This manuscript evaluates the local and non-local effects of land use and land management change (LULMC) under present-day climate in three state-of-the art Earth system models. To do so the authors use a dedicated set of experiments of different idealised land use states. 

The manuscript is methodologically sound and presents an important contribution to understanding the biogeophysical impacts of LULMC. In particular, the use of three different ESMs enables to provide a more comprehensive picture on the uncertainties and robustness of the simulated features. The number of simulations, ESMs, and variables discussed make a fairly dense reading of the results, which, in my opinion, could be somewhat improved by following a more rigid structure (see below). 

Main comments:

The results and section 3.2.1 in particular are hard to follow. Consider clearly separating local and non-local (e.g. move ln 349f to ln 358) and sticking to one order in which you discuss the models in each section (e.g. CESM, followed by EC-Earth, then MPI) throughout the results (or manuscript in general). 

Non-local effects - Winkler et al (2017) recommend prescribed SSTs to isolate non-local LCC effects from background climate. Using a dynamic ocean as done here has the advantage to identify non-local LULMC impacts such as the cooling/warming response in the north Atlantic in CESM. However, it also seems to blur the non-local effects from LULMC and background climate judging based on the appendix figures' widespread - and seemingly robust (?) - non-local effects. I feel this should be discussed to give an indication on how robust these non-local effects are (not requesting another simulation, although this would be instructive in really separating those non-local effects). 

Minor comments:

ln 23 Depending on the land use forcing higher estimates exist (139 PgC for the last 1000 years in Kaplan et al 2011). 

Kaplan, Jed O., Kristen M. Krumhardt, Erle C. Ellis, William F. Ruddiman, Carsten Lemmen, and Kees Klein Goldewijk. “Holocene Carbon Emissions as a Result of Anthropogenic Land Cover Change.”  The Holocene 21 (5): 775–91. https://doi.org/10.1177/0959683610386983.

ln 103ff In the CESM description it would be informative to add the horizontal resolution (currently buried in ln 159)

ln 125 If possible give grid resolution TL255 as degree or km2 equivalent

ln 128 This seems important, could you please state briefly what happends instead?

ln 160 consider moving the information on the grid resolution into the model description

ln 181 - Do you mean "the native parameterization of irrigation"? 

ln 274 Are the 0.1 W m-2 the global imbalance?

ln 338-344 Consider moving this into the discussion.

ln 354 - Please clarify, do  you mean MPI-ESM shows a similar pattern to CESM in the local effects?

ln 365ff - Consider moving this extended comparison to other published work into the discussion.

 

ln 390 - So really just the extratropics, not most of the globe?

ln 482 - "This indicates that extra-tropical afforestation

is dominating the global picture for these models due to a strong albedo response largely counteracted by the turbulent heat

fluxes" - Unlear to me, do you mean that the albedo response to extra-tropical afforestation is dominating the global energy balance response to afforestation?

ln 574 - irrigation as an adaption strategy? Surely only where there is no future water stress as the authors point out further down. Is there any specific reference supporting this idea? 

ln 660 - URL missing

ln 696 - space missing

Review on De Hertog et al.

The manuscript investigates local and non-local climatic effects of LCLMC by three ESMs using four unique idealized experiments under the current climate. The study involves much work and brings novelty to the scientific community. I recommend accepting it after considering the comments below that could help to improve the paper.

Main comments

• As I understand and as the abstract states, there is much uncertainty in the three models on the central questions of the study. The differences are well explained for each model, but could you go one step further and give overall conclusions (taking into account various model biases, etc.) and specific recommendations to the model developers on reducing this uncertainty?
• As the other reviewer recommended, please consider revising the structure. It may be helpful for the reader if some text on the differences of the results (experiments/models/local and non-local effects) could be summarized in a table. Many findings are hidden in the text. A table that summarizes what models agree on disagree with, and overall conclusions would be helpful.

• Could you add some info on how the PFTs distribution differs spatially among the models in the year 2014 (I expect they all may deviate from LUH2)? Can part of the inter-model differences be attributed to this? Also, is there a large inter-model difference in the distribution of forest types? (deciduous and evergreen forests may have very different implications for albedo). Similarly, can you state that any afforestation (from any previous land cover and to any type of forest) causes warming in the north and cooling in the tropics? Is it universal, or can you make some conclusions (although this may require additional analysis), e.g., conversion of specific forest type to forest gives more/less warming to make more implementable conclusions?

Minor comments

• Figures: consider adding a row of ensemble means of three models to some figures where appropriate
• Figures 2 and 3: consider merging these two figures
• Line 91: Second,
• Line 159: space missing
• Lines 160-170: Explanation of the methods. As I understood from the text, if the grid has at least some forest, then the grid becomes 100% forest. If it does not have forest but has other vegetation, then this vegetation becomes forest in a ratio of different forest types determined via latitudinal averaging. Is this so? The explanation requires clarification as it is quite misleading in its current state.
• Line 165: for forest PFTs or for grids that have forest PFTs?
• Line 182 and around: Please double-check the letters of figure panels shown in text, they do not always match
• Lines 351-354: “conditions in which the frozen soils are less extensive throughout the year which causes a soil warming”. Please revise the sentence to make it clearer
• There is much warming in the CROP experiment simulated by EC-EARTH due to local effects (fig. 5), but I cannot see to what local effects can this warming be attributed to in the decomposition analysis.
• Figure 5 and C1 seem to have the same legend but vary. What are they standing for?
• There seems to be a mix-up as the appendix does not always give info that is promised in the text, and there is no Appendix D that you refer to.
• Line 354: less cooling
• Lines 388-400: Although you mention that Boysen et al. (2020) already show the biophysical impact of LULMC on AMOC, I think this finding deserves more discussion.
• Line 458: both in both?