人工智能(AI)等技术的蓬勃发展带动了中国数据中心的“爆发式增长”,同时也推高了能源消耗和碳排放。
截至2023年底,中国以449个数据中心的数量位居亚太地区之首。
国际能源署(IEA)最新报告显示,2024年中国数据中心用电量已占全球数据中心用电总量的25%,成为仅次于美国的全球第二大电力消耗国。
与各国情况类似,中国数据中心用电量预计将在未来几年持续快速增长,人工智能的兴起是重要推动因素之一。
不过,当前实际需求规模及未来增速仍存在不确定性。
现阶段,其他驱动因素对电力需求增长的影响仍远大于数据中心。
虽然各方对数据中心的预测数据存在差异,但有报告指出,其电力需求可能从2025年的100-200TWh(太瓦时)激增至2030年的600TWh,相应的CO2排放量或将达到200MtCO2e(百万吨二氧化碳当量)。
尽管中央和地方政府已出台多项政策以应对数据中心的环境影响,但挑战依然存在。
电力需求不断增长
中国国务院援引官媒《中国日报》2021年的一份报道称,2020年中国数据中心耗电量达200TWh,约占当年全国总用电量的2.7%,预计到2030年将增至400TWh(占比3.7%)。政府最新数据显示,2022年数据中心用电量为77TWh,2025年预计为150-200TWh,2030年或达400TWh。
2025年初,彭博社援引高盛(Goldman Sachs)更高预估称,中国数据中心的电力需求“预计将增长两倍多(从目前的200TWh),到2030年可能接近600TWh”。

相比之下,国际能源署(IEA)的预测则更为保守,其预计2024年中国数据中心用电量仅为100TWh,到2027年可能翻倍。
无论从占全国电力需求的比重,还是作为需求增长的驱动力来看,数据中心的规模仍然有限。
不同机构的数据显示,当前中国数据中心用电量约占全国总用电量的0.9%至2.7%。
彭博社指出,数据中心的用电量“不到制造业的十分之一”,并提到仅2024年一年,工业用电需求就增加了300TWh。
国际能源署表示,自2022年以来,数据中心仅占新增电力需求的3%,到2027年这一比例可能升至6%。该机构认为,中国电力需求增长的主要推动力来自工业领域,包括工业电气化及供热和交通电气化。
不过,国务院发展研究中心资源与环境政策研究所副主任韩雪表示,到2025年底,数据中心相关的CO2排放量预计将占全国总排放量的1%。
建设“绿色数据中心”
2021年,中国宣布了一项为期三年的行动计划,旨在建设“高效、清洁、集约、循环”的“新型数据中心”。
该行动计划包括提高数据中心PUE(电能利用效率)的措施。PUE是衡量数据中心能源效率最常用的指标。
其计算方式是将数据中心总能耗除以IT设备能耗。该比值越高,表明数据中心的能效越低。
截至行动计划结束,全国数据中心平均PUE已从上年的1.54降至1.48。
2024年提出的新目标是到2025年将大型数据中心的PUE控制在1.25以下。相比之下,拥有欧洲最多数据中心的德国要求现有数据中心从2027年起平均PUE需达到1.5。
与此同时,中国于2022年启动了备受期待的“东数西算”国家工程,旨在处理东部人口稠密省份产生的数据。该项目鼓励在西部太阳能和风能资源丰富的地区建设新数据中心,以支持东部繁忙的大都市。
根据该工程规划,中西部地区的数据中心将处理更多非实时云计算需求,如离线分析和存储备份,而对时效性要求高的数据服务仍由东部地区提供。
内蒙古等北方地区的地方政府也出台了配套政策,推动数据中心与可再生能源设施协同建设。
此外,北京地方政府已为数据中心提供资金支持,用于改善其PUE。而南方科技中心广东省则选择将部分数据中心建在海底,以减少冷却需求并降低能耗。
自2020年起,中国政府持续跟踪数据中心能源转型进展。2024年最新数据显示,全国已有50余个数据中心达到“绿色”能源标准,其中国家电网1个、互联网企业14个。
面临可再生能源挑战
到2030年,中国数据中心预计将消耗400TWh至600TWh的电力,相关排放量可能达到200MtCO2e。
当前,中国可再生能源资源主要集中在北方地区,而电力需求仍集中在东南沿海。这意味着,即便有“东数西算”工程的支持,数据中心通常也依赖于长距离输电来使用可再生能源。
“绿色电力在数据中心行业应用前景广阔,但仍面临诸多挑战。”绿色和平气候与能源资深项目主任吕歆说。
她向Carbon Brief指出:“完成跨省绿色电力交易仍然非常困难。”她解释道,这主要受限于可再生能源发电的不稳定性以及长距离输电线路的高昂运维成本。
中国已出台相关政策,支持绿电直供数据中心,并建设了配备专用可再生能源和储能设施的“绿色电力产业园区”。
“这些政策的推进和市场机制的完善将促进数据中心使用绿电。”吕歆补充道。
另一项挑战是数据中心的用水需求。由于需要大量冷却用水,数据中心可能加剧西部和北部地区本就紧张的水资源压力。
为应对这一问题,北京、宁夏和甘肃等地政府已出台强制性措施,要求提升数据中心用水效率,并逐步淘汰电力和水效率低下的数据中心。
随着数据中心规模不断扩大以满足人工智能运算需求,未来可能出现更多耗电量达数千兆瓦的”超大规模”数据中心,这将带来更大的电力供应压力。在国家整体电力结构中,采用更清洁的燃料组合有助于减少排放。
但研究机构SemiAnalysis指出,由于中国对煤炭的依赖,当前中国数据中心“在排放方面处于明显劣势”。
目前煤炭在中国能源结构中占比约60.5%。国际能源署数据显示,中国大部分数据中心所在的东部地区,约70%电力来自煤电。不过该机构预测,2030年后可再生能源与核能的快速发展将“推动煤炭的退出”。
该报告预计,到2035年,可再生能源和核能将“共同满足中国数据中心60%的电力供应”。
The post 解读:中国如何应对数据中心能源增长的需求 appeared first on Carbon Brief.
Climate Change
Night Skies and Shifting Stars: How Indigenous Celestial Knowledge Tracks a Changing Climate
When the land no longer answers the stars the way it once did, Indigenous peoples are among the first to notice — and the first to ask why.
A Sky Full of Knowledge
Look up on a clear night on Turtle Island and you’re seeing a sky that has guided human life for thousands of years. Across Indigenous nations in Canada, detailed systems of celestial knowledge developed not as abstract science but as living, practical guides —telling people when to plant, when to harvest, when herds would move, and when ice would come. This astronomical knowledge was woven into language, ceremony, and everyday life, passed down through generations with remarkable precision.
The Mi’kmaq and the Celestial Bear
Among the Mi’kmaq of Atlantic Canada, star stories are ecological calendars, precise and functional. The story of Muin and the Seven Bird Hunters connects the annual movement of what Western astronomy calls Ursa Major to the seasonal cycle of hunting and harvest: the bear rises in spring, is hunted through summer, and falls to earth in autumn. This knowledge was brought to broader public attention in 2009 during the International Year of Astronomy, when Mi’kmaq Elders Lillian Marshall of Potlotek First Nation and Murdena Marshall of Eskasoni First Nation shared the story through an animated film produced at Cape Breton University narrated in English, French, and Mi’kmaq.¹ The story encodes specific observations about when and where to hunt, and which species to expect at which time of year. It is science in narrative form.
The Anishinaabe and the Seasonal Star Map
Among the Anishinaabe peoples of the Great Lakes and northern Ontario, celestial knowledge forms part of a comprehensive seasonal understanding. Knowledge keepers like Michael Wassegijig Price of Wikwemikong First Nation have described how Anishinaabe constellations quite different from those of Western astronomy connect the movement of the heavens to naming ceremonies, seasonal gatherings, and land practices.² The Royal Astronomical Society of Canada now offers planispheres featuring Indigenous constellations from Cree, Ojibwe, and Dakota sky traditions, recognizing their value as both cultural heritage and ecological knowledge systems.³
When the Stars and the Land Fall Out of Rhythm
Here’s the challenge that climate change has introduced: the stars still move on their ancient, reliable schedule. But the land no longer always responds as expected. Migratory birds that once arrived when certain constellations appeared are now showing up earlier or later. Ice that once formed in predictable windows is forming weeks late, or not at all. Berry harvests, fish runs, animal migrations, all once timed by celestial cues accumulated over millennia are shifting. Indigenous knowledge holders across Canada describe this as a kind of dissonance: the sky remains faithful, but the land has changed.⁴
Long-Baseline Ecological Records
Far from being historical curiosity, Indigenous celestial knowledge systems are now being recognized by researchers as long-baseline ecological calendars —records of how nature behaved over centuries, encoded in story and ceremony. When an Elder observes that a particular star rising no longer predicts the arrival of certain geese, that observation represents a departure from a pattern that may have held true for hundreds of years. The Climate Atlas of Canada integrates Indigenous knowledge observations alongside western climate data, recognizing that both contribute meaningfully to understanding ecological change.⁵
Keeping the Knowledge Alive
Language revitalization and land-based education programs are helping ensure this knowledge reaches the future. From youth astronomy nights on-reserve to the integration of Indigenous sky stories in school curricula, there is growing recognition that these knowledge systems belong to what comes next, not only what came before. As Canada grapples with accelerating ecological change, the quiet precision of thousands of years of skyward observation offers something no satellite can fully replicate: a continuous record of the relationship between the cosmos and a living land.
Blog by Rye Karonhiowanen Barberstock
Image Credit: Dustin Bowdige, Unsplash
References
[1] Marshall, L., Marshall, M., Harris, P., & Bartlett, C. (2010). Muin and the Seven Bird Hunters: A Mi’kmaw Night Sky Story. Cape Breton University Press. See also: Integrative Science, CBU. (2009). Background on the Making of the Muin Video for IYA2009. http://www.integrativescience.ca/uploads/activities/BACKGROUND-making-video-Muin-Seven-Bird-Hunters-IYA-binder.pdf
[2] Price, M.W. (Various). Anishinaabe celestial knowledge. Wikwemikong First Nation. Referenced in: Royal Astronomical Society of Canada Indigenous Astronomy resources.
[3] Royal Astronomical Society of Canada. (2020). Indigenous Skies planisphere series. RASC. https://www.rasc.ca/indigenous-skies
[4] Neilson, H. (2022, December 11). The night sky over Mi’kmaki: A Q&A with astronomer Hilding Neilson. CBC News. https://www.cbc.ca/news/canada/newfoundland-labrador/hilding-neilson-indigenizing-astronomy-1.6679072
[5] Climate Atlas of Canada. (2024). Prairie Climate Centre, University of Winnipeg. https://climateatlas.ca/
The post Night Skies and Shifting Stars: How Indigenous Celestial Knowledge Tracks a Changing Climate appeared first on Indigenous Climate Hub.
https://indigenousclimatehub.ca/2026/04/night-skies-and-shifting-stars-how-indigenous-celestial-knowledge-tracks-a-changing-climate/
Climate Change
World ‘will not see significant return to coal’ in 2026 – despite Iran crisis
A much-discussed “return to coal” by some countries in the wake of the Iran war is likely to be far more limited than thought, amounting to a global rise of no more than 1.8% in coal power output this year.
The new analysis by thinktank Ember, shared exclusively with Carbon Brief, is a “worst-case” scenario and the reality could be even lower.
Separate data shows that, to date, there has been no “return to coal” in 2026.
While some countries, such as Japan, Pakistan and the Philippines, have responded to disrupted gas supplies with plans to increase their coal use, the new analysis shows that these actions will likely result in a “small rise” at most.
In fact, the decline of coal power in some countries and the potential for global electricity demand growth to slow down could mean coal generation continues falling this year.
Experts tell Carbon Brief that “the big story isn’t about a coal comeback” and any increase in coal use is “merely masking a longer-term structural decline”.
Instead, they say clean-energy projects are emerging as more appealing investments during the fossil-fuel driven energy crisis.
‘Return to coal’
The conflict following the US-Israeli attacks on Iran has disrupted global gas supplies, particularly after Iran blocked the strait of Hormuz, a key chokepoint in the Persian Gulf.
A fifth of the world’s liquified natural gas (LNG) is normally shipped through this region, mainly supplying Asian countries. The blockage in this supply route means there is now less gas available and the remaining supplies are more expensive.
(Note that while the strait usually carries a fifth of LNG trade, this amounts to a much smaller share of global gas supplies overall, with most gas being moved via pipelines.)
With gas supplies constrained and prices remaining well above pre-conflict levels, at least eight countries in Asia and Europe have announced plans to increase their coal-fired electricity generation, or to review or delay plans to phase out coal power.
These nations include Japan, South Korea, Bangladesh, the Philippines, Thailand, Pakistan, Germany and Italy. Many of these nations are major users of coal power.
Such announcements have triggered a wave of reporting by global media outlets and analysts about a “return to coal”. Some have lamented a trend that is “incompatible with climate imperatives”, while others have even framed this as a positive development that illustrates coal’s return “from the dead”.
This mirrors a trend seen after Russia’s invasion of Ukraine in 2022, which many commentators said would lead to a surge in European coal use, due to disrupted gas supplies from Russia.
In fact, despite a spike in 2022, EU coal use has returned to its “terminal decline” and reached a historic low in 2025.
Gas to coal
So far, the evidence suggests that there has been no return to coal in 2026.
Analysis by the Centre for Research on Energy and Clean Air found that, in March, coal power generation remained flat globally and a fall in gas-fired generation was “offset by large increases in solar and wind power, rather than coal”.
However, as some governments only announced their coal plans towards the end of March, these figures may not capture their impact.
To get a sense of what that impact could be, Ember assessed the impact of coal policy changes and market responses across 16 countries, plus the 27 member states of the EU, which together accounted for 95% of total coal power generation in 2025.
For each country, the analysis considers a maximum “worst-case” scenario for switching from gas to coal power in the face of high gas prices.
It also considers the potential for any out-of-service coal power plants to return and for there to be delays in previously expected closures as a result of the response to the energy crisis.
Ember concludes that these factors could increase coal use by 175 terawatt hours (TWh), or 1.8%, in 2026 compared to 2025.
(This increase is measured relative to what would have happened without the energy crisis and does not account for wider trends in electricity generation from coal, which could see demand decline overall. Last year, coal power dropped by 63TWh, or 0.6%.)
Roughly three-quarters of the global effect in the Ember analysis is from potential gas-to-coal switching in China and the EU.
Other notable increases could come from switching in India and Indonesia and – to a lesser extent – from coal-policy shifts in South Korea, Bangladesh and Pakistan.
However, widely reported policy changes by Japan, Thailand and the Philippines are estimated to have very little, if any, impact on coal-power generation in 2026. The table below briefly summarises the potential for and reasoning behind the estimated increases in coal generation in each country in 2026.
Dave Jones, chief analyst at Ember, stresses that the 1.8% figure is an upper estimate, telling Carbon Brief:
“This would only happen if gas prices remained very high for the rest of the year and if there were sufficient coal stocks at power plants. The real risk of higher coal burn in 2026 comes not from coal units returning…but rather from pockets of gas-to-coal switching by existing power plants, primarily in China and the EU.”
Moreover, Jones says there is a real chance that global coal power could continue falling over the course of this year, partly driven by the energy crisis. He explains:
“If the energy crisis starts to dent electricity demand growth, coal generation – as well as gas generation – might actually be lower than before the crisis.”
‘Structural decline’
Energy experts tell Carbon Brief that Ember’s analysis aligns with their own assessments of the state of coal power.
Coal already had lower operation costs than gas before the energy crisis. This means that coal power plants were already being run at high levels in coal-dependent Asian economies that also use imported LNG to generate electricity. As such, they have limited potential to cut their need for LNG by further increasing coal generation.
Christine Shearer, who manages the global coal plant tracker at Global Energy Monitor, tells Carbon Brief that, in the EU, there is a shrinking pool of countries where gas-to-coal switching is possible:
“In Europe, coal fleets are smaller, older and increasingly uneconomic, while wind, solar and storage are becoming more competitive and widespread.”
In the context of the energy crisis, Italy has announced plans to delay its coal phaseout from 2025 to 2038. This plan, dismissed by the ECCO thinktank as “ineffective and costly”, would have minimal impact given coal only provides around 1% of the country’s power.
Notably, experts say that there is no evidence of the kind of structural “return to coal” that would spark concerns about countries’ climate goals. There have been no new coal plants announced in recent weeks.
Suzie Marshall, a policy advisor working on the “coal-to-clean transition” at E3G, tells Carbon Brief:
“We’re seeing possible delayed retirements and higher utilisation [of existing coal plants], as understandable emergency measures to keep the lights on, but not investment in new coal projects…Any short-term increase in coal consumption that we may see in response to this ongoing energy crisis is merely masking a longer-term structural decline.”
With cost-competitive solar, wind and batteries given a boost over fossil fuels by the energy crisis, there have been numerous announcements about new renewable energy projects since the start of war, including from India, Japan and Indonesia.
Shearer says that, rather than a “sustained coal comeback” in 2026, the Iran war “strengthens the case for renewables”. She says:
“If anything, a second gas shock in less than five years strengthens the case for renewables as the more secure long-term path.”
Jones says that Ember expects “little change in overall fossil generation, but with a small rise in coal and a fall in gas” in 2026. He adds:
“This would maximise gas-to-coal switching globally outside of the US, leaving no possibility for further switching in future years. Therefore, the big story isn’t about a coal comeback. It’s about how the relative economics of renewables, compared to fossil fuels, have been given a superboost by the crisis.”
The post World ‘will not see significant return to coal’ in 2026 – despite Iran crisis appeared first on Carbon Brief.
World ‘will not see significant return to coal’ in 2026 – despite Iran crisis
Climate Change
Disaster Declarations Ripple Through South Texas Amid Water Crisis
Small towns around Corpus Christi worry where they’ll fall on the pecking order if the region’s water runs out.
At least six small cities and towns in the Coastal Bend region of Texas issued disaster declarations in the last two weeks, begging not to be forgotten amid a spiraling water crisis.
Disaster Declarations Ripple Through South Texas Amid Water Crisis
-
Climate Change9 months ago
Guest post: Why China is still building new coal – and when it might stop
-
Greenhouse Gases9 months ago
Guest post: Why China is still building new coal – and when it might stop
-
Greenhouse Gases2 years ago嘉宾来稿:满足中国增长的用电需求 光伏加储能“比新建煤电更实惠”
-
Climate Change2 years ago
Bill Discounting Climate Change in Florida’s Energy Policy Awaits DeSantis’ Approval
-
Climate Change2 years ago嘉宾来稿:满足中国增长的用电需求 光伏加储能“比新建煤电更实惠”
-
Renewable Energy6 months agoSending Progressive Philanthropist George Soros to Prison?
-
Climate Change Videos2 years ago
The toxic gas flares fuelling Nigeria’s climate change – BBC News
-
Carbon Footprint2 years agoUS SEC’s Climate Disclosure Rules Spur Renewed Interest in Carbon Credits








