AI Maps China’s Renewable Energy Infrastructure in New Study
Chinese scientists have completed the nation’s first comprehensive, high-precision inventory of wind and solar energy infrastructure, using artificial intelligence to analyze satellite imagery and map hundreds of thousands of renewable energy facilities across the country. The landmark study, published in Nature, reveals that nationwide inter-provincial coordination could unlock approximately 100 billion kilowatt-hours of additional clean energy annually without adding a single new turbine or solar panel.
A Digital Twin of the Nation’s Clean Energy Assets
Led by researchers from Peking University’s School of Earth and Space Sciences in collaboration with Alibaba DAMO Academy, the team processed 7.56 terabytes of 0.5-meter resolution satellite imagery covering all of China. Using deep learning models based on open-source frameworks, they identified and mapped 319,972 solar photovoltaic facilities and 91,609 wind turbines across 1,915 Chinese counties.
According to Xinhua News Agency, Prof. Liu Yu of Peking University described the resulting dataset as providing a “bird’s eye view of the nation’s renewable energy landscape.” The research, published in Nature on May 20, 2026, represents the first time China’s renewable energy infrastructure has been comprehensively mapped at high resolution using actual facility locations rather than hypothetical deployments.
The Challenge: Wasted Wind and Solar Power
Despite China’s position as the world’s largest and fastest-growing renewable energy market, the country has long grappled with “curtailment” — where generated renewable electricity goes to waste because of mismatches between when and where power is produced. In 2025 alone, China added over 430 GW of new wind and solar capacity, with renewable installations surpassing thermal power additions for the first time, according to the National Energy Administration.
Yet the core issue is not insufficient generation capacity, but spatial and temporal mismatches: wind often peaks at night while solar peaks during the day, and renewable-rich regions in western China are far from demand centers along the eastern coast. As People’s Daily reported, Prof. Liu Yu noted that “in the past, people knew from experience that wind and solar could ‘help each other out’ in time — when wind is strong, solar is often weak, and vice versa. But how much this complementarity could actually alleviate absorption pressure has lacked a quantitative answer based on real geographic distribution.”
The Power of Spatial Coordination
The study’s most striking finding is that the effectiveness of wind-solar complementarity depends critically on geographic scale. At the county level, less than one-quarter of China’s counties can achieve effective wind-solar complementarity within their own borders. However, when coordination expands to the national level, almost any location can find a highly complementary wind or solar counterpart elsewhere in the country.
“If wind-solar matching is only done within county boundaries, less than a quarter of the country’s regions can form effective complementarity — a severe limitation,” said Assistant Prof. Zhang Fan of Peking University, as quoted by Guangming Daily. “But once the coordination scope is expanded, the effects quickly become apparent.”
Quantifying the Opportunity
The research team modeled different coordination scenarios and found that nationwide inter-provincial coordination could unlock 99.88 TWh (~100 billion kWh) of additional renewable energy absorption annually — equivalent to 9.1% of total solar and wind generation calculated in the study. This is enough to power the national average load for approximately 120 hours.
Crucially, this represents clean energy recovered from curtailment without any additional generation capacity. “This is not electricity generated out of thin air, but wind and solar power that would have had to be discarded, ‘picked back up’ through scientific dispatch,” Prof. Liu Yu explained. “Compared to simply piling up energy storage facilities, this approach can more effectively reduce system regulation pressure.”
At China’s average industrial electricity price of approximately 0.6-0.8 yuan per kWh, the recovered energy represents 60-80 billion yuan ($8-11 billion USD) in potential economic value annually. Environmentally, recovering 100 TWh of clean energy could displace significant coal-fired generation, reducing CO₂ emissions by an estimated 80-100 million tonnes per year.
Policy Implications and the Road Ahead
The research arrives at a pivotal moment for China’s energy policy. The “Power Mutual Aid Project” has been listed among 109 major projects in China’s 15th Five-Year Plan (2026-2030), underscoring the national priority on inter-regional power coordination. The study provides scientific backing for investments in ultra-high-voltage transmission lines and cross-provincial green electricity trading mechanisms.
Yu Chaohui, a researcher at Alibaba DAMO Academy, told QQ News that “with the help of AI models, we have built a new data foundation for academia and industry, which is expected to promote systematic research on wind-solar power planning, further助力 the construction of a new-type power system and accelerate the achievement of the ‘dual carbon’ goals.”
Limitations and Future Directions
The analysis assumes an 80% dispatchable-flexibility system, which may not reflect current grid capabilities. Realizing the full potential would require significant upgrades to grid infrastructure and market mechanisms. Additionally, the data reflects 2022 installations, and China’s rapid renewable expansion means current numbers are significantly higher. High-resolution geospatial data on individual facilities cannot be publicly released due to critical energy infrastructure sensitivity, limiting independent verification.
Nevertheless, the study elegantly demonstrates that the path to a high-renewable power system lies not just in building more capacity or storage, but in building a smarter, more interconnected spatial network that can match generation with demand across vast distances. As China pursues its “dual carbon” goals of peaking emissions by 2030 and achieving carbon neutrality by 2060, this AI-powered blueprint offers a data-driven roadmap for one of the most consequential energy transitions in human history.
The methodology developed by the Peking University and Alibaba team could also be applied to other countries seeking to optimize their renewable energy systems, potentially offering a template for global clean energy coordination.