Data Centers Heat up Downwind Neighborhoods

Research aims to reduce impact of heat pollution on downwind neighborhoods

Peer-Reviewed Publication

Arizona State University

News Release 18-May-2026

Waste heat from data centers can boost air temperatures in downwind neighborhoods by as much as 4 degrees Fahrenheit, researchers at Arizona State University report in a new study conducted in the Phoenix metro area, the hottest in the U.S.

“As we do more measurements under different kinds of atmospheric conditions, I think we're going to see more significant impacts around data centers,” said lead author David Sailor.  

With hundreds of megawatts of data center capacity operating in many cities, and thousands more proposed, the combined impact on urban temperature could be substantial. U.S. data center capacity is projected to more than double by 2030. Sailor and co-authors said the overlooked heat hazard demands attention from city planners and industry developers. The researchers aim to help develop solutions that could significantly reduce downwind impacts.

The waste heat produced by a single data center can surpass the amount emitted by 40,000 households, according to Sailor. Air-cooled condenser arrays discharge air heated to 14 to 25 degrees F above the surrounding air temperature, creating thermal plumes that move downwind over neighboring areas.

“They're such a concentrated load of electricity consumption and hence heat emissions that we became concerned about the impact that they could have locally, and also in the downwind neighborhoods,” said Sailor, a professor at Arizona State University and director of ASU’s School of Geographical Sciences and Urban Planning.

Other researchers have tried to use remote sensing data from satellites to estimate the heat impact of data centers historically. The ASU study is the first to directly measure air temperatures downwind and upwind of data centers to record the real-time effects of waste heat on surrounding communities. Sailor and co-authors Soroush Samareh Abolhassani and Eli Martin are publishing their findings in the Journal of Engineering for Sustainable Buildings and Cities.

The researchers mounted data-logging high-accuracy and fast-response temperature sensors on cars that drove around Phoenix-area data centers and throughout nearby neighborhoods from June 18 to October 25, 2025. Using multiple cars allowed them to simultaneously measure temperatures upwind and the downwind of the four selected facilities ranging from a 36-megawatt single-building data center in Mesa to a 169-megawatt colocation campus in Chandler. The chosen centers reflect the typical design of “hyperscalers” that house many thousands of servers and use primarily air-based cooling systems.

Temperatures downwind of data centers averaged 1.3 to 1.6 degrees F warmer than upwind temperatures and reached as high as 4 degrees F above upwind temperatures. The heat impact was detectable up to a third of a mile, or about five city blocks, distant from the perimeter of datacenters.

“Even if these data centers only contribute to an additional heat island magnitude of one degree or two degrees, that can still have a very significant impact on our lives,” Sailor said. That’s especially true in places where extreme heat already poses serious public health risks.

A one-degree boost in air temperature, for example, is enough to drive higher use of air conditioning across entire neighborhoods. Those air conditioners, in turn, put even more heat into the surroundings.

Sailor and colleagues are planning a more extensive effort collect data over a wider range of times and weather conditions. That data will allow them to develop an accurate atmospheric model to study the effects of measures to lessen the heat impact on downwind neighborhoods.

“Data centers are inherently an important part of our society, and they're going to become even more necessary going forward,” Sailor said. Rather than just highlight adverse consequences, his goal is to collaborate with data center providers and other stakeholders to develop the knowledge needed to reduce the heat pollution problem.

Design modifications to facilities and cooling equipment informed by high-resolution microclimate modeling, for example, could lower the thermal footprint of a data center without compromising data center operations. Greenbelts or parks could buffer heat pollution. Cities could require such fixes in siting and permitting of data centers.