Estimating Global Anthropogenic CO2 Gridded Emissions Using a Data-Driven Stacked Random Forest Regression Model
Round 1
Reviewer 1 Report
The manuscript submitted by Zhang et al. estimates global CO2 emissions by using a machine learning model. The paper is well organized with relatively adequate discussions. However, there are several more serious flaws that need to be revised before it can be published further in the journal. Specifically the following points are included.
1) Regarding the validation of the results: Was a rigorous cross-validation method used, such as dividing the input data into a training set, a validation set, and a test set? The model input used ODIAC data, which is not suitable to be used again for validation. A third-party carbon emission dataset, such as the Carbon Monitor (https://www.nature.com/articles/s41597-020-00708-7) or other carbon emission inventory data should be considered.
2) Regarding the importance of model input features: why add Lat variable, how to explain its such high importance, and is it necessary? The importance of satellite observation of CO2 is very low for the model, try to discuss it. Is there overfitting of some features or the model itself, suggest to add some necessary details of corresponding model diagnosis.
3) It is suggested to add validation results and characterization about the estimated seasonal and annual variation of CO2 emission.
Author Response
Thank you for your valuable suggestions for our manuscript. The manuscript has been modified based on your comments and our specific replies to each point were laid out within the file attached.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Dear authors,
First of all, congratulations on drafting this very interesting work. Most importantly, I like the fact the the findings were extrapolated at the global scale, with focus on specific parts of the World. The methods are well described and the findings are well supported, in my humble opinion.
I have laid out some comments, suggestions and few additions to the manuscript. Please find them enclosed within the file attached.
I recommend a minor review of this work.
Comments for author File: Comments.pdf
Author Response
Thank you for your recognition and valuable suggestions for our manuscript. The manuscript has been modified based on your comments and our specific replies to each point were laid out within the file attached.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The authors have addressed well my previous concerns, and the revised paper can be published in the journal.
