This paper describes a land cover change dataset for use with regional climate models run for any period from 1950 to 2100 over the Euro-CORDEX region. Future land cover maps are derived for each of the SSP scenarios and are based on the Land-Use Harmonization 2 (LUH2) dataset. This dataset is an important enabler of downscaled climate projections using the SSP scenarios such as those under the CORDEX-CMIP6 project.

The article is appropriate to support the dataset. It presents the methodology in sufficient detail to enable replication and evaluates the dataset against multiple products and includes an estimate of the uncertainty.

The dataset was easily accessible and downloadable after a registration step. It comes as netCDF format files making it very easy to use with the appropriate metadata embedded in the files.

Overall the dataset is unique, useful and complete for the Euro-CORDEX region. It would be great to see this same methodology be applied to produce similar datasets for all CORDEX regions globally at some point in the future.

The paper is well written and worthy of publication. I have only technical corrections that should be addressed before publication below.

Technical corrections

1. ln 72: delete extra “keeping”

2. ln 399: “lager” should be “larger”

3. ln 546: delete the first “based”

4. ln 609:”emphasize” should be “emphasizes”

5. ln 621: “giving” should be “given”

6. ln 619: “changes” should be “change”

7. ln637: “documentation” should be “document”

8. ln 556:”currently validated” should be “currently being validated”

General comments

This preprint describes a newly developed high-resolution land use/cover dataset for Europe, ready-to-use for Regional Climate Models. The presented dataset represents a novelty, since it provides higher spatial resolution and more thematic detail for both the historic and the future land use/cover trajectory in Europe, compared to the current state-of-the-art land use forcing LUH2. The method embodies a novel approach of combining coarse-resolution land use transitions (from LUH2) with high-resolution land cover information (from ESA CCI LC; forest types) for generating time series of Plant Functional Types (PFT) maps tailored to the modelling community. The presented dataset/topic shows high significance and large potential to be used for further studies. It is unique, useful for future modelling experiments and presented in a complete, comprehensible form (accompanying the Reinhart et al. 2022 paper). Resulting datasets are accessible via the given identifier, accompanied with metadata, and a discussion of uncertainties in the article. Further, the results are compared to other recently published land use/cover change datasets, which places the findings in the current state of research. In addition, the paper is interestingly written and clearly structured.

Overall, the paper is well written and its content of wide scientific interest. I recommend publication after only a few minor, mostly technical revisions.

Specific comments

Line 48

Why is a historic time series of 65 years required? Why is 1950 the starting year, is that a requirement from the RCMs?

Line 100: “First, the ESA-CCI LC map for the year 2015, which has a native resolution of ∼300 m globally, is aggregated to 0.1 ◦ resolution”

It was not clear to me how the aggregation was done. Which resampling method did you use, which classes/grid values did you include? I take notice that it is described in more detail in Reinhart et al. 2022, but a little more information would be good here.

Lines 182-184: “Following the recommendations [..] natural vegetation (i.e., forest and shrubland) is only cleared and converted into grassland for land-use class transitions to pasture, while it remains unchanged for land-use class transitions from non-forested vegetation to rangeland.”

What is the reason for this? Maybe the recommendations provide more justification. I am not sure if I understood it correctly: So, the dataset only includes transitions from natural vegetation to anthropogenic land use and no forest to shrubland or shrubland to natural grasslands? Perhaps you could provide some more justification and examples for this.

Lines 282-232: Treatment of irrigated cropland

Was there a reason to include irrigation as a land management option and not e.g. fertilizer application, pesticide usage, cropping frequency, crop types, etc.? It would be good to have more information/justification of why irrigation was selected as the only indicator of land management.

Tables 5 and 6

I do not quite understand the “forward in time” or “backward in time” from the captions in relation to the column and row labels in Table 5 and especially Table 6. I think that headers are needed to describe the “From…” column labels and the “To…” row labels, e.g. “PFT in time step 0” and “PFT in time step 1”. If Table 5 shows the “forward in time” transitions and Table 6 the “backward in time” transitions, why are there several entries in the “from URB” column in Table 5, whereas there are none in the “from URB” column of Table 6 (if it is read as backward in time, it actually means transitions from something else to urban?). Also, the FOR-NFV transitions (no entries in Table 5 and 2 entries in Table 6) are not clear to me.

Lines 571-572

It is not surprising that LUCAS LUC land cover changes are similar to LUH2, as the LUH2 is a main input for generating the dataset. I suggest adding in the introduction why LUH2 is so important for the methodology and in the end for the emergence of LUCAS LUC (possibly because there are no other annual land use change datasets with future scenarios).

Technical corrections

• Line 72: Word duplication, please remove one “keeping”.
• Line 289: I think “Tsendbazar et al. (2021)” is not the right reference for the Copernicus LC100 dataset and brackets are missing. Please put the correct dataset reference here:
  
  Buchhorn, M., Smets, B., Bertels, L., Lesiv, M., Tsendbazar, N.E., Herold, M., Fritz, S., 2020. Copernicus Global Land Service: Land Cover 100m: collection 3: epoch 2015-2019: Globe. Version V3. 0.1.
  
  https://doi.org/10.5281/zenodo.3939038; https://doi.org/10.5281/zenodo.3518026; https://doi.org/10.5281/zenodo.3518036; https://doi.org/10.5281/zenodo.3518038, https://doi.org/10.5281/zenodo.3939050
• Line 386: Remove „in“ from „and in especially HILDA+”.
• Line 399: “lager” should be “larger”.
• Line 404: “HIDLA+” should be “HILDA+”.
• Figure 6, caption: “show” should be “shown”.
• Line 471: A comma is missing after “While grassland cover strongly increases in one scenario (Fig. 13b) in the IP region”.
• Line 494: “a increase” should be “an increase”.
• Line 544: Commas are missing before and after “averaged over Europe”.
• Line 608: “a extreme” should be “an extreme”.
• Line 664: Add an “and” after “The LUCAS LUC datasets were already produced for other CORDEX regions (Hoffmann et al., 2021)”.