Dear Authors,

I am suggesting that the manuscript is published once the further recommendations and queries raised by referee 2 have been addressed:

'I have a major concern regarding the use of only one LST image for model evaluation. Former model evaluations that used LST tried to acquire as many images as possible for one period. The argument that Landsat 8 images is more representative of the heterogeneity of the surface urban climate is receivable but not sufficient. These heterogeneities are also captured by MODIS data for instance, although smoothed by the lower resolution. Bechtel et al. 2019 who did a somparative study of LST in different cities using LCZ, MODIS and Landsat data found out that there was a great spatial correlation between the two. I believe they should be used jointly. Hu et al. (2014), Wouters et al. (2016) and Brousse et al. (2020) also used MODIS LST for evaluation of their model simulations. I would therefore like the authors to include such additional evaluation before acceptance of the manuscript.

Finally, I still have some questions which I would like to see answered before accepting the manuscript for publication:

1. I still did not understand the rationale with having air temperature measured at the Pinggu station and upwind from the airport. The reasoning with the wind measurements being WMO standard makes sense. But then why not simply use the air temperature from the airport itself as it can be considered as rural too?

2. The implications on the results from adapting the surface resistance to evaporation for LCZ 4 6 and 9 need to be discussed.'

Best wishes,
Lisa

Dear Michael,

I appreciate your response to the reviewer’s first query against the use of the coarser MODIS data for a model comparison. I am suggesting, however, that some reference to the previous research where MODIS data has been used for model evaluations should be added to the manuscript – perhaps around line 86 – to highlight to the reader that MODIS data can be used for model evaluations.
Once these references are included, I would be happy to recommend that the manuscript is published.

Kind Regards,

Lisa