Review of ‘Projected climate change in Fennoscandia — and its relation to ensemble spread and global trends’ by Strandberg et al.

The manuscript presents results how temperature and precipitation change under projected climate change over Fennoscandia and Sweden using an ensemble CORDEX regional models with different GCMs, RCMs and three different rcp scenarios. The results provide important information about climate change in this region and should be published if the authors consider some major concerns described below. My main concern is that the Method section is not clearly presented, and several results are introduced without sufficient explanation of how they were calculated. For the findings to be properly understood and evaluated, methods should be linked to the results

The Introduction is generally well written. The choice of RCMs (CORDEX )is described, however there are some known issues with CORDEX data that should be mentioned, e.g. lack of aerosol impacts, precipitation and temperature biases (Vautard et al., 2021).

The Method section should be revised to enhance clarity and allow for a better understanding of the results. What variables are bias corrected, and what are the results of this correction. Over what period is this correction performed? How does it affect future projections? The effect of the bias correction is missing in the representation of the results further in the manuscript. If some models are biased corrected more than others, I guess this would impact the comparisons. If it is described and represented in Berg et al (missing in references) the results should be briefly presented. Is it only over land? (Region A covers the ocean as well)

The selection and analyses of sub-ensembles section is also unclear. Is it only these 17 models that is used further in the manuscript? What are these 17 models used for? This is the first mention of GWL (not written out). How is the GWL found? In the GCMs that drive the RCM? This should be described better. How and when GWL 2 is reached in different models and projections should be included in a Figure. Only results for tas and csu are presented in the manuscript, while several are stated here.

The methods section should also include a brief description of which periods the results are calculated.

Figure 1, is region B used in any analysis, I did not find any further reference to this region.

Figure 2 and 3, Is this the mean over all the CORDEX models available or only the 17? The introduction and title of the manuscript focuses on the spread of ensembles that are not shown in this figure. Could this be included in the figure by shading be added where there is a large spread between the models?

For climate adaptations, the change in seasonality of precipitation is important, as summers become drier and autumn wetter for crop security. Although mentioned, the results could be included in supplementary. ‘The signal is not robust’, this needs to be shown, maybe by showing the spread between the RCMs in the Figures. As a general comment, since supplementary material is already included, the results described in the manuscript should also be provided. Referring to results and then stating 'not shown' is inconsistent and reduces clarity for the reader

The calculations in section 3.3 are not described in the method section. Are these also bias corrected? What models are included? Should also include CMIP6 models as they have been available for several years(CMIP7 results are available soon).

Figure 4 should have a legend, and darker color for the RCP4.5 and RCP8.5 models.

Figure 5- 8. These results would be more significant for future work if the different RCMs and GCMs were identified by model name on the y-axis. As some GCMs are too warm or too dry (or cold and wet) and labeling them would help highlight which models show large discrepancies.

Figure 9 Change colors (same as Figure 4) and the different time periods could have different markers. The calculations in section 3.5 is also not mentioned in Methods, how is the global temperature calculated?

Need to justify the statement ‘and that a GWL could be based on only one RCP’. This is not shown in the manuscript.

The references section should be checked. I could not find the Berg et al. 2022 reference and IPPC chapters should be referenced as they state on the first(second) page.

Minor Comments

P2. L38: Give some examples of the physical processes that is better represented.

P2. L42-43: Give a reference for this statement.

p.2 l.49-51. This sentence could be improved

p3. L.67: CMIP5 has been written out previously.

P5. L78-79: SMHIGridClim should not be in the same parenthesis as the reference (split), it is confusing if it is a reference or an abbreviation.

P6. Table 2 Could be simplified by putting ‘season or year’ in the legend of the table.

P7. L108. Split the parenthesis.

P11 Figure 3. Are these yearly means?

P12 l 183. Use another word than distance.

P15. L 233: Use different word than instead. Both results are presented.

Dear authors please find my detailed review comments in the attached file.

The paper “Projected climate change in Fennoscandia – and its relation to ensemble spread and global trends” presents temperature and precipitation projections for Scandinavia using bias-adjusted EURO-CORDEX GCM-RCM ensembles. It compares these with CMIP5 GCM signals and analyzes the influence of GCM, RCM, and emission scenarios on projection differences. The study finds that the EURO-CORDEX ensemble aligns well with the broader CMIP5 GCM ensemble and supports using global warming levels to reduce emission scenario uncertainty. It also notes that warming in Scandinavia since pre-industrial times has been about twice the global average, though this regional amplification is expected to decrease as global temperatures continue to rise.

Major revisions

Bias-Adjusted RCM vs. Raw GCM Data:

Comparing bias-adjusted RCM data with raw GCM data is methodologically inconsistent. To avoid misinterpretation due to the effects of bias adjustment, it would be more appropriate to compare raw RCM and raw GCM simulations directly. In this context, a more detailed description of the applied quantile-mapping approach would be valuable—for example, whether it is trend-preserving and whether it modifies climate change signals.

Sub-Ensemble Construction:

The rationale behind the construction of the sub-ensembles remains unclear. The purpose and in particular the objectives of the analyses should be explained more explicitly.

Language and Clarity:

The manuscript requires revision for language and writing style. Scientific clarity and fluency should be improved throughout. Several sections would benefit from more concise descriptions, while others lack necessary technical details—particularly regarding the ensemble setup and methodology.

Tables:

Table 1 – The asterisk (*) is unnecessary, as the information it provides is already evident from the table.

Table 2 – The selection of indices is likely relevant for the Scandinavian climate; however, presenting absolute changes in some indicators without referencing the baseline climate limits their interpretability. For example, many regions in Scandinavia experience few or no summer days, while others have around 30 in the reference period. Similarly, heavy precipitation of 10 mm/day is common along the Norwegian coast but rare in the polar regions of northern Sweden. Including percentile-based metrics or relative changes in indicators like Rx1day could improve clarity. Additionally, for consistency with other studies, replacing the Dry Days (dd) index with wet-day frequency and presenting its relative change may provide more meaningful interpretation.

Figures:

The figure requires major revision. Figures should be self-explanatory, featuring clear captions, meaningful legends, and clean visualization to convey the message without relying on the main text.

Figure 2 – As this paper aims to provide an overview of projected climate change in Scandinavia, including precipitation changes would be beneficial. Additionally, while the paper focuses on global warming levels, it presents end-of-century temperature changes for different emission scenarios, which is somewhat inconsistent with the emphasis on global warming levels expressed elsewhere in the text.

Figure 3 – Panels c, d, e, g, and h require representation of historical climatology to enhance interpretability (see also comment in the tables section).

Figure 5 and 6 – The content presented is unclear. The terms “South” and “North” are undefined, and the GCMs/RCMs corresponding to the y- and x-axis indices are not indicated. The meaning of “csu” is also unclear. Captions require substantial revision to improve clarity. Figures should be self-explanatory, with comprehensive captions and legends that convey the intended message without relying on the main text.

Some specific comments

L30 – The temperature response in Europe is correlated but warms at stronger rates.

L33 – The paragraph on climate models needs to be introduced stating why climate models are employed.

L40 – best available –> most comprehensive

L42 – they also allow for a probabilistic assessment of potential changes. In general, I would state the ability to employ a wider set of statistical tests at the end not as prominent.

L46 – this sounds as if this paper would introduce the EURO-CORDEX data, however, it is rather the presentation of an analysis of a dynamically downscaled ensemble.  

L58 – The “construction” section that follows is inconclusive and appears primarily driven by the choices of the EURO-CORDEX initiative and subdividing the ensemble by GCM, RCM and RCP, as clarified in later parts of the study.

L75 – L80 (2.2 Bias adjustment) – Please add more detail on the set-up of the quantile-mapping implementation. This is particularly important as some quantile-mapping implementation preserve trends (climate change signals) while other can modify it. This is of high relevance for the present paper as you compare quantile-mapped RCM data with raw GCM data, which is generally not a proper and fair comparison (see also later).  

L180 – GMC à GCM

L183 – sentence seems to be incomplete.

L202 – consistency in the rate of global warming.

L267 – warming since pre-industrial conditions. ¨

L274 – Out of interest, is there a possible explanation for (1) the warming being twice as large as the global average since pre-industrial times, and (2) the subsequent leveling off of this trend in the future (e.g., due to sea-ice dynamics, polar processes, or topography-related factors)?