General comments:

The authors took two surface layer schemes (Noah-MYJ and Noah-MP) out of WRF, and simplified them as stand-alone modules to evaluate their performance for temperature inversions over forests in the Arctic winter. Additionally, a conceptual model was also developed to investigate the impact of individual variables. To correct the limits of the WRF schemes on turbulent collapse, some modifications were inserted. The research provides some ideas on improving the surface layer models, especially under stable conditions. However, I found the structure of the manuscript is not well organized. The interpretation of the results is limited too. So, I would like to recommend a major revision before it can be published in this journal.

Major comments:

1. The title is misleading, since you didn’t evaluate the surface layer schemes inside WRF. Especially, what’s the difference of the surface layer models in the version 4.5.1 compared to the previous versions?
2. In section 2 the authors first introduce the conceptual model, then the two schemes from WRF, and then in section 3 the modified schemes are described after the measurements. I feel this organization is not straightforward and confusing. The connection between the conceptual model and WRF schemes is not clear.
3. The validation of the models (Figure 7) is based on a lot of input parameters regarded as constant values; however, these parameters should change with time. It’s difficult to draw the conclusion that the modifications improve the model performance. This part should be expanded to gain more confidence.
4. The investigations on the model results are limited, especially the last part with all measurements input. The discussion should be extended.

Minor comments:

1. The full name of WRF should be mentioned somewhere.
2. Line 56: LMDZ model should be explained.
3. The language needs to be improved. There are some spelling and grammar mistakes. For example, line 227&404: a comma should be inserted before which. Line 401 is confusing.