A Drought Index: The Standardized Precipitation Evapotranspiration Irrigation Index
, Tianao Gao 4, Yanan Ding 5,*, Yibo Ding 1,2, Yiyang Zhao 6 and Wei Fan 1,2
Round 1
Reviewer 1 Report
The paper is well done, ready for the publication I guess.
Author Response
Thank you for your high evaluation of this study.
Reviewer 2 Report
The authors propose a new drought index (SPEII) that accounts for the main human water use source, i.e. irrigation. The new index is then compared to established ones (e.g., SPEI) over China, thus showing an improvement in the reproduction of the vegetation drought especially over heavily irrigated areas.
My main concerns are about the formulation of the SPEII. When calculating the SPEI, the water deficit is properly defined as the difference between precipitation and PET, namely between the water supplied and the water required. In the SPEII formulation, the actual instead of the potential ET is used, why not still PET? If the sum of precipitation and irrigation is lower than PET, the crop can still evapotranspirate at a rate that is lower to the potential one. Also, at which temporal step is eq. (17) implemented? And what is the source of irrigation data? Furthermore, the reason for splitting the case of Kc data available or not is not clear to me (lines 232-233). It seems that if kc data is not available, irrigation is calculated as the discrepancy between precipitation and ET, which is reasonable but it is not real irrigation. Again, if kc data is available (the other case) where does the irrigation information come from? Without clarifying these aspects, it is difficult to interpret the results.
Please find some additional comments in the following:
Lines 108-111: I would talk about land cover types rather than “vegetation”.
The irrigation fraction comes from the FAO’s GMIA. The data appear in Figs 2 and 3, but the source is mentioned later in the text. Also, it is not exactly the portion of irrigated area for each pixel but the portion of area equipped for irrigation, which is pretty different (see, e.g., Ozdogan et al., 2010, DOI: 10.3390/rs2092274, Dari et al. 2022, https://doi.org/10.1016/j.advwatres.2022.104130 , Elwan et al., 2022, https://doi.org/10.3390/w14050804). Nevertheless, I think that analysing the results at different irrigation levels is very useful to assess the uncertainties linked to the adoption of such a data set.
La bels of Figs. 5 and 6 are hard to read.
Author Response
We would like to thank Reviewer #2 most sincerely for his/her constructive comments and valuable suggestions, which have helped us to thoroughly revise and improve the manuscript. Below please find our responses to reviewer #2’s comments, and we have also incorporated them to the revised manuscript.
Reply to Reviewer #2’s comments:
The authors propose a new drought index (SPEII) that accounts for the main human water use source, i.e. irrigation. The new index is then compared to established ones (e.g., SPEI) over China, thus showing an improvement in the reproduction of the vegetation drought especially over heavily irrigated areas.
- My main concerns are about the formulation of the SPEII. When calculating the SPEI, the water deficit is properly defined as the difference between precipitation and PET, namely between the water supplied and the water required. In the SPEII formulation, the actual instead of the potential ET is used, why not still PET? If the sum of precipitation and irrigation is lower than PET, the crop can still evapotranspirate at a rate that is lower to the potential one. Also, at which temporal step is eq. (17) implemented? And what is the source of irrigation data?
[Reply] Because PET (potential evapotranspiration) was overestimate actual evapotranspiration (ET) in calculating evapotranspiration of crops. Meanwhile, the purpose of this study developed a convenient drought index to monitor agricultural drought in region. Therefore, ET was more appropriate than PET to develop a drought index for monitoring agriculture. The equation (17) was calculated after equation (18 ) and (19). Based on GMIA (Digital Global Map of Irrigated Area), the irrigation data was estimate by equation (19).
- Furthermore, the reason for splitting the case of Kc data available or not is not clear to me (lines 232-233). It seems that if kc data is not available, irrigation is calculated as the discrepancy between precipitation and ET, which is reasonable but it is not real irrigation. Again, if kc data is available (the other case) where does the irrigation information come from? Without clarifying these aspects, it is difficult to interpret the results.
[Reply] Kc data was avail abled from Irrigation and Drainage Engineering[1]. Irrigation and Drainage Engineering showed the Kc data of winter wheat and summer corn in Shaanxi, Shanxi, Henan and Hebei Provinces. Northeast China Plain of Major crop-producing areas was composed by those provinces. Meanwhile, Kc data in Irrigation and Drainage Engineering come from field experiments results. Therefore, Kc data was reliable and accurate. The irrigation data was estimated based on water deficit and Kc data. Because, the purpose of this study developed a convenient drought index to monitor agricultural drought in region. We sacrificed some degree accuracy of irrigation data for convenience to apply in region. However, the results showed SPEII had relatively great performance in monitor regional agricultural drought. Moreover, we explained some limitations of SPEII in section 4.1.
References
[1] Wang, Z., Irrigation and Drainage Engineering. China Agriculture Press: BeiJing, 2010.
3.Please find some additional comments in the following:
Lines 108-111: I would talk about land cover types rather than “vegetation”.
[Reply] We have revised it in Line 109.
- The irrigation fraction comes from the FAO’s GMIA. The data appear in Figs 2 and 3, but the source is mentioned later in the text. Also, it is not exactly the portion of irrigated area for each pixel but the portion of area equipped for irrigation, which is pretty different (see, e.g., Ozdogan et al., 2010, DOI: 10.3390/rs2092274, Dari et al. 2022, https://doi.org/10.1016/j.advwatres.2022.104130, Elwan et al., 2022, https://doi.org/10.3390/w14050804). Nevertheless, I think that analysing the results at different irrigation levels is very useful to assess the uncertainties linked to the adoption of such a data set.
[Reply] Thank you for your references. We have moved these Figures behind the data source (see Lines 187-198). And we also agree with you in irrigation that irrigated area was different from irrigation level in field scale. However, the irrigation degree was different irrigation level. Firstly, irrigation level was considered as similar in different region of China. Secondly, the grid cells were considered as a unit in this study. Therefore, the irrigation degree was considered as an irrigation water amount in cropland in a grid cells unit. Because, the purpose of this study developed a convenient drought index to monitor agricultural drought in region, we might not very accurate to calculate the actual irrigation amount in the field. Moreover, the results showed SPEII had relatively great performance in monitor regional agricultural drought. Moreover, we explained some limitations of SPEII in section 4.1 (see Lines 523-537).
- Labels of Figs. 5 and 6 are hard to read.
[Reply] We have revised it. Please see the Labels of Figs. 5 and 6.
Author Response File: 
Author Response.docx
Reviewer 3 Report
Please see my comments in the attached file.
Comments for author File: 
Comments.pdf
Author Response
We would like to thank Reviewer #3 most sincerely for his/her constructive comments and valuable suggestions, which have helped us to thoroughly revise and improve the manuscript. Below please find our responses to reviewer #3’s comments, and we have also incorporated them to the revised manuscript.
Reply to Reviewer #3’s comments:
The objectives of this study were: "(1) to develop a new drought 90 index considering irrigation factors based on traditional drought index framework; (2) to compare the performance the SPEii, SPEI and scPDSI in cropland vegetation response to drought; and (3) to evaluate drought change and provide some strategies for regional irrigation and water resource scheduling management in the future based on CMIPS models." It is known that droughts are not only associated with climatic factors, but can be exacerbated by human actions. The novelty of the work lies in the incorporation of the human factor, in this case the irrigation component to compose a more suitable indicator for droughts. The work methodology is robust and well explained. Results and discussions clearly show the advantages of using "socio-climatic" indicator. However, the authors could improve the conclusions, to be more compelling and assert ive. I recommend the publication with minor revisions.
Specific comments and suggestions:
- Keywords: include "SPEII index"
[Reply] We have revised it (see Line 34).
- Please increase the font size of the figures. Many of them are too small and therefore illegible.
[Reply] We have increased the font size of the Figure 5, Figure 6, Figure 9 and Figure 10.
- Page 3, line 100: correct for 9.63 x 106 km2
[Reply] We have revised it (see Line 100).
- Page 3, line 122: Please add accessed date; Date Month Year, the exact date when you last accessed the link. Do this for all links.
[Reply] We have revised it for all link in section 2.2.
- Page 4: Table 1 shows 8 crop-producing areas, but the authors mention nine (9). The authors consider NCP or NEP?
[Reply] We have revised it (see Table 1).
- Page 5, line 149: "The study area were covered by 826…11 to "The study area is covered by 826 meteorological stations (Figure lc)".
[Reply] We have revised it (see Line 151).
- Page 9: Figure 4 "The difference value between the vegetation response times to the SPEii and SPEI (a minus b) (f), correct to " … (d minus e) "
[Reply] We have revised it in Figure 4.
- Page 11: Why did the authors used irrigation levels intervals in Figure 6 (four levels) instead using the same as in Figure 5? For comparison purpose is better use the same levels.
[Reply] We have revised it in the same level (see Figure 5, Figure 6 and Line 361-364).
- Page 11, lines 244-348: what is the percentage of the area with correlation coefficient different less or equal to zero? And what are the irrigation levels for this area?
[Reply] We revised it in Lines 290-296. Meanwhile, Fig. 4c showed that the SPEII mainly had higher correlation coefficient areas than SPEI in eastern China. However, SPEI mainly had higher correlation coefficient than SPEII in western China. Meanwhile, the correlation coefficient difference had 37.1%, 10.5% and 52.3% areas more than 0, less than 0 and equal 0, respectively. Only 0.35% of the correlation coefficient difference areas less than -0.01. By comparing Fig. 3 and Fig. 4c, we found that SPEII had higher correlation coefficient than SPEI in irrigation area of China. However, SPEII had lower correlation coefficient than SPEI in non-irrigation area of China.
- Assessing the frequency and intensity of droughts in the future using the SPEII can "mask" changes and trends due to the use of irrigation. In this case, would not it be better to use the SPEI to evaluate only in terms of changes associated with climatic variables?
[Reply] The purpose of this study developed a convenient drought index to monitor agricultural drought in region. Meanwhile, SPEII also use to monitor future agricultural drought under considering human irrigation activities. To combine climatic variables and SPEI might have more performance monitor agricultural drought (not considering human irrigation activities). This might be another interesting research. Therefore, we will consider this problem in future research.
Author Response File: Author Response.docx
Reviewer 4 Report
Can you propose classification range for the new drought index
Specifically in which context/type of region we need to consider this index ?
Is it sensitive to the issues related to the component indices .. like SPEI/NDVI/PDSI ..if so to what extend
How can you incorporate the non stationarity of the new index ?
(these are some questions which can be addressed; but not mandatory to respond)
Author Response
We would like to thank Reviewer #4 most sincerely for his/her constructive comments and valuable suggestions, which have helped us to thoroughly revise and improve the manuscript.
Reply to Reviewer #4’s comments:
Can you propose classification range for the new drought index Specifically in which context/type of region we need to consider this index? Is it sensitive to the issues related to the component indices. like SPEI/NDVI/PDSI. if so to what extend? How can you incorporate the non-stationarity of the new index?
(these are some questions which can be addressed; but not mandatory to respond)
[Reply] Thank you for your high evaluation of this study. The SPEII was calculated by SPEI. Therefore, the classification of drought index was consistent with SPEI. However, we did not consider the sensitivity of SPEII to the component and non-stationarity of the SPEII. This might be another interesting research. Therefore, we will consider this problem in future research.
Round 2
Reviewer 2 Report
In my opinion, there are still some pending issues concerning my previous comments.
The authors are using crop evapotranspiration, which can be higher than PET (for Kc > 1 ). Irrigation is calculated as the residual of ET with respect to rainfall (eq. 19). The authors state in the manuscript (lines 232-235):
“??? could be expressed by Equation 18. We had ?? data for winter wheat and summer maize in typical crop-producing areas from Irrigation and Drainage Engineering [57]; however, when ?? data were deficient or insufficient, ??? could be estimated by ????:”
Eq. 19 comes after this sentence. It is still not clear to me what happens if kc data is not reliable/sufficient. Do you compute Irrigation as rainfall – PET? If so, please clarify. Also, I suggest mentioning the time step at which irrigation is computed (daily, weekly, monthly, …).
I suggest mentioning in the manuscript what specified in previous comment n.4, namely the difference between actually irrigated area and area equipped for irrigation (with associated uncertainties) should be pointed out.
Labels of Fig 5 are still hard to read. The title of Section 3.2 is in the caption of Fig 6.
Author Response
We would like to thank Reviewers most sincerely for his/her constructive comments and valuable suggestions, which have helped us to thoroughly revise and improve the manuscript. The attached word is our reply to the reviewers' comments, and we have also incorporated them to the revised manuscript (RM).
Author Response File: Author Response.docx
Round 3
Reviewer 2 Report
The authors keep on canging the caption of Figs 5 and 6, but I was asking for simply making the labels bigger. The paper can be accepted after this very minor change without nedd for further review from mi side.
