Predict Seasonal Maximum Freezing Depth Changes Using Machine Learning in China over the Last 50 Years

Round 1

Reviewer 1 Report

Dear Authors, 

In this study, were employed machine learning model MLM with reanalysis data and observations seasonally frozen 148 ground regions covering 12 provinces in China from 600 meteorological stations over 50 years to predict the annual maximum freezing depth (MFD) in China. The model accurately predicted MFD values and revealed a decreasing trend in freezing depth over time. Changes in the distribution of freezing depth ranges were also observed. Machine learning proves effective for studying and predicting changes in MFD, particularly in the absence of recent data on seasonally frozen ground in China. The research necessitates revisions and modifications that have the potential to significantly impact and benefit readers within this particular field.

(Attached is a file with all comments)

 

Comments for author File: Comments.pdf

In some instances, there may be a need for greater consistency and accuracy in the usage of articles: "a," "an,"  "the." , "such as.", and "a few.". There exist several occurrences of repetitive phrasing or word choice that could be diversified in order to improve the readability of the text.

The clarity and comprehensibility of the English language in the text are generally satisfactory. However, attending to these minor concerns would further augment the writing's quality.

 

Author Response

Dear professor and editor. Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

I received a manuscript entitled 'Machine Learning to Predict Changes in the Seasonal Maximum Freezing Depth in China over the Last 50 Years' for review. I undertook the review with the caveat that I am interested in ground freezing, but I am not an expert in modelling and machine learning.

I read the manuscript with attention and great interest. I note that the text is polished. Perhaps it has been reviewed by others before, or the authors have much experience publishing research results. I am positively surprised by the quality of the ‘introduction’ chapter. It is tough to find any shortcomings here. The authors are aware of the limitations of the modelling carried out. They also present well the factors responsible for the ground freezing process. The modelling based on spatial data allows for a very interesting summary. To improve the material presented, I recommend referring to a few remarks, which I list below:

1. Unfortunately, The hypsometric map shown in Figure 1 is not in accordance with cartographic principles. I understand that the authors intended to indicate areas of high elevation in ‘cool’ colours, which are supposed to be associated with ground temperature. However, this isn't easy to accept. The map needs to use a consistent hypsometric colour scale. The Tibetan Highlands, the Karakorum or the Himalayas are currently associated with the seabed. The perception of the human eye makes fields with cold colours (blue) appear further/lower than fields with warm colours.

2. Chapter 3 discussed the factors responsible for ground freezing, summarised in Figure 2. Unfortunately, it is not clear exactly what factor the DEM represents. Presumably, it is the height above sea level and should be described as such. Just so you know, the DEM can also calculate the aspect, slope, or incoming solar radiation.

3. Among the factors taken for modelling, my concern is always the quality of the spatial soil data. This factor's variability is so significant that local conditions can differ significantly from the class assigned to a vector field or raster. I think this can be commented on in the discussion section – which factors are specific and result from an approximation of reality. The thickness of the snow cover is also shown, but after all, more important than the thickness is the snow cover's duration, in the snow cover in the continuity period of that snow cover.

I know that many compromises have to be accepted for modelling purposes. I, therefore, appreciate the authors' cautious approach to many issues. I suggest authors refer to the research in the SFG field, like "Seasonally Frozen Ground"

I wish you the successful completion and publication of these exciting research results.

Author Response

Dear professor and editor, please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Dear Authors,

Thank you for addressing the review comments and making the necessary revisions to your manuscript.