Construction and Assessment of a Drought-Monitoring Index Based on Multi-Source Data Using a Bias-Corrected Random Forest (BCRF) Model

Round 1

Reviewer 1 Report

The manuscript named 'Construction and assessment of a drought monitoring index based on multi-smulti-source data using a bias-corrected random forest (BCRF) model' is a very meaningful work. This manuscript comprehensively considered the soil moisture information, the influence of extreme weather, and the changes in crop growth before and after drought stress. Combining this information with underlying surface environment data, this paper utilized a bias-corrected random forest model to construct a comprehensive drought index and to evaluate its accuracy. Based on the results, the temporal and spatial patterns of agricultural drought and the dominant climate variables of the crop growing season in Northeast China were discussed with scientific significance and innovation. However, there are still a few details to be ponder in this paper, before it can be published. Below are my major concerns and detailed comments:

Firstly, agricultural drought in Northeast China is affected by various factors, such as topography, planting structure and farmland environment. A single index is indeed difficult to accurately describe the actual drought situation. The new index constructed in this paper reflects the scope and degree of drought stress to a certain extent. Here, it is recommended to increase the comparison experiment. Is there randomness or contingency in RFSDI only doing accuracy verification with a specific index? It is recommended to supplement the discussion with other meteorological or remote sensing indices (such as SPI, CWDI and VTCI) or to compare with the drought monitoring accuracy of the independent variable index mentioned in this paper to further prove the drought monitoring accuracy and effectiveness of the RFSDI index.

Secondly, the manuscript concludes that there are differences in the dominant climate variables of agricultural drought in different months of the crop growth season in Northeast China, which is an important conclusion drawn. In section 4.3, they focus more on the cumulative impact of multiple climatic factors on agricultural drought. It is suggested that increasing the discussion of agricultural drought-dominated climatic factors in different months on the occurrence of agricultural drought.

There are some details in the manuscript:

1. In lines 57-58, it is recommended to delete the phrase 'In recent years, frequent droughts worldwide have serious negative impacts on agricultural production and socioeconomic development due to climate change.' as it is already mentioned in the first paragraph.

2. In line 94, It is suggested that the main models of remote sensing drought monitoring should be described before, followed by the corresponding monitoring indexes or methods of three kinds of remote sensing monitoring models, respectively.

3. In the introduction, it is recommended to provide a related elaboration on the distribution of crop planting areas, topographic conditions, climatic conditions and planting systems in Northeast China, and then lead to the complex geographical factors, planting systems and climatic factors. A single remote sensing index has been difficulty accurately describing the occurrence of agricultural drought. The paper should also emphasize that it comprehensively considers  the interaction of soil, vegetation, atmosphere and other factors, constructs a drought index, and more accurately identifies agricultural drought.

4. Line 170 should be revised to more accurately describe Northeast China, and it is recommended to change 'Inner Mongolia to the eastern part' of the 'Inner Mongolia Autonomous Region'.

5. It is suggested that the paper provide a brief introduction to the topography in the study area, including information on the main planting areas and characteristics of drought distribution and occurrence in Northeast China, particularly in the main planting areas of maize and rice.

6. Line 192 Clarify the meaning of 'Per'.

7. In Section 2.2.3, it would be helpful to include the sample points, survey routes, or areas. Field survey data are suggested to be reflected in 'figure 1a'.

8. In line 353, line 378 and line 373 are partially described, please shows the results obtained by comparing RFSDI with SPEI at what time scale.

9. In lines 398-399 '3.4 Research on drought-dominating climate variables during the crop-growing season, 3.4.1. Multi-scale characteristics of drought from 2003 to 2020.' There are logical problems, in general. The description of 400-412 lines is more in line with the use of 'Multi-scale characteristics of drought from 2003 to 2020' as the title. This section suggests deleting 398 lines and changing 3.4.1 to 3.4.

10. Section 4.2 Suggestions are divided into two parts. The first part is the discussion of the spatial and temporal pattern of agricultural drought in Northeast China, corresponding to section 3.3 of the results of this paper. The second part is the discussion of the drought trend in different agricultural areas, corresponding to 3.4 of the results of this paper. At the same time, combined with the relevant conclusions obtained in this study, combined with the relevant conclusions of the predecessors, the logic will be clearer.

11. Table 1: Spelling errors of some words: 'Developmenntal' is modified to 'Developmental'. It is recommended to check the spelling of the full text after the revision is completed.

12. Whether the '3.3 The spatio-temporal pattern of drought in NEC based on the RFSDI' considers the frequency of drought occurrence or the percentage of drought stress under different altitude conditions, soil texture, and slope environments, which will more clearly reflect which farmland environment has a higher frequency of drought.

13. Using a consistent format for expressing time intervals throughout the text. In lines 411, 412 and 452, 'from 2010-2020' and other similar issues, it is recommended that the full text be unified and expressed as 'from 2010 to 2020'.

14. In lines 154-163, the expression is repeated with the above paragraph, modified as appropriate, and ressorted in this part.

15. In lines 276-277, it is recommended to increase the interpretation of the relevant parameters in Formula 3-4.

16. In line 358, where does the threshold for drought determination and nondrought determination come from? It is recommended to quote relevant literature.

17. Making appropriate adjustments to improve the clarity of Figure 5f legend, horizontal and vertical axes of Figure 10, and the linear equation in Figure VII. The chart in Table 5 should be self-evident.

 

English edit is required.

Author Response

Please see the attachment.

Author Response File: Author Response.doc

Reviewer 2 Report

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.docx