPI, SPEI, e.g. see Advances in Water Resources). You can use few of them for the short scientific comparable discussion between some regions.
There are small language errors and misprints (e.g. should be indices instead of indexes). Please check carefully references and citations (e.g. DeGaetano).
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
The manuscript explores the possibility of obtaining monthly SPI values for stations with short data series using data from stations with long observation series to gain more detailed spatial information about the extent of drought in China. Overall, the paper is interesting due to the importance of the subject in the light of climate change impacts, and I think that it has the potential for publication in Atmosphere after major revisions, considering the list of comments below:
Line 15:
‘Then result shows that it is feasible for short-sequence …’ Please correct this for - The result shows that it is feasible for short-sequence …
Line 55-57:
The sentence occupying these lines seems to be unfinished. I would finish it with the wording 'based on them'.
Line 174:
Is this a self-developed method? If not please provide a citation. Looking at the schematic presentation of the method, the question arises how the diameter of the circle, which determines the set of stations that will take the average values of distribution parameters from N stations, is selected.
And how is the problem solved when the stations are in the common part of two circles?
Line 183-184:
What type of kriging was used? Please describe in more detail. Please also complete the citation.
Line 191-195:
Based on the description of lines 191-195 the process of verification of the used methods is not fully understood.
I agree with counting the SPI values for all stations in the long data series to obtain the real parameters of the gamma distribution (for comparative and error analysis). In the next step, stations with long data series should be randomly selected from the set of stations on the basis of which the analysis will be performed (I would suggest in a similar proportion as stations with long data series and short data series - 2416/51.062). The gamma distribution parameters for the selected stations should later be used to obtain SPI values for the remaining long data series stations.
Was a division made among stations with a long data series? If so, please provide on the basis of parameters from how many stations with long data series were determined SPI values using indirect methods (at the verification stage when only stations with long data series were used).
Line 243:
Please correct Figure 5 (a). I suggest varying the line thickness (starting with the one at the bottom) from thickest to thinnest. This will make the plot more readable.
Line 213-216:
Please correct the style
Line 220:
In the sentence: 'To ensure the accuracy of the SPI and reduce the computational complexity when using the indirect methods,...' . I would change to:
'To ensure the best possible accuracy of the SPI and reduce the computational complexity when using the indirect methods,...'
Line 244:
I would suggest changing the Figure 5 caption. E.g.: Comparison of real SPI values for Fangxian Station with values obtained using different indirect methods. (a) SPI time series; (b) SPI-SPI plot.
The wording 'size comparison' does not fit. We simply compare the index values.
Line 343-344:
The sentence occupying these lines seems to be unfinished.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Dear authors,
Thank you for the detailed response to my comments. I am glad my suggestions helped improve the results. There are a few minor grammatical errors in the manuscript. Kindly revise the manuscript before the final submission.
Good luck!
Reviewer 3 Report
After the author's explanations and after reading the manuscript, I think that the authors have done significant changes. In its present form, the manuscript can be accepted.
