Summary

The authors investigate the effect of including the correct external forcing in the regional climate model for past climate states. They found that orbital changes and volcanic forcing has some impact on the temperature in the region of the eastern Mediterranean and the Arabic Peninsula. The precipitation response is rather weak and not significant.

General

Overall, this is a well structured and well written manuscript, which investigates an interesting technical aspect of paleo climate modelling. The authors showed that the impact of not including the full forcing in the regional climate model (RCM) seems to be of second order relevance compared to the gain of higher resolution although in some variables it might be important to include the forcing in the RCM correctly. Such a clear statement is missing and can be included in the manuscript if the authors agree.  

Thus, I think this is certainly of importance to the scientific community and I recommend publication after some minor revisions.  

Major

1. Fig. 6: One problem is that the authors only have one simulation each for one forcing. I know that the simulations are expensive, so a small ensemble of say 10 simulations for REF and FULL might not be feasible but at least the authors need to mention this shortcoming in the results past and the conclusions.

Minor

Middle East is maybe problematic as it is a geopolitical concept.  Please use a different phrasing for the region. At least some of my colleagues react on using this for a geographical region.

P1,l4: Please change to “Here,”

P1,l15: I think that the current study does not provide “a unique source of information for the comparison of paleoclimate simulations with proxy records” as only sensitivity simulations are presented for a short period in the past. The full 2500 year long simulation certainly will do this but for this study I suggest to remove the sentences “The study is part of the new 2500-year-long transient, fully forced RCM simulation over the Eastern Mediterranean / Middle East. This work introduces a unique source of information for the comparison of paleoclimate simulations with proxy records and reconstructions” .

P2,l44: For LGM there are also a couple of studies using even convection permitting resolutions:. It may be nice to mention this here.

Velasquez P., J. O. Kaplan, P. Ludwig, M. Messmer, and C. C. Raible, 2021: The role of land cover on the climate of glacial Europe. Climate of the Past, 17, 1161-1180.

Velasquez, P., Messmer, M., and Raible, C. C., 2022: The role of ice-sheet topography on the Alpine hydro-climate at glacial times, Climate of the Past, 18, 1579–1600.

Russo, E., Buzan, J., Lienert, S., Jouvet, G., Velasquez, P., Davis, B., Ludwig, P., Joos, F. and Raible, C. C., 2023: High resolution reconstruction of LGM climate over Europe and the Alpine region using WRF, Climate of the Past, 20, 449–465, https://doi.org/10.5194/cp-20-449-2024.

P3,l66: please change “Section 2 details the models, their configurations,”

P3,l68: “Section 3 focusses ” reads better.

P3,l71: “In Section 4,  we present the main conclusions and outline potential directions for future research.” Reads better.

P3-4: Section 2.1 I would suggest to include part o the discussion presented in this section already in the introduction to make the motivate why the authors focus on this region and this period.  

P5,l120: Please remove “is used” after (Tiedtke, 1988).

P5,l126 “In this study, only the”

Caption Fig 3: “Transient forcings ...”

Fig.5 For the mean and the STDD plots it would be better to use discrete colours rather than a continuous colour scale.  

L191-195: This is a rather lengthy description for textbook knowledge so just write “The better representation of lower temperatures over orography is due to the lapse rate effect.”

Section 3.1.2/ Fig.6: It would be interesting to also include the ESM result so we can see how the RCM changes the ESM response to volcanic eruptions. Maybe this is interesting for all plots.

Fig 7: I would suggest using a non parametric test as the authors have really a small number of years, So Mann-Whitney-U test would be better here.

L315ff: I think the authors suggest that the decrease in precipitation in 1258 is due to volcanic forcing but given the variability of the time series I am not convinced that there is causality. For this statement the authors need to run an ensemble of simulations. So please rephrase.

Section 3.2.3 I guess it is always mm/month, please adapt this in the manuscript and  the figures.

Table 3: It makes no sense to show Min, 5% and 10% as precipitation is not Gaussian distributed.

L389: I would not fully agree that the effect of volcanic forcing in the RCM is pronounced. With just one simulations the authors need to be a bit more careful.

L393-395: “Land-use changes and in particular the choice of external data set have a substantial impact on simulation outcomes. Although not detailed in this study, this was tested, emphasizing the significance of the land-use dataset”.  This is fully new result and not presented in the main paper, so I recommend including this result or remove the sentence.

411: please change Methodology to Method

This detailed study by Hartmann et al. examines the use of a climate model output (MIP-ESM-LR) to drive a higher-resolution regional climate model (COSMO-CLM), where the external forcing implemented follows Jungclaus et al., 2017. Apparently this study is part of a 2500 year regional climate model simulation over the Eastern Mediterranean and the Middle East region that I agree is an important region in many aspects. The selected time period for this particular study is one decade and covers the 1257 Samalas eruption. The contribution of each forcing in the simulated regional climate variability is tested through several single-run sensitivity simulations (single and combined forcing using the MPI-ESM output) in addition to a present-day COSMO-CLM reference simulation. It is mentioned that this method leads to a novel paleo-regional climate model and this paper is dedicated to distinguishing between the relative contribution of each forcing and the internal variability with the aim to see if the incorporation of external forcing into the regional climate model enhances realism at the regional scale.

The methodology used is good in general, biases and detected (regional) climate extremes associated with each forcing are studied in detail, and the results have the potential to contribute to the overall enhancement of reliability within climate and regional model communities on various timescales. This manuscript is also well structured and reads well for the most part and so I would say that it is worthy of being published in Climate of the Past after the authors address some of my concerns. Although I suggest major revision, it is mainly due to lack of discussions regarding certain aspects and/or requested clarification regarding parts of the methodology - but the authors can of course use these suggestions to go as deep as they see fit.

These concerns mainly relates to the fact that these single-run sensitivity simulations are only but one realisation of climate variability over the study area and this needs to be addressed in more details and how this might impact the results obtained (and if this will be accounted for in the larger 2500yr run). This becomes quite clear in Figure 6 and 7. In Fig. 7 it seems that the cooling that followed Samalas did not contribute significantly to the mean seasonal temperature anomalies (in any season) in the FULL experiment. As can also be seen in Figure 6, the T anomaly decrease is small, -0.4°K (perhaps within the range of natural climate variability), so if the authors were to run an ensemble of these simulations, result might only show a weak significant T decrease in 1258. Also, the reference period is highly important. Here the reference period (for the RCM) is tuned according to present-day climate, did the authors check how/if the results would change according to pre-industrial reference period? I think that could give us important information on how relevant these post-volcanic anomalies were with respect to the current time period, I urge the authors to at least discuss this. It would also make sense to discuss the response detected here and compare with the known climate impact of Samalas partly discussed in the intro (maybe in discussions). I feel that too much focus is put on absolute temperature changes when referring to other studies with respect to the results.

It is also interesting that the seasonal mean of precipitation is more significant (summer/autumn) than the seasonal mean temperature, since precipitation is usually more sensitive to parametrisation in these models and thus less robust compared to simulated temperature. How well in general does the regional model simulate precipitation (compare to other models/obs/etc)?

Here below are more specified comments:

L53: „Various external forcing become relevant, ...„. This sentence sounds incomplete, become relevant in regional climate variability?

L120: remove „is used“.

L134: Indeed volcanic forcing can have decadal climate impacts through secondary processes beyond the well-known direct surface cooling.

L142: maybe hundreds of millennia? Eccentricity oscillates in periods of 100.000 to 400.000yr.

L143: Annually -> Annual ?

L208: 1256 is also more than 1K cooler than in 1255. Although the direct temperature response is interesting, the anomalies reveal more about the significance of this T decrease.

L220: should this not be 1262 as in Fig. 6?

L293: So in general, first the well-known precipitation decrease followed by a modified/reversed pattern? Could you reflect on the timescale these changes might act on? This is also related to the comment above in L134.

L323-324: This could depend on the potential modified patterns followed by the strong surface cooling, so maybe it does not disappear but it certainly become less dominating.

Figure 1 caption: Does this mean that this is the author's representation of Toohey and Sigl?