This paper describes a new large ensemble with the climate model EC-Earth3 and a relatively unique combination of emissions scenarios. The analysis examples are largely uncontroversial/unsurprising, which is perhaps intentional and adequate for a description paper. The paper is written very clearly, with sound methods and results that support the conclusions. I only have a few small comments that could be considered in a revised version, after which I can recommend publication. I want to express my thanks to the authors for making this data available to the community; it is very valuable.

Three requests as a curious reader:

Could the authors make a few intuitive comparisons with CMIP6? In particular, I was under the impression that EC-Earth3 has among the largest decadal global temperature variability of any CMIP6 model (at least in the piControl). How does this affect its signal/noise ratio in comparison to other CMIP6 models and other available large ensembles?

It would be interesting to divide the change patterns from the different emissions scenario by global mean temperature, so the pattern become comparable. I actually think that would be a more interesting analysis (and easy to do) than the absolute change maps that are shown here. The authors kind of have to show the absolute change patterns for the description paper, I get that, but we already knew how they would look. I think a comparison of the normalized patterns would be very interesting for the overshoot vs non-overshoot path.

I encourage a comparison with or at least discussion of Sanderson et al. (2017) and several references therein, which investigate what might have been the first “large” (n=10) ensemble of overshoot simulations, albeit with CMIP5 forcing and not CMIP6 forcing. 

Other comments:

There have been some concerns with the transition of biomass burning (BB) forcing from the satellite period to the future emissions scenarios in the CMIP6 forcing files. The forcing file has high variability during the satellite era and abruptly lower variability thereafter, at least in certain regions of prominent BB. A large ensemble is perfect to investigate whether this has an appreciable effect on the simulated climate during that transition period. It’s not a priori clear whether it affects each model, as aerosol forcing is implemented quite differently across models. It would be valuable information for the community whether there’s any sensitivity to this issue in EC-Earth.

Fig. 2: Please add some observations (at least for tas, pr, and sic) as a minimal model validation and to illustrate how the ensemble range compares to real-world variability.

Some map colorbars are a bit unintuitive, as they have a color switch offset from zero (e.g., Fig 3c-j). Also, rainbow or rainbow-like colorbars aren’t encouraged.

L188: “mainly due to reduced variability of the change across members” sounds confusing to me. Do the authors just mean “due to the larger signal under stronger forcing”?

L218: “highly relevant” rather than “highly important”

L151: that’s actually quite interesting. Could the authors speculate what could cause this?

L100: “There are differences”

L116: “tell us”

References:

Sanderson, B. M., et al., 2017: Community climate simulations to assess avoided impacts in 1.5 and 2°C futures. Earth Syst. Dyn., 8, 827–847.

Review of The SMHI Large Ensemble (SMHI-LENS) with EC-Earth3

This paper is written as an overview/introduction to the SMHI-LENS. The paper is well written and provides a sufficient introduction to this model. However, the paper misses some relevant literature in the introduction, need some more detail on the initialization of the ensemble and could use minor changes to the Figures to help with interpretation by the reader. I recommend that the paper is revised before it is accepted.

Comments are as follows

Section 1:

While this provides a good introduction, it is unclear why the authors cite specific large ensembles and not others (see line by line comments).

The introduction would benefit from a paragraph describing some of the interesting work already done using large ensembles. While the literature is too large to include everything, some references perhaps relating to what is shown later in the manuscript, or a brief introduction to new science done with large ensembles should be included.

Section 2.2 Initial conditions:

Please include the specific years that you used for the initial states in a table.

Figures: are rainbow colorbars the best choice? Perhaps you can find a better colorbar

F2 – poor quality and fuzzy

two orange colors are difficult to distinguish by eye on my computer screen

F6 – b) the orange line seems to come from nowhere

c) I don’t see the orange line at all

perhaps different symbols or dots, dashes could be used so we can see all colors

F7 – Please describe in the caption how you compute the 10 member result. Do you pick one set of 10 members or resample 10 members many times?

Given most large ensembles have 30 members, as you note in your introduction. It would be good to do this for 30 members as well as 10 members and add a panel to the Figures. Would it be worth considering precipitation for these Figures as well given the pathway dependence of this variable:

e.g https://journals.ametsoc.org/view/journals/clim/30/11/jcli-d-16-0441.1.xml

https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/2016GL070869

https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2018JD028821#:~:text=We%20find%20a%20robustly%20larger,GHGs%20across%20all%20available%20models.&text=This%20is%20because%20of%20a,by%20the%20GHG%20atmospheric%20forcing.

Line by line comments:

Line 34- This is also shown using large ensembles in the following two papers:

https://esd.copernicus.org/articles/11/491/2020/

https://iopscience.iop.org/article/10.1088/1748-9326/ab7d02/pdf – this could also be compared to the results on line 185-186 in the discussion

Line 46 – MPI-GE is not MPI-ESM-LR but MPI-ESM1.1 – additionally the correct acronym for this large ensemble is MPI-GE not MPI-ESM-GE

Line 46 – I am confused about the choice of models introduced here. The large ensemble archive introduced by Deser et al 2020 includes more models, why not introduce all of the ones in this archive?

Line 51 – Also GFDL-SPEAR is now available online: https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2019MS001895

Line 70 – RCM large ensembles already exist. It would be worth citing these here:

https://journals.ametsoc.org/view/journals/apme/58/4/jamc-d-18-0021.1.xml

106 – is there a citation for SSPs and ScenarioMPI?

155- I believe this is usually called TAS? Would it be more understandable to use the standard acronym – also please be consistent as you use tas in Figure2’s caption

163 – perhaps 3K and higher is better wording

163 – ‘the’ northern hemispheric

173 – is increasing → ‘increases’

177 – should this be ‘divided by’?

221 – it would be interesting to add whether the Aleutian low is too pronounced in all ensemble members as we would not expect observations to agree with the ensemble mean. This applies for all the metrics discussed on these lines.

236 – however this result contrasts with the following work, which should be added on this line

https://journals.ametsoc.org/view/journals/clim/30/11/jcli-d-16-0441.1.xml

509 – specify what the nino3.4 region is