the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 20 Mar 2025
Arctic temperature and precipitation extremes in present-day and future storyline-based variable resolution Community Earth System Model simulations
Abstract. Over the last few decades, the Arctic region has warmed up at a greater rate than elsewhere on the globe, partly resulting from the on-going loss of sea ice and seasonal snow over land. It is projected that the amplified warming of the surface will continue in the future. In addition, the intensity and frequency of temperature and precipitation means and extremes are projected to change, which may pose serious threats for human infrastructure and livelihoods. To assess (future) climate extremes, advanced modelling approaches with (regionally) refined resolution could be helpful.
In this study, we use the variable-resolution Community Earth System Model version 2.2 (VR-CESM) to evaluate and assess present-day and future temperature and precipitation extremes, such as heat waves and heavy precipitation, over the Arctic. Applying a globally uniform 1° grid and a VR grid with regional grid refinements to 28 km over the Arctic and Antarctica, we run 30-year present-day (1985–2014), 10-year present-day (2005–2014), and future (2090–2099) simulations with interactive atmosphere and land surface models, and prescribed sea ice and sea surface temperatures. We use the 30-year simulation to evaluate the ability of the VR grid to simulate climate extremes by comparison with gridded outputs of the globally uniform 1° grid, reanalysis-based datasets, and a regional climate model. The 10-year simulations follow two storylines of Arctic climate change representing a combination of strong/weak Arctic tropospheric warming and strong/weak SST warming in the Barents-Kara Seas and are used to assess future climate extremes by focussing on temperature and precipitation extremes. The outcomes show that the VR grid generally performs better in simulating precipitation extremes, while the globally uniform 1° grid generally performs better in simulating temperature extremes. Future projections suggest that high temperature extremes will generally increase both in intensity and duration, whereas low temperature extremes will decrease in intensity and duration, especially over regions dominated by SST warming and large sea ice loss. Further, wet precipitation extremes are projected to increase in intensity and frequency. The outcomes of this study may contribute to an improved understanding on future climate extremes and its implications.
- Preprint
(33177 KB) - Metadata XML
-
Supplement
(30983 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2025-1070', Anonymous Referee #1, 28 Apr 2025
Review of ‘Arctic temperature and precipitation extremes in present-day and future storyline-based variable resolution Community Earth System Model Simulations’ by Wijngaard et al
This article examines changes to mean and extreme metrics of temperature and precipitation in VR CESM simulations, where mesh refinement is over the poles. Comparisons are made with standard resolution CESM output. Simulations of present day and of two storylines of future change based on other work are performed with and without mesh refinement over the region of interest. They find that high extremes of temperature and precipitation will increase in many of the study regions encompassing high latitude land and ocean.
This is a well written and thorough paper. I think that the application of different storylines of future change within a high-resolution model is a novel and interesting way to use high-resolution modelling at a lower computational cost. Throughout there is thought put into trying to physically understand the results, and with trying to square these results with past studies, placing it nicely within the larger body of literature. Figures are high quality and contain a lot of information that is well presented.
I have mainly minor comments I’d like to see addressed before publication.
General:
In terms of readability, using capital letters to refer to simulation names (e.g. POLARRES), regions (e.g ECG, SCA..), and metrics of extremes (e.g. WSDI, CSDI) did at times make it a bit difficult to understand. Using bolding, italics, or perhaps single quotation marks to differentiate between them (depending on what the journal might allow), or even simply spelling out the whole name of, for example, a region, would improve ease of reading.
The Data & Methods are very comprehensive but probably a bit more specific and detailed than is necessary for this paper, more suited toward a paper published in a journal about geophysical model development. For example, details of CLM5, computational costs, parameter tuning, where reanalysis data is retrieved from, an unused dataset, etc could be pared down to the necessary information not available in other published work.
I think it’s important to keep in mind that agreement/disagreement between a reanalysis product and a model over regions of sparse observations should always be taken with a grain of salt, and that reanalyses are not equivalent to observations. Treatment of differences between a model and a reanalysis product in some regions is akin to comparing model and model, and reanalyses themselves have many biases. Perhaps a bit more discussion to this end would be beneficial.
Specific/Technical
L 13 & elsewhere: ‘SST warming’ should either say ‘sea surface warming’ or ‘SST increase’ because a temperature can’t warm.
L30-32: I don’t think this is a complete treatment of the literature on whether cold extremes are becoming more common in the mid-latitudes, as more recent studies (e.g. Cohen et al 2023, Van Olgdenberg 2019, Blackport 2024) have not come to the conclusion that cold extremes are occurring more often. Additionally, it’s not particularly relevant to the regions discussed later in the paper so probably best to just not include it.
Paragraph beginning at line 81: Might want to add references to Morris et al papers from 2023-2025 using VR CESM to look at wind extremes. They find that sometimes the low resolution CESM gets a different sign of response than VR CESM for extreme winds, and perhaps it’s something worth looking into for the Arctic region as well.
L95: I’m curious why there is an enhancement of resolution over the Antarctic as well?
L283-285: difficult sentence to understand, I’d suggest re-writing to clarify.
L353: missing comma before ‘as for example’
L408: I think here and below should read ‘cold’ anomalies and not ‘cooling’ anomalies.
L471: I find this sentence a bit confusing. Do wet regions become more variable?
L500: comma after regimes, remove ‘as well’ from end.
L532: Does POLARRES produce more storms/more strong storms than the reanalyses do for this region?
L547-549: Is this sentence at odds with that at L471?
Figure 2 caption & elsewhere: ‘outputted’ -> ‘output’
Citation: https://doi.org/10.5194/egusphere-2025-1070-RC1 - AC1: 'Reply on RC1', Rene Wijngaard, 25 Jun 2025
-
RC2: 'Comment on egusphere-2025-1070', Anonymous Referee #2, 02 May 2025
My comments are provided in the attached .pdf file.
- AC2: 'Reply on RC2', Rene Wijngaard, 25 Jun 2025
Status: closed
-
RC1: 'Comment on egusphere-2025-1070', Anonymous Referee #1, 28 Apr 2025
Review of ‘Arctic temperature and precipitation extremes in present-day and future storyline-based variable resolution Community Earth System Model Simulations’ by Wijngaard et al
This article examines changes to mean and extreme metrics of temperature and precipitation in VR CESM simulations, where mesh refinement is over the poles. Comparisons are made with standard resolution CESM output. Simulations of present day and of two storylines of future change based on other work are performed with and without mesh refinement over the region of interest. They find that high extremes of temperature and precipitation will increase in many of the study regions encompassing high latitude land and ocean.
This is a well written and thorough paper. I think that the application of different storylines of future change within a high-resolution model is a novel and interesting way to use high-resolution modelling at a lower computational cost. Throughout there is thought put into trying to physically understand the results, and with trying to square these results with past studies, placing it nicely within the larger body of literature. Figures are high quality and contain a lot of information that is well presented.
I have mainly minor comments I’d like to see addressed before publication.
General:
In terms of readability, using capital letters to refer to simulation names (e.g. POLARRES), regions (e.g ECG, SCA..), and metrics of extremes (e.g. WSDI, CSDI) did at times make it a bit difficult to understand. Using bolding, italics, or perhaps single quotation marks to differentiate between them (depending on what the journal might allow), or even simply spelling out the whole name of, for example, a region, would improve ease of reading.
The Data & Methods are very comprehensive but probably a bit more specific and detailed than is necessary for this paper, more suited toward a paper published in a journal about geophysical model development. For example, details of CLM5, computational costs, parameter tuning, where reanalysis data is retrieved from, an unused dataset, etc could be pared down to the necessary information not available in other published work.
I think it’s important to keep in mind that agreement/disagreement between a reanalysis product and a model over regions of sparse observations should always be taken with a grain of salt, and that reanalyses are not equivalent to observations. Treatment of differences between a model and a reanalysis product in some regions is akin to comparing model and model, and reanalyses themselves have many biases. Perhaps a bit more discussion to this end would be beneficial.
Specific/Technical
L 13 & elsewhere: ‘SST warming’ should either say ‘sea surface warming’ or ‘SST increase’ because a temperature can’t warm.
L30-32: I don’t think this is a complete treatment of the literature on whether cold extremes are becoming more common in the mid-latitudes, as more recent studies (e.g. Cohen et al 2023, Van Olgdenberg 2019, Blackport 2024) have not come to the conclusion that cold extremes are occurring more often. Additionally, it’s not particularly relevant to the regions discussed later in the paper so probably best to just not include it.
Paragraph beginning at line 81: Might want to add references to Morris et al papers from 2023-2025 using VR CESM to look at wind extremes. They find that sometimes the low resolution CESM gets a different sign of response than VR CESM for extreme winds, and perhaps it’s something worth looking into for the Arctic region as well.
L95: I’m curious why there is an enhancement of resolution over the Antarctic as well?
L283-285: difficult sentence to understand, I’d suggest re-writing to clarify.
L353: missing comma before ‘as for example’
L408: I think here and below should read ‘cold’ anomalies and not ‘cooling’ anomalies.
L471: I find this sentence a bit confusing. Do wet regions become more variable?
L500: comma after regimes, remove ‘as well’ from end.
L532: Does POLARRES produce more storms/more strong storms than the reanalyses do for this region?
L547-549: Is this sentence at odds with that at L471?
Figure 2 caption & elsewhere: ‘outputted’ -> ‘output’
Citation: https://doi.org/10.5194/egusphere-2025-1070-RC1 - AC1: 'Reply on RC1', Rene Wijngaard, 25 Jun 2025
-
RC2: 'Comment on egusphere-2025-1070', Anonymous Referee #2, 02 May 2025
My comments are provided in the attached .pdf file.
- AC2: 'Reply on RC2', Rene Wijngaard, 25 Jun 2025
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 378 | 196 | 18 | 592 | 46 | 21 | 36 |
- HTML: 378
- PDF: 196
- XML: 18
- Total: 592
- Supplement: 46
- BibTeX: 21
- EndNote: 36
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1