Dear author, co-authors,

I have checked your revised ms and response to the review provided by Bart van de Hurk. Unfortunately, despite various reminders, at which I received a positive response that the second review would also be provided, I never received the second review and consequently provide this editors review on your ms. Overall, the ms reads well and adresses a interesting study indicating on the relevance of LULCC in climate change aiming to differentiate between the biogeophysical and biogeochemical effects on climate. It also shows the issues on how to arrive at a consistent consideration of this features in the ESMs and the complexity of really being able to fully appreciate the results with these complicated model systems. I am mentioning this since it is challenging as a reader to keep fully track of some of the subtle differences that might explain some of the reported signals.

Here my main (but generally minor) comments:

You modified the abstract in response to the comment raised by the referee; Since here the term storages refers according to me to CO2 uptake whereas it is actually increases/decreases I would suggest to change this to: “Modifications of the land carbon pool by LULCC are disentangled in accordance with processes that can lead to increases and decreases in this carbon pool”

Line 51: “...land–atmosphere energy and water exchange.”

Line 175: It really took me some time again to carefully read the explanation of the different simulations you have done. I am convinced that you have made a maximum effort to find an optimal way to do so but think that any reader will have a hard time to easily understand what has been done going through all the selected acronyms. I suggest to change the main sentence that at the end summarizes and explains it well removing some more of these acronyms: “The difference between the ESM and L1A simulations therefore yields the biogeophysical effects of LULCC on climate (dTBGP) whereas the difference between the L1A and L1B simulations yields the biogeochemical effects of LULCC on climate (dTBGC). And is it not an idea to, instead of using the reference L1A and L1B to refer to the BGP and BGC simulations?

Line 255; you added here the statement that “The modification of local BGC-induced temperature signals leads, for example, to a warming in all models in Australia where trees have been replaced by pastures (Fig. 1b).” I would suggest to add here that this is a clear example that despite replacing the trees by pasture, which results in a simulated increase in albedo, there is a simulated warming due to the BGC effect.

Line 266: I had to read a couple of times through this explanation why your signals are less robust compared the Pitman result. You state that “In the RCP scenario, the area undergoing LULCC is relatively small and is mainly located outside regions with strong snowmasking effects, unlike in the past. BGP cooling..”. I would rephrase this explaining more clearly that in the past LULCC BGP effects were more dominant in the mid- to high latitude regions, mainly associated with changes in albedo including changes in snow cover, the latter being less relevant in the analysis of future LULCC BGP effects where the LULCC is mainly found in tropical regions and where the albedo changes are counteracted by simulated decreases in evapotranspiration in the tropics.

Line 286: “While this holds for all models..”

Line 299: “but also increases the snow-masking effect in periodically snow-covered regions.” I checked again what is exactly meant with an increase in the snow-masking effect and where this is introduced in the beginning of the paper referring to the work by Claussen et al. (2001). I infer that this effect is referring to the fact that with an increase in forests, the impact of snow cover on albedo is reduced due to the fact that the snow is mainly found under the forest canopy. It might be worthwhile to shortly mention this in the
beginning of the ms to explain well what is meant with this effect.

Line 318: “is mainly attributed to enhanced changes in South America, Africa and Australia. These might be related to changes in latent heat fluxes or cloud cover.” Have to admit that this doesn’t make a strong impression in terms of really understanding the models signals. Is there a possibility that this explanation for having larger changes in these regions can be really corroborated?

Line 342: Suggestion to change to “which emphasizes the importance of including pastures in the model simulations, and being another source of uncertainty of considering the LULCC effects in climate change, as IPSL, which does not consider the distinction between grassland and pastures, does not show significant changes”. I added here this suggestion but don’t know if this is correct. I anyhow suggest to mention here explicitly what is not considered by IPSL compared to the other models.

Line 407: “This suggests that the use of annual land-use transition maps rather than annual land cover states maps (gross instead of only net LULCC transitions; Hurtt et al., 2011) leads to substantial increases in land-use emissions”. This appears to be an important feature of your study pointing at some of the different approaches how to implement again the land use information in the different ESMs. When the reader would not be completely informed about some of these subtle differences it would be not that straightforward to appreciate this. So you are saying that MPI and MIR use the land-use transition maps whereas the other ESMs applied the land cover state maps? I thought that the models all implemented the same land use information in these simulations?

Regards, Laurens Ganzeveld

Dear authors, co-authors, just checked your response and last round of edits to address my comments on your ms. Based on this, I accept your paper for publication in ESD. Lets see how this paper on the land use component in climate change will further initiate/motivate follow-up studies on this important topic,

Laurens Ganzeveld