I sent this out for a second round of review and received the following comment from Reviewer 1. Please address the comments and provide an updated draft of the paper. The reviewer also provided a marked up version of the paper with suggestions about how to improve the text with regards to clarity. See the attached document.

Finally, I think that some comments are warranted on whether or not the conclusions of the paper are impacted by the use of the relatively simple SURFEX model and any biases in surface energy fluxes that can be attributed to the use of a simple land model. Other regional modeling / forecasting systems include more advanced land models, so the results should be put into the context of the wider spectrum of modeling systems.

Reviewer 1 comments:

I would like to thank the authors for addressing my concerns raised in the first review of this manuscript. There has been a substantial revision of the scope, a more comprehensive description of the experimental design and an appropriate level of restructuring to improve the overall narrative. There are a couple of points that I think still need to be addressed:

1. While the reporting of the results has vastly improved there is still very limited discussion on why certain results are obtained with respect to particular simulation configurations. For example why does FS simulate better temperatures during summer? Is this because this simulation benefits from soil moisture memory by allowing the land surface to be fully interactive? The reduced performance in all configurations for precipitation is explained by the wet bias of the forcing data but perhaps more needs to be said on why FS is particularly worse than DRI and CON in winter.

2. I was really concerned about the massive bias in the ground heat flux. I think the authors were too quick to dismiss this and should say more about why this bias persists particularly because it ranges from 40 to >100 W m-2. In particular, is the ground heat flux calculated or updated to include the residual energy imbalance (e.g. G = RN – H – LE) in order to maintain energy balance…because it certainly looks like it is. Perhaps something needs to be said on what steps will be taken next to improve the surface turbulent energy fluxes and their partitioning because the biases for the surface energy balance are quite large. It looks like including (or improving) a soil resistance for soil evaporation and a stomatal resistance for transpiration may improve the excessive evaporation in the model.

3. In the first review I suggested to augment the validation of the surface energy balance with gridded observational products such as GLEAM or LandFlux. I’m not convinced by the reason given as to why this request was dismissed. It is difficult to evaluate the skill of the land surface fluxes at the grid cell values from a model that represents here a 20 km2 grid cell average to point observations, particularly if the vegetation ‘mix’ contrasts to that of the point observations. The biases are quite large and it would be good to check if this is associated with comparing grid cell averages to a point observation and whether the model can at least capture the spatial variability of the domain. Comparing against two points is not very rigorous and given how large the biases are, avoiding comparison to products such as the MERRA reanalysis, GLEAM or LandFlux really undermines the credibility of this model validation study. Could you look at how well the model captures the anomalies to assess more critically the ability of the model to simulate the temporal variability?

4. There are still sections of text where the language / sentence composition is awkward. Rather than note them all here I attach these comments in an annotated PDF of the revised manuscript. These are my suggestions for further refinement of the manuscript text and include some minor requests for further clarification.

Thank you for your thorough revisions to the paper. The revised manuscript is acceptable for publication.