Simulation of wind and solar energy generation over California with E3SM SCREAM regionally refined models at 3.25 km and 800 m resolutions

Zhang, Jishi; Golaz, Jean–Christophe; Signorotti, Matthew Vincent; Lee, Hsiang–He; Bogenschutz, Peter; Monteagudo, Minda; Ullrich, Paul Aaron; Arthur, Robert S.; Po–Chedley, Stephen; Cameron–smith, Philip; Watson, Jean–Paul

doi:10.5194/egusphere-2025-3947

Preprints

https://doi.org/10.5194/egusphere-2025-3947

Preprints

04 Sep 2025

| 04 Sep 2025

Simulation of wind and solar energy generation over California with E3SM SCREAM regionally refined models at 3.25 km and 800 m resolutions

Jishi Zhang, Jean–Christophe Golaz, Matthew Vincent Signorotti, Hsiang–He Lee, Peter Bogenschutz, Minda Monteagudo, Paul Aaron Ullrich, Robert S. Arthur, Stephen Po–Chedley, Philip Cameron–smith, and Jean–Paul Watson

Abstract. This study produces wind and solar power generation estimates derived from the US Department of Energy’s Simple Cloud-Resolving Energy Exascale Earth System Model (E3SM) Atmosphere Model (SCREAM) by leveraging regional mesh refinement over California (CARRM) simulations at 3.25 km and 800 m horizontal resolutions, using the Python wrapper of System Advisor Model (PySAM). The resulting wind and solar energy generation estimates are compared to monthly capacity factors from the Energy Information Administration (EIA), the High-Resolution Rapid Refresh (HRRR; 3 km resolution) forecast model, and the E3SM North American Regionally Refined Model (NARRM; 25 km resolution). We systematically assess the impacts of generation modeling assumptions, meteorological models, and horizontal resolution. Results show that resolution plays a dominant role for wind energy: increasing from 25 km to 3.25 km brings qualitative and quantitative improvements, most notably by resolving the phase error in the seasonal cycle found in coarser simulations. However, further refinement to 800 m offers minimal gains. SCREAM’s performance for solar generation surpasses HRRR, likely due to more accurate surface radiation. The sensitivity of PySAM to system configuration, particularly for axis-tracking modeling in photovoltaics, is also highlighted. Overall, SCREAM-RRM shows strong potential for high-resolution energy assessments, with future progress depending on more in situ observations and clearer quantification of generation modeling uncertainties.

Received: 13 Aug 2025 – Discussion started: 04 Sep 2025

Competing interests: At least one of the (co-)authors is a member of the editorial board of Geoscientific Model Development. The authors declare that they have no other competing interests.

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Status: closed

RC1:
'Comment on egusphere-2025-3947', Anonymous Referee #1, 20 Oct 2025
This is a well-written and structured manuscript, easy to read and follow meanwhile without losing the rigour of scientific work. It deals with the evaluation of atmospheric models for simulating solar and wind power in California, with multiple layers of investigation, i.e., evaluating the impact of using different atmospheric models, spatial resolutions, and solar and wind power generation models. Answering these questions would help users to understand the impact of different factors on solar and wind power calculations, and to decide on the best solution given the available resources. This is particularly important in the context of the current global transition toward a carbon-neutral energy system. I believe this manuscript could be further improved by addressing the points listed below:
The necessity or irreplaceability of the work needs to be sharpened. What makes it so important to evaluate solar and wind power generation derived from the atmospheric models (or meteorological data sets) you mentioned in California? What are their special things compared with other alternatives, or can they represent the others?

Adding information about the status quo of renewable energy installation in California and how much they cover the local electricity demand may help readers further understand the importance of your work.

The limitations of this study should be uncovered and discussed in the manuscript.

The conclusions you have made based on the results in the work, are they applicable to other regions? Or, are they applicable to other models?

Are monthly CF observations really able to identify the difference of the high- and coarse-resolution atmospheric modeling outputs by evaluating their derived power results? The fluctuations in renewables, especially in wind, have been largely filtered out at the monthly scale, for which high-resolution models are good at solving while coarse-resolution models are not. Would not evaluating the direct solar and wind output variables from atmospheric models against observations from, e.g., synoptic stations and weather masts, make more sense for this purpose? Since hourly or finer time scale CF are not reachable as mentioned in the manuscript.

Please justify in this manuscript why this work chose to use grid-cell (figure 3, 4, 5, 8; 11, 12) based evaluation, instead of evaluating over every plant (figure 9; 15) all the time?

The study indicated in line 99 that the only difference between SCREAM-3kmCARRM and SCREAM-800mCARRM is the horizontal resolution, and in line 100-101 it also indicated that the physical parameterization is sensitive to horizontal resolution. Then, why do you think using the same suit of physical parameterization in these two experiments is fair and reasonable?

Since this work recognized the effect of using different loss schemes in the generation estimate models, it should disclose the details of the wake scheme used in the PySAM/SAM, even though it is the default one. A brief intro as you did to PLUSWIND in lines 263-264 would do the job for PySAM/SAM. A great solution would also be adding words justifying why a certain wake scheme is chosen to be used in the workflow.

Now follows some technical comments:
Make sure every acronym mentioned in the abstract has its full name explained there (e.g., SCREAM-RRM is a bit surprised to me)

To be precise with the used language, solar and wind power are not renewable technologies, wind turbines and solar PV systems are. Solar and wind power are renewable energy sources. Please reformulate line 23 accordingly, and check throughout the manuscript to get rid of similar errors.

In line 97 the study used “water years”, which is a field-specific terminology. It is suggested to explain its meaning at its first occurrence (line 97, while its first explanation now is in line 172) or use a more common term like weather years.

The headline index might be wrong in section 2.1 and 2.2. Now it appears nothing presented under the headline 2.1, which should not be the case.

Please check the entire Methods section, many acronyms mentioned without introducing its full name at the 1st occurrence, such as SVD, MPAS, HICCUP

Line 216, repetitive to line 114-115, please reformulate and avoid using the exact same sentence

Line 274, repetitive word used “preprocessing”

Please add references to underpin your perspectives in line 315-316

Figure 3: suggest to rearrange the legend into multiple columns (like 3) to save space, and to add index for subplots like a), b), c), which applies to other figures as well in this manuscript

Figure 4: suggest to keep legend only one time for one row since they are the same, the same applies to Figure 5 and other figures from solar

Line 423: I don’t see an underestimation in SoCal from HRRR, at least not from the bold line. Also it seems to me that NorCal has the least discrepancy present instead of SoCal.

Figure 6: to save space, suggest to use a shared colorbar, put repetitive names as row name and column name, and show latitudes and longitudes only at the first column and the last row, the same applies to other figures from solar, otherwise some numbers are hard to read at the moment in these figures

Figure 7: texts are hard to recognize, save space by reducing repetitive information as suggested before, try to use a landscape layout or split them into two panels to improve the visualization

Figure 14: why not showing the “difference” subplots you have in Figure 7?

Figure 9 and 15: try to adjust the legend so that it would not overlap the plot
Citation: https://doi.org/10.5194/egusphere-2025-3947-RC1
- AC1: 'Reply on RC1', Jishi Zhang, 22 Mar 2026
  
  We thank the reviewer for their careful and thorough assessment of our manuscript. Please see the attached document for a detailed response to each point.
  
  Citation: https://doi.org/10.5194/egusphere-2025-3947-AC1
RC2:
'Comment on egusphere-2025-3947', Anonymous Referee #2, 08 Feb 2026

Please see the attached file for comments on egusphere-2025-3947.

Citation: https://doi.org/10.5194/egusphere-2025-3947-RC2
- AC2: 'Reply on RC2', Jishi Zhang, 22 Mar 2026
  
  We thank the reviewer for their careful and thorough assessment of our manuscript. Please see the attached document for a detailed response to each point.
  
  Citation: https://doi.org/10.5194/egusphere-2025-3947-AC2

Status: closed

RC1:
'Comment on egusphere-2025-3947', Anonymous Referee #1, 20 Oct 2025
This is a well-written and structured manuscript, easy to read and follow meanwhile without losing the rigour of scientific work. It deals with the evaluation of atmospheric models for simulating solar and wind power in California, with multiple layers of investigation, i.e., evaluating the impact of using different atmospheric models, spatial resolutions, and solar and wind power generation models. Answering these questions would help users to understand the impact of different factors on solar and wind power calculations, and to decide on the best solution given the available resources. This is particularly important in the context of the current global transition toward a carbon-neutral energy system. I believe this manuscript could be further improved by addressing the points listed below:
The necessity or irreplaceability of the work needs to be sharpened. What makes it so important to evaluate solar and wind power generation derived from the atmospheric models (or meteorological data sets) you mentioned in California? What are their special things compared with other alternatives, or can they represent the others?

Adding information about the status quo of renewable energy installation in California and how much they cover the local electricity demand may help readers further understand the importance of your work.

The limitations of this study should be uncovered and discussed in the manuscript.

The conclusions you have made based on the results in the work, are they applicable to other regions? Or, are they applicable to other models?

Are monthly CF observations really able to identify the difference of the high- and coarse-resolution atmospheric modeling outputs by evaluating their derived power results? The fluctuations in renewables, especially in wind, have been largely filtered out at the monthly scale, for which high-resolution models are good at solving while coarse-resolution models are not. Would not evaluating the direct solar and wind output variables from atmospheric models against observations from, e.g., synoptic stations and weather masts, make more sense for this purpose? Since hourly or finer time scale CF are not reachable as mentioned in the manuscript.

Please justify in this manuscript why this work chose to use grid-cell (figure 3, 4, 5, 8; 11, 12) based evaluation, instead of evaluating over every plant (figure 9; 15) all the time?

The study indicated in line 99 that the only difference between SCREAM-3kmCARRM and SCREAM-800mCARRM is the horizontal resolution, and in line 100-101 it also indicated that the physical parameterization is sensitive to horizontal resolution. Then, why do you think using the same suit of physical parameterization in these two experiments is fair and reasonable?

Since this work recognized the effect of using different loss schemes in the generation estimate models, it should disclose the details of the wake scheme used in the PySAM/SAM, even though it is the default one. A brief intro as you did to PLUSWIND in lines 263-264 would do the job for PySAM/SAM. A great solution would also be adding words justifying why a certain wake scheme is chosen to be used in the workflow.

Now follows some technical comments:
Make sure every acronym mentioned in the abstract has its full name explained there (e.g., SCREAM-RRM is a bit surprised to me)

To be precise with the used language, solar and wind power are not renewable technologies, wind turbines and solar PV systems are. Solar and wind power are renewable energy sources. Please reformulate line 23 accordingly, and check throughout the manuscript to get rid of similar errors.

In line 97 the study used “water years”, which is a field-specific terminology. It is suggested to explain its meaning at its first occurrence (line 97, while its first explanation now is in line 172) or use a more common term like weather years.

The headline index might be wrong in section 2.1 and 2.2. Now it appears nothing presented under the headline 2.1, which should not be the case.

Please check the entire Methods section, many acronyms mentioned without introducing its full name at the 1st occurrence, such as SVD, MPAS, HICCUP

Line 216, repetitive to line 114-115, please reformulate and avoid using the exact same sentence

Line 274, repetitive word used “preprocessing”

Please add references to underpin your perspectives in line 315-316

Figure 3: suggest to rearrange the legend into multiple columns (like 3) to save space, and to add index for subplots like a), b), c), which applies to other figures as well in this manuscript

Figure 4: suggest to keep legend only one time for one row since they are the same, the same applies to Figure 5 and other figures from solar

Line 423: I don’t see an underestimation in SoCal from HRRR, at least not from the bold line. Also it seems to me that NorCal has the least discrepancy present instead of SoCal.

Figure 6: to save space, suggest to use a shared colorbar, put repetitive names as row name and column name, and show latitudes and longitudes only at the first column and the last row, the same applies to other figures from solar, otherwise some numbers are hard to read at the moment in these figures

Figure 7: texts are hard to recognize, save space by reducing repetitive information as suggested before, try to use a landscape layout or split them into two panels to improve the visualization

Figure 14: why not showing the “difference” subplots you have in Figure 7?

Figure 9 and 15: try to adjust the legend so that it would not overlap the plot
Citation: https://doi.org/10.5194/egusphere-2025-3947-RC1
- AC1: 'Reply on RC1', Jishi Zhang, 22 Mar 2026
  
  We thank the reviewer for their careful and thorough assessment of our manuscript. Please see the attached document for a detailed response to each point.
  
  Citation: https://doi.org/10.5194/egusphere-2025-3947-AC1
RC2:
'Comment on egusphere-2025-3947', Anonymous Referee #2, 08 Feb 2026

Please see the attached file for comments on egusphere-2025-3947.

Citation: https://doi.org/10.5194/egusphere-2025-3947-RC2
- AC2: 'Reply on RC2', Jishi Zhang, 22 Mar 2026
  
  We thank the reviewer for their careful and thorough assessment of our manuscript. Please see the attached document for a detailed response to each point.
  
  Citation: https://doi.org/10.5194/egusphere-2025-3947-AC2

Data sets

Code, model, and analysis data for simulation of wind and solar energy generation over California with E3SM SCREAM regionally refined models at 3.25 km and 800 m resolutions Jishi Zhang https://zenodo.org/records/16809290

Model code and software

Viewed

Total article views: 2,413 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
1,797	578	38	2,413	37	31

HTML: 1,797
PDF: 578
XML: 38
Total: 2,413
BibTeX: 37
EndNote: 31

Views and downloads (calculated since 04 Sep 2025)

Month	HTML	PDF	XML	Total
Sep 2025	1,361	34	4	1,399
Oct 2025	100	33	7	140
Nov 2025	36	42	4	82
Dec 2025	20	56	7	83
Jan 2026	86	72	8	166
Feb 2026	54	87	1	142
Mar 2026	57	132	6	195
Apr 2026	83	122	1	206

Cumulative views and downloads (calculated since 04 Sep 2025)

Month	HTML	PDF	XML	Total
Sep 2025	1,361	34	4	1,399
Oct 2025	100	33	7	140
Nov 2025	36	42	4	82
Dec 2025	20	56	7	83
Jan 2026	86	72	8	166
Feb 2026	54	87	1	142
Mar 2026	57	132	6	195
Apr 2026	83	122	1	206

Viewed (geographical distribution)

Total article views: 2,265 (including HTML, PDF, and XML) Thereof 2,265 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 01 May 2026

Short summary

We ran a convection-permitting model with regional mesh refinement (3.25 km and 800 m) to simulate present-day wind and solar capacity factors over California, coupling it to an energy generation model. The high-resolution models captured realistic seasonal and diurnal cycles, with wind markedly better than a 25 km model and solar outperforming a 3 km operational forecast. We highlight the critical role of resolution, modeling assumptions, and data reliability in renewable energy assessment.


Total:	0
HTML:	0
PDF:	0
XML:	0