AI data centres raise local land temperatures by up to 9.1°C

Data centres powering artificial intelligence are causing land surface temperatures to rise by an average of 2°C near their locations, with extremes reaching 9.1°C, according to researchers at the University of Cambridge. The heat effects extend up to 10 kilometres away, potentially impacting more than 340 million people worldwide. The findings come from a study analysing satellite data over the past two decades.

Researchers led by Andrea Marinoni at the University of Cambridge examined satellite measurements of land surface temperatures from the past 20 years, cross-referencing them with the locations of more than 8,400 AI data centres. To isolate the centres' impact, they focused on facilities situated away from densely populated areas, accounting for other potential influences on temperature readings. Their analysis revealed an average temperature increase of 2°C (3.6°F) in the months following a data centre's operational start, with the most extreme cases showing a rise of 9.1°C (16.4°F). The warming persisted up to 10 kilometres from the sites, diminishing by only 30 per cent at seven kilometres. Marinoni described the results as 'quite surprising' and warned, 'This could become a huge problem.' Real estate firm JLL forecasts that global data centre capacity will double between 2025 and 2030, with AI driving half of that growth amid surging energy demands. Using population data, the team estimated that over 340 million people reside within 10 kilometres of such facilities, experiencing warmer conditions than they otherwise would. Specific regions like Mexico's Bajío and Spain's Aragon province registered unexplained 2°C increases between 2004 and 2024. Chris Preist at the University of Bristol noted the findings' nuance, calling for further research to distinguish heat from computation versus the structures themselves absorbing sunlight. Marinoni maintained that data centres elevate ground temperatures regardless and urged caution in their design and development. The study appears on arXiv under DOI arXiv:2603.20897.

Articoli correlati

President Trump shakes hands with tech CEOs signing the Ratepayer Protection Pledge at the White House, with AI data centers symbolized in the background.
Immagine generata dall'IA

Tech giants sign White House pledge to cover AI data center power costs amid backlash

Riportato dall'IA Immagine generata dall'IA

On March 4, 2026, leading tech firms including Amazon, Google, Meta, Microsoft, OpenAI, Oracle, and xAI signed the non-binding Ratepayer Protection Pledge at the White House, committing to fund new power generation and infrastructure for AI data centers to shield consumers from rising electricity bills. President Trump hailed it as a 'historic win,' but critics question its enforceability amid growing environmental and economic concerns.

A Wired article explores the idea of launching AI data centers into orbit to mitigate their environmental impact. It highlights the rapid growth of these facilities amid the AI boom and their massive energy consumption. The proposal aims to address rising electricity demands and associated global warming.

Riportato dall'IA

Nearly half of planned US data centers for this year risk delays or cancellation due to import issues from China, exacerbated by tariffs. Community opposition is fueling moratoriums, with Maine poised to halt new construction until 2027. These hurdles challenge President Trump's push for rapid AI infrastructure buildout.

Building on recent studies like Stefan Rahmstorf et al.'s analysis showing a doubling of Earth's warming rate to ~0.36°C per decade since 2014, scientists disagree on whether reductions in aerosol pollution or natural fluctuations are driving the speedup. Nearly all agree warming has accelerated, but views differ on causes, rate, and future trajectory—with implications for climate sensitivity and adaptation.

Riportato dall'IA

Residents in rural areas across the United States are opposing new data center projects over concerns about water supplies, energy costs, and farmland loss. In Tazewell County, Illinois, locals successfully blocked a proposed facility after packed council meetings and petitions. The backlash highlights tensions between AI infrastructure growth and agricultural needs.

Hut 8 announced a 15-year lease valued at $9.8 billion for a large-scale AI data center in Texas, sending its shares up nearly 30 percent on Wednesday. The deal covers 352 megawatts of capacity at the company's Beacon Point campus and includes options that could raise the total value to $25.1 billion.

Riportato dall'IA

A massive data center proposed for Imperial Valley, California, would require 750,000 gallons of water daily, raising fears among local residents about rising costs and shortages. Developer Sebastian Rucci plans to complete the 330-megawatt facility by 2028, but it faces lawsuits and regulatory hurdles. The project is one of 24 data centers set to open in the state by 2030.

 

 

 

Questo sito web utilizza i cookie

Utilizziamo i cookie per l'analisi per migliorare il nostro sito. Leggi la nostra politica sulla privacy per ulteriori informazioni.
Rifiuta