the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 26 May 2026
GCAM-China-v8: An Integrated Global-to-Provincial Framework for Assessing China’s Energy and Emission Futures
Abstract. This paper describes Global Change Analysis Model-China version 8 (GCAM-China-v8), an open-source integrated assessment model that represents interactions among energy, economic, and water systems within a globally consistent framework, with explicit subnational representation for China. GCAM-China-v8 builds on the GCAM and represents the world as 31 geopolitical regions outside China, while disaggregating China into 31 province-level regions to capture regional heterogeneity. GCAM-China-v8 can be used to explore how changes in socioeconomic drivers, technological progress, and policy assumptions affect energy and water demand and production at the subnational level in China, while maintaining consistency with national and international boundary conditions. This paper documents the model structure and data inputs, with particular emphasis on the methodological updates introduced in GCAM-China-v8, including enhanced sectoral and temporal representations. To demonstrate the capabilities of the updated model, we apply GCAM-China-v8 to two illustrative scenarios with contrasting assumptions about future socioeconomic development and energy system transformation. This paper provides a transparent and extensible modeling framework for future research on China’s long-term energy and climate transitions. It also contributes to the broader Integrated Assessment Models (IAMs) community by advancing national-scale model development within an open and consistent framework.
- Preprint
(2761 KB) - Metadata XML
-
Supplement
(815 KB) - BibTeX
- EndNote
Status: open (until 30 Jul 2026)
-
CEC1: 'Comment on egusphere-2026-1919 - No compliance with the policy of the journal', Juan Antonio Añel, 21 Jun 2026
reply
-
CC1: 'Reply on CEC1', Shuling Xu, 23 Jun 2026
reply
Dear Prof. Arel,
Thank you for your comments.
We fully acknowledge the journal’s requirement that all code and data necessary to reproduce the results must be openly available at the time of submission.
The GCAM-China v8 modeling framework has already been fully open-sourced and is publicly accessible at: https://github.com/umd-cgs/gcam-china/releases. In addition, all scenario configuration files and model outputs used in this study have been archived and made publicly available on Zenodo: https://doi.org/10.5281/zenodo.20790929.
We confirm that both the code repository and the Zenodo archive are openly accessible without restriction and fully comply with the GMD Code and Data Policy requirements.
Please let us know if any additional modifications are required.
Shuling Xu
Citation: https://doi.org/10.5194/egusphere-2026-1919-CC1 -
CC2: 'Reply on CEC1', Shuling Xu, 23 Jun 2026
reply
Dear Editor Juan Antonio Añel,
Thank you for your comments.
We fully acknowledge the journal’s requirement that all code and data necessary to reproduce the results must be openly available at the time of submission.
The GCAM-China v8 modeling framework has already been fully open-sourced and is publicly accessible at: . In addition, all scenario configuration files and model outputs used in this study have been archived and made publicly available on Zenodo: .
We confirm that both the code repository and the Zenodo archive are openly accessible without restriction and fully comply with the GMD Code and Data Policy requirements.
Please let us know if any additional modifications are required.
Sincerely,
Shuling Xu
Citation: https://doi.org/10.5194/egusphere-2026-1919-CC2 -
CEC2: 'Reply on CC2', Juan Antonio Añel, 23 Jun 2026
reply
Dear authors,
Thanks for your reply. However, we can not accept GitHub to host the assets of your manuscript. GitHub is not suitable for scientific publishing. This is clearly explained in the policy of the journal which you should have consulted before submitting your manuscript, and read after I pointed it out in my first comment.
Therefore, you must store the code in a repository we can accept. Please, reply to this comment with the information about it, including link and permanent handler.
Juan A. Añel
Geosci. Model Dev. Executive Editor
Citation: https://doi.org/10.5194/egusphere-2026-1919-CEC2 -
CC3: 'Reply on CEC2', Shuling Xu, 23 Jun 2026
reply
Dear Editor Juan Antonio Añel,
Thank you for your clarification.
The GCAM-China v8 modeling framework is also available at the following Zenodo repository: , including the model code and all required input datasets. In addition, all scenario configuration files and model simulation outputs in this study have been archived and made publicly available on Zenodo: .
Sincerely,
Shuling Xu
Citation: https://doi.org/10.5194/egusphere-2026-1919-CC3 -
CEC3: 'Reply on CC3', Juan Antonio Añel, 23 Jun 2026
reply
Dear authors,
In your reply the information about the repository is missing. IT is the second time this happens in one of your replies. Please, reply with the requested information, and double-check that your reply contains it before submitting.
Juan A. Añel
Geosci. Model Dev. Executive Editor
Citation: https://doi.org/10.5194/egusphere-2026-1919-CEC3 -
AC1: 'Reply on CEC3', Yang Ou, 23 Jun 2026
reply
Dear Editor
Sorry about the confusion and I highly respect GMD's open-source policy and take this matter seriously!
Our GCAM-China v8 model is openly available at: https://doi.org/10.5281/zenodo.19471594,
The source code, input files, and all original model outputs of our four scenarios are publicly available at: https://doi.org/10.5281/zenodo.20790929.
We will keep working at our best to support open-source community and facilitate the review process here, please don't hesitate to reach out if there's anything else we could assist.
Yang Ou (corresponding author)
On behalf of all authorsCitation: https://doi.org/10.5194/egusphere-2026-1919-AC1 -
CEC4: 'Reply on AC1', Juan Antonio Añel, 23 Jun 2026
reply
Dear authors,
Many thanks for the reply. We can consider now the current version of your manuscript in compliance with the policy of the journal.
Juan A. Añel
Geosci. Model Dev. Executive Editor
Citation: https://doi.org/10.5194/egusphere-2026-1919-CEC4
-
CEC4: 'Reply on AC1', Juan Antonio Añel, 23 Jun 2026
reply
-
AC1: 'Reply on CEC3', Yang Ou, 23 Jun 2026
reply
-
CEC3: 'Reply on CC3', Juan Antonio Añel, 23 Jun 2026
reply
-
CC3: 'Reply on CEC2', Shuling Xu, 23 Jun 2026
reply
-
CEC2: 'Reply on CC2', Juan Antonio Añel, 23 Jun 2026
reply
-
CC1: 'Reply on CEC1', Shuling Xu, 23 Jun 2026
reply
-
RC1: 'Comment on egusphere-2026-1919', Anonymous Referee #1, 24 Jun 2026
reply
The manuscript presents GCAM-China-v8, an updated China-focused version of GCAM with provincial-level representation and enhanced sectoral detail. The model is potentially valuable for the integrated assessment modeling community and for studies of China’s long-term energy and emissions pathways. However, several aspects of the manuscript require further clarification before publication.
- The abstract states that GCAM-China-v8 includes enhanced sectoral and temporal representations (Line 59), but these two improvements are not sufficiently highlighted in the subsequent model description. The authors should clearly identify what has been improved in sectoral resolution and temporal representation relative to previous GCAM-China versions.
- The consideration of vintage-specific existing power infrastructure, particularly coal-fired power units, appears to be an important methodological improvement. This contribution should be emphasized earlier in the manuscript, for example in the abstract, introduction, or model overview.
- The main novelty of GCAM-China-v8 relative to earlier GCAM-China, GCAM-TU, CNCAP, GCAM-China-ABaCAS, and other China-focused GCAM variants remains insufficiently clear. I suggest adding a compact comparison table that explicitly summarizes differences in model version, base year, spatial resolution, sectoral coverage, technology detail, temporal structure, policy representation, openness, and typical applications.
- For water withdrawal and water consumption, the manuscript should explain the difference between the model base-year data and official statistics. For example, China’s 2024 Water Resources Bulletin reports total national water use of approximately 600 billion m3.
- Please clarify whether the model explicitly represents mitigation technologies or control measures for non-CO2 greenhouse gases. This is especially important for assessing China’s 2035 NDC-related climate targets. The authors should explain how non-CO2 mitigation contributions are quantified, how they interact with CO2 mitigation.
- The calibration procedure should be described more systematically. Please state the calibration targets, benchmark datasets, calibration years, and sector-specific adjustment methods. It would be helpful to provide a table summarizing calibration by sector, including energy supply, electricity, industry, building, transport, water, and land.
- The REF and CN scenarios need a more quantitative description. Please provide a concise table summarizing key assumptions. This would help readers distinguish model assumptions from model results.
- The simplified representation of inter-provincial electricity trade requires stronger justification. Please clarify whether transmission capacity, congestion, losses, and grid topology are ignored, approximated, or implicitly represented. In addition, please explain whether the model explicitly excludes electricity transmission between provinces that are not geographically adjacent and are not connected by ultra-high-voltage transmission lines.
- The nuclear incentive policy example appears somewhat arbitrary in the current manuscript. The authors should either justify the policy design more clearly, including its empirical or policy basis, or move it to the SI as a methodological demonstration rather than presenting it as a central scenario result.
- Please check the citation to “China Industrial Statistical Yearbook (2026).” This appears likely to be erroneous unless the 2026 edition is already available and intentionally used. If not, the citation should be corrected to the appropriate name (for example China Industrial Statistical Yearbook 2025 (2026)).
- Please ensure consistent terminology for water withdrawal, water consumption, water demand, and water use throughout the manuscript and supplementary materials.
Overall, the manuscript documents an important model development effort, but its contribution would be clearer if the authors more explicitly positioned GCAM-China-v8 relative to previous China-focused GCAM models and strengthened the documentation of calibration, scenario assumptions, water accounting, electricity trade, and non-CO2 mitigation.
Citation: https://doi.org/10.5194/egusphere-2026-1919-RC1 -
RC2: 'Comment on egusphere-2026-1919', Anonymous Referee #2, 25 Jun 2026
reply
General comments:
This paper presents GCAM-China-v8, an open-source integrated assessment model (IAM) that disaggregates China into 31 province-level regions within a globally consistent framework. By bridging the gap between global climate targets and local implementation, this model serves as a tool for analyzing subnational energy and emission pathways. National-scale IAMs are essential because climate policies are increasingly implemented at the subnational level. GCAM-China-v8 provides the necessary resolution to investigate not just if China can reach carbon neutrality, but where and how these transitions will physically and socioeconomically occur across its provinces.
The authors should be commended for the massive data-gathering and harmonization effort required to calibrate a 31-province model to a base year as recent as 2021. Unlike many previous China-focused models that were project-specific or poorly documented, GCAM-China-v8 is a community-shared framework, significantly enhancing the reproducibility and extensibility of China’s climate research. A major technical advance is the explicit disaggregation of coal power by construction vintage. This allows the model to simulate realistic, staggered retirement pathways based on the actual age structure of provincial fleets rather than assuming a uniform national phase-out. The model disaggregates residential building demand into 20 consumer groups based on income deciles and urban-rural status. This enables vital research into energy equity and how technology adoption is driven by income heterogeneity. Reconstructing provincial supply curves for wind and solar using 10 km × 10 km grid-cell data ensures that the model reflects the physical and economic limits of renewable expansion more accurately than previous versions.
However, there are several limitations on the methodological choices and room for model improvements. The paper is worth considering for the publication if the following specific comments are addressed.
Specific comments:
While the paper represents a significant step forward, several architectural choices and omissions require further clarification or more rigorous critical analysis.
- Socioeconomic projections: The future population and GDP growth at the provincial level follow the SSP growth rates. This ignores the region-specific growth patterns, internal migrations and economic activities. This is critical for the future socioeconomic dynamics across the provinces. Further, the GDP values are used to project the sectoral demands. This typically provides the model with a downscaled nature rather than capturing the provincial level nuances, which further limits the reliability of the results in provincial scale.
- Ambiguity in international transport: While the main text mentions "international freight shipping" as one of four final transport demands, it is not explicitly clear how these emissions and energy demands are treated or allocated. The supplementary information indicates that “freight aircrafts” are excluded contradicting the main text. Such information needs to be made explicit and clear in the main text.
- Temporal resolution vs. high-VRE findings: The model predicts a Carbon Neutrality scenario where wind and solar reach 75% of total generation by 2060. However, the power sector still operates on an annual balance. This methodological choice is a major limitation: it cannot capture the diurnal and seasonal variability of renewables or the physical feasibility of the storage and grid flexibility required to support such a mix.
- Top-down reconciliation risks: The authors apply "province- and sector-specific scaling factors" when bottom-up facility data exceeds IEA national totals. The magnitude of these scaling factors is not disclosed. This is a critical omission; if the adjustments are large, the top-down national constraints may be masking the very "regional heterogeneity" the model claims to capture.
- Simplistic trade and transmission: Electricity trade is modeled using empirical ratios rather than physical grid topology or transmission constraints. Since grid bottlenecks are a primary driver of provincial energy decisions in China, this simplification may lead to overly optimistic projections of inter-provincial energy sharing.
- Static spatial mappings: The mapping of provinces to water basins is fixed at 2010 values. This assumes spatial demand patterns remain unchanged for 50 years, ignoring the massive geographic shifts in industrial and urban activity projected by the model itself, which could mask future regional water stress.
- LCOE curve parametrization: Provincial supply curves for wind and solar are shaped by three parameters: MaxResource, MidPrice, and CurveEx. The paper needs more explanation on how grid-cell technical potentials (10km x 10km) were aggregated into these specific parameters. Specifically, how were land-use restrictions and terrain constraints weighted when moving from a high-resolution grid to a provincial-level curve?
- Disaggregation gaps in non-CO₂ emissions: China is the world’s largest emitter of methane, largely from coal sector followed by agriculture. For a model focused on provincial transition pathways, leaving out the largest methane sources (e.g., coal mine methane and rice cultivation) from the provincial resolution is a significant omission.
- CCS storage and CO2 transport: Several types of CCS technologies are mentioned, such as CCS in energy supply, industries and CDR technologies. However, it’s not mentioned how the provincial geological storage potential is applied in the calculation. The assumptions such as cumulative CCS and geological capacity relationship and annual constraints are unclear. Furthermore, mismatch in source of carbon capture and destination of geological storage, as well as geological storage limitation in a particular province would require CO2 transport between provinces. Such mechanisms are not described.
- Hydrogen and its derivatives: Hydrogen shows a significant share in industry and transport sector in the Carbon Neutrality scenario. The energy production methods from hydrogen and its derivatives (synthetic fuels, ammonia) are not described.
- Results vs historical data: Likewise figure 15, major variables of energy system, land-use and agriculture should be compared with statistical values. Further, there should be some explanations why calibration year has gap with statistical record in emissions.
- No vintage in building sector: The building sector uses a Gompertz function to relate floorspace to income, but it does not include building vintaging. This means existing building stock is treated the same as new construction. The model's inability to represent the inertia of existing insulation and efficiency levels needs more justification. How does the model prevent unrealistic, instantaneous improvements in building efficiency across the entire provincial stock?
- Top-down vs bottom-up: While the major attraction of the model is disaggregation into provincial scale, calculations involve both top-down and bottom-up approaches. Sectoral energy consumptions are reconciled with the IEA Energy Balances, which are further used to estimate the sectoral non-CO2 emissions. This erases the provincial nuances and reliability on such scale. The implications of such mixed approaches should be critically addressed.
- Lack of critical analysis in discussion: While Section 4 lists model limitations, it lacks a critical analysis of how the methodological choices impact the findings. Specifically, the authors should explicitly delineate the boundaries of the model's utility, guiding readers/users on the contexts where its provincial-level insights are most robust and where structural simplifications necessitate a cautious interpretation of its projected transition pathways.
Minor comments
- Novelty in model development: The paper mentions several versions of GCAM-China, some of which already considered provincial level development. However, the major advances of GCAM-China-v8 compared to previous GCAM-China versions are unclear in the abstract and introduction section. There should be an explicit mention regarding what are new developments in this version compared to previous versions.
- Visualizations: The differences between the compared scenarios are not clear in figure 8, figure 9, figure 13. Adding a difference chart would be better for such small changes.
- Reproducibility: Information required to reproduce the model simulation, such as system requirements, installation guides, specifications of environment on which model was run, simulation time, etc. are important for other users. It’s better to have such information in the supplementary material.
Citation: https://doi.org/10.5194/egusphere-2026-1919-RC2
Model code and software
GCAM-China release website GCAM-China contributors https://doi.org/10.5281/zenodo.19471594
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 187 | 69 | 7 | 263 | 17 | 9 | 10 |
- HTML: 187
- PDF: 69
- XML: 7
- Total: 263
- Supplement: 17
- BibTeX: 9
- EndNote: 10
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Dear authors,
Unfortunately, after checking your manuscript, it has come to our attention that it does not comply with our "Code and Data Policy".
https://www.geoscientific-model-development.net/policies/code_and_data_policy.html
in the "Code and Data Availability" section of your manuscript you state "The processing code and scenario data of this study will be publicly available upon publication." We can not accept this statement. The policy of the journal clearly establishes that all the code and data necessary and related to a manuscript must be published and openly available without restriction before submitting a manuscript to the journal. Therefore, please, publish the processing code and scenario data in one of the repositories that we can accept according to our policy, and reply to this comment with their details. The GMD review and publication process depends on reviewers and community commentators being able to access, during the discussion phase, the code and data on which a manuscript depends, and on ensuring the provenance of replicability of the published papers for years after their publication.
Later, if the Topical Editor decides to continue with the review or publication process of your manuscript and you are requested to upload a new version of it, then The 'Code and Data Availability’ section of your manuscript must also be modified to cite the new repository locations, and corresponding references added to the bibliography.
I must note that if you do not fix this problem, we cannot continue with the peer-review process or accept your manuscript for publication in GMD.
Juan A. Añel
Geosci. Model Dev. Executive Editor