the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 27 Jun 2023
High resolution LGM climate over Europe and the Alpine region using the regional climate model WRF
Abstract. In this study we present a series of sensitivity experiments conducted for the Last Glacial Maximum (LGM, ~21000 years ago) over Europe using the regional climate Weather Research and Forecasting model (WRF). Using a 4 step 2-way nesting approach, we are able to reach a convection-permitting horizontal resolution over the inner part of the study area, covering Central Europe and the Alpine region. The main objective of the paper is to evaluate a model version including a series of new developments better suitable for the simulation of paleo-glacial time slices with respect to the ones employed in former studies. The evaluation of the model is conducted against newly available pollen-based reconstructions of the LGM European climate and takes into account the effect of two main sources of model uncertainty: a different height of continental glaciers at higher latitudes of the northern hemisphere and different land cover. Model results are in good agreement with evidence from the proxies, in particular for temperatures. Importantly, the consideration of different ensemble members for characterizing model uncertainty allows to increase the agreement of the model against the proxy reconstructions that would be obtained when considering a single model realization. The spread of the produced ensemble is relatively small for temperature, beside areas surrounding glaciers in summer. On the other hand, differences between the different ensemble members are very pronounced for precipitation, in particular in winter over areas highly affected by moisture advection from the Atlantic. This highlights the importance of the considered sources of uncertainty for the study of European LGM climate and allows to determine where the results of an RCM are more likely to be uncertain for the considered case study. Finally, the results are also used to demonstrate the added value of convection-permitting resolutions, at both local and regional scales, under glacial conditions.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(20683 KB)
-
Supplement
(1130 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(20683 KB) - Metadata XML
-
Supplement
(1130 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-1197', Anonymous Referee #1, 27 Jul 2023
The manuscript presents the results of a series of high-resolution regional climate simulations with the model WRF covering Western Europe and focusing on Switzerland for the Last Glacial Maximum. The model set- up improves some of the perceived limitations of previous simulations. In particular, this set of simulations includes some that are convection-permitting and a new estimation of land cover. The simulations are compared with a new pollen-based reconstruction of the Last Glacial Maximum.
The main conclusion is that convection and land cover do alter the distribution of mean temperature and precipitation in this region.
Recommendation: The manuscript is, in my opinion, well and clearly written, but the motivation of the study does not come across as clearly. I have some recommendations that the authors may want to consider in a revised version
Main points
1) The research question and motivations of the study are somewhat unclear. The study presents a new set of simulations, but why are they necessary? What were the deficiencies of previous simulations? Why is a convection-permitting resolution in principle necessary to better simulate the climate of the LGM ? One of the main conclusions is that the resolving convection does have a clear effect on the simulated precipitation, but it remains unclear - unless I missed it in the manuscript - whether the convection scheme improves the simulation of precipitation (and maybe temperature) compared to the reconstructions.
2) The model set-up also includes a small ensemble with different initial conditions (?). The study finds that the ensemble spread can be large for precipitation, but not so much for temperature. However, the length of the simulations is short, just 11 years. Is it possible that the ensemble spread is just due to the short length of the simulations? Could this spread be compared to the decadal variability of precipitation in the present climate?
Particular points
3) ‘However, these increases in model complexity have not generally led to improved model performance when compared against proxy ...’
Could the authors be more specific here? What are the deficiencies of previous simulations that remain unexplained?
4) ‘ Model results are evaluated against a newly developed pollen-based reconstruction database for the European LGM climate.
A reference to the new reconstructions would be helpful here.
5) The starting point of the presented simulations is the results of earlier studies using the same model version (Velasquez et al., 2020, 2021, 2022).’
Same model version? The sentence a bit later in the paragraph says 'previous version'. Could the specific model version used by Velasquez et al. be mentioned here?
6) ‘ D01 and D02, down to a spatial resolution of 18 km and with the convection parameterization switched on, is performed. This experiment is indicated as DEF_noconv in Table 3’
switched off, I guess.
7) These differences are in some cases of the same order of the differences between the different ..’
...order of magnitude as the differences between.
Citation: https://doi.org/10.5194/egusphere-2023-1197-RC1 -
AC1: 'Reply on RC1', Emmanuele Russo, 31 Oct 2023
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-1197/egusphere-2023-1197-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Emmanuele Russo, 31 Oct 2023
-
RC2: 'Review of "High resolution LGM climate over Europe and the Alpine region using the regional climate model WRF" by E. Russo and coauthors', Anonymous Referee #2, 14 Sep 2023
SUMMARY
Russo and coauthors apply a regional model nesting approach to dynamically simulate Last Glacial Maximum temperature and precipitation patterns over Europe. They apply the regional Weather Research and Forecasting model (WRF) with two-way nesting down to 2km resolution, based on a 28-year high-resolution atmosphere-only Community Earth System Model simulation driven by surface fields out of an equilibrium simulation with the fully coupled atmosphere/ocean version of CESM at lower resolution. They proceed to compare the fields of the simulated climate with, and without, convective parameterizations to that of a pollen-based climate reconstruction (Davis et al., 2022).
The paper is clearly structured, and fairly well written, albeit somewhat dry, dispassionate and with little curiosity for meteorological phenomena underlying the high-resolution weather simulated, and restricting itself to the bare minimum in the climatological analysis in this version of the paper. At 4s/2km resolution, investigating for example storms, or regional effects, could be quite interesting. Some further points are detailed below. All in all a paper that can be improved.
Key Points
- Goals: The authors aim to evaluate an apparently bug-fixed version of WRF against data, and contrast model uncertainties from ice height, land cover and convective parameterizations. These are great points to study. However, the study design does not really allow to understand how the different uncertainties play out against one another (nonfactorial), and looking at the figures in the results and discussion section, as well as the supplement, does not elucidate this further.- CO2: One potential reason why the model simulations appear biased dry is not discussed: Namely that the used pollen data suggests dryer conditions than warrented. The water-use efficiency under low CO2 conditions is lower, implying that plants are more stressed under similar climatic conditions [1] -- so perhaps the model is less biased than it appears.
- Discussion: Here you could bring in more depth. You could elaborate whether you expect that the results found here dependent on the version of WRF, and on CESM as a host model? The fact that the 28-year global simulation providing input does lead to significant spread in the regional model results is surprising: Where does this divergence come from? Are these nonstationary effects that suggest that the simulation period is too short? This would also imply that averaging over such a short time period may be inappropriate, weakening the justification of one of the assumptions set out (p5 last paragraph).
- Vegetation cover discussion: Given the substantial differences between the land surface conditions fed into the high-resolution simulations -- don't you expect to see effects arising simply from the strongly different land cover, for example in North Africa?Detailed Points
p2l30 "a series of LGM studies have shown..." this sentence needs references.
p2l35/p3l1-3 Here a differentiation to statistical/statistical-dynamical downscaling should be added.
p3l25 "The starting point .... are the results of earlier studies using hte same model version"... so what? What are the results of the earlier studies that imply one should do the same things? It feels like something is missing here.
p3l29 delete space after 2.3
p4l13 add space after precession
p4l4 these sentences on the glacier scheme are confusing. Does ice become supercritical in NOAH-MP? Or is what is meant that there are melt/refreeze processes in the version used in Velasquez et al. (2021) that produce unphysical temperatures?
p7 sec 2.4 -- A key weakness of Davis et al. (2022) is that it does not address the CO2-caused precipitation bias in the reconstructions, which would be expected to cause a dry bias under the low CO2 conditions.
p8 l13-15 The narrow distribution of precipitation estimates out of the pollen-based reconstructions is perhaps indicative of the dry bias (s. above)
p9 l27 remove "¨"
p9 l32-35 Indeed, the large differences between the ensemble members are remarkable. But going back to the ensemble description, can this be simply due to internal variability in the non-overlapping subsections of the 28-year simulations? (The description of the ensemble design is confusing).
p11 Code and data availability: Fix broken reference.References:
[1] Prentice et al., https://doi.org/10.1016/j.gloplacha.2022.103790Citation: https://doi.org/10.5194/egusphere-2023-1197-RC2 -
AC2: 'Reply on RC2', Emmanuele Russo, 31 Oct 2023
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-1197/egusphere-2023-1197-AC2-supplement.pdf
-
AC2: 'Reply on RC2', Emmanuele Russo, 31 Oct 2023
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-1197', Anonymous Referee #1, 27 Jul 2023
The manuscript presents the results of a series of high-resolution regional climate simulations with the model WRF covering Western Europe and focusing on Switzerland for the Last Glacial Maximum. The model set- up improves some of the perceived limitations of previous simulations. In particular, this set of simulations includes some that are convection-permitting and a new estimation of land cover. The simulations are compared with a new pollen-based reconstruction of the Last Glacial Maximum.
The main conclusion is that convection and land cover do alter the distribution of mean temperature and precipitation in this region.
Recommendation: The manuscript is, in my opinion, well and clearly written, but the motivation of the study does not come across as clearly. I have some recommendations that the authors may want to consider in a revised version
Main points
1) The research question and motivations of the study are somewhat unclear. The study presents a new set of simulations, but why are they necessary? What were the deficiencies of previous simulations? Why is a convection-permitting resolution in principle necessary to better simulate the climate of the LGM ? One of the main conclusions is that the resolving convection does have a clear effect on the simulated precipitation, but it remains unclear - unless I missed it in the manuscript - whether the convection scheme improves the simulation of precipitation (and maybe temperature) compared to the reconstructions.
2) The model set-up also includes a small ensemble with different initial conditions (?). The study finds that the ensemble spread can be large for precipitation, but not so much for temperature. However, the length of the simulations is short, just 11 years. Is it possible that the ensemble spread is just due to the short length of the simulations? Could this spread be compared to the decadal variability of precipitation in the present climate?
Particular points
3) ‘However, these increases in model complexity have not generally led to improved model performance when compared against proxy ...’
Could the authors be more specific here? What are the deficiencies of previous simulations that remain unexplained?
4) ‘ Model results are evaluated against a newly developed pollen-based reconstruction database for the European LGM climate.
A reference to the new reconstructions would be helpful here.
5) The starting point of the presented simulations is the results of earlier studies using the same model version (Velasquez et al., 2020, 2021, 2022).’
Same model version? The sentence a bit later in the paragraph says 'previous version'. Could the specific model version used by Velasquez et al. be mentioned here?
6) ‘ D01 and D02, down to a spatial resolution of 18 km and with the convection parameterization switched on, is performed. This experiment is indicated as DEF_noconv in Table 3’
switched off, I guess.
7) These differences are in some cases of the same order of the differences between the different ..’
...order of magnitude as the differences between.
Citation: https://doi.org/10.5194/egusphere-2023-1197-RC1 -
AC1: 'Reply on RC1', Emmanuele Russo, 31 Oct 2023
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-1197/egusphere-2023-1197-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Emmanuele Russo, 31 Oct 2023
-
RC2: 'Review of "High resolution LGM climate over Europe and the Alpine region using the regional climate model WRF" by E. Russo and coauthors', Anonymous Referee #2, 14 Sep 2023
SUMMARY
Russo and coauthors apply a regional model nesting approach to dynamically simulate Last Glacial Maximum temperature and precipitation patterns over Europe. They apply the regional Weather Research and Forecasting model (WRF) with two-way nesting down to 2km resolution, based on a 28-year high-resolution atmosphere-only Community Earth System Model simulation driven by surface fields out of an equilibrium simulation with the fully coupled atmosphere/ocean version of CESM at lower resolution. They proceed to compare the fields of the simulated climate with, and without, convective parameterizations to that of a pollen-based climate reconstruction (Davis et al., 2022).
The paper is clearly structured, and fairly well written, albeit somewhat dry, dispassionate and with little curiosity for meteorological phenomena underlying the high-resolution weather simulated, and restricting itself to the bare minimum in the climatological analysis in this version of the paper. At 4s/2km resolution, investigating for example storms, or regional effects, could be quite interesting. Some further points are detailed below. All in all a paper that can be improved.
Key Points
- Goals: The authors aim to evaluate an apparently bug-fixed version of WRF against data, and contrast model uncertainties from ice height, land cover and convective parameterizations. These are great points to study. However, the study design does not really allow to understand how the different uncertainties play out against one another (nonfactorial), and looking at the figures in the results and discussion section, as well as the supplement, does not elucidate this further.- CO2: One potential reason why the model simulations appear biased dry is not discussed: Namely that the used pollen data suggests dryer conditions than warrented. The water-use efficiency under low CO2 conditions is lower, implying that plants are more stressed under similar climatic conditions [1] -- so perhaps the model is less biased than it appears.
- Discussion: Here you could bring in more depth. You could elaborate whether you expect that the results found here dependent on the version of WRF, and on CESM as a host model? The fact that the 28-year global simulation providing input does lead to significant spread in the regional model results is surprising: Where does this divergence come from? Are these nonstationary effects that suggest that the simulation period is too short? This would also imply that averaging over such a short time period may be inappropriate, weakening the justification of one of the assumptions set out (p5 last paragraph).
- Vegetation cover discussion: Given the substantial differences between the land surface conditions fed into the high-resolution simulations -- don't you expect to see effects arising simply from the strongly different land cover, for example in North Africa?Detailed Points
p2l30 "a series of LGM studies have shown..." this sentence needs references.
p2l35/p3l1-3 Here a differentiation to statistical/statistical-dynamical downscaling should be added.
p3l25 "The starting point .... are the results of earlier studies using hte same model version"... so what? What are the results of the earlier studies that imply one should do the same things? It feels like something is missing here.
p3l29 delete space after 2.3
p4l13 add space after precession
p4l4 these sentences on the glacier scheme are confusing. Does ice become supercritical in NOAH-MP? Or is what is meant that there are melt/refreeze processes in the version used in Velasquez et al. (2021) that produce unphysical temperatures?
p7 sec 2.4 -- A key weakness of Davis et al. (2022) is that it does not address the CO2-caused precipitation bias in the reconstructions, which would be expected to cause a dry bias under the low CO2 conditions.
p8 l13-15 The narrow distribution of precipitation estimates out of the pollen-based reconstructions is perhaps indicative of the dry bias (s. above)
p9 l27 remove "¨"
p9 l32-35 Indeed, the large differences between the ensemble members are remarkable. But going back to the ensemble description, can this be simply due to internal variability in the non-overlapping subsections of the 28-year simulations? (The description of the ensemble design is confusing).
p11 Code and data availability: Fix broken reference.References:
[1] Prentice et al., https://doi.org/10.1016/j.gloplacha.2022.103790Citation: https://doi.org/10.5194/egusphere-2023-1197-RC2 -
AC2: 'Reply on RC2', Emmanuele Russo, 31 Oct 2023
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-1197/egusphere-2023-1197-AC2-supplement.pdf
-
AC2: 'Reply on RC2', Emmanuele Russo, 31 Oct 2023
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 372 | 167 | 32 | 571 | 55 | 25 | 19 |
- HTML: 372
- PDF: 167
- XML: 32
- Total: 571
- Supplement: 55
- BibTeX: 25
- EndNote: 19
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 131 | 23 |
| Germany | 2 | 81 | 14 |
| Switzerland | 3 | 60 | 10 |
| China | 4 | 31 | 5 |
| France | 5 | 30 | 5 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 131
Cited
1 citations as recorded by crossref.
Emmanuele Russo
Jonathan Buzan
Sebastian Lienert
Guillaume Jouvet
Patricio Velasquez Alvarez
Basil Davis
Patrick Ludwig
Fortunat Joos
Christoph Raible
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(20683 KB) - Metadata XML
-
Supplement
(1130 KB) - BibTeX
- EndNote
- Final revised paper