One-Quarter of Global Data Centers Face Water Stress, Study Reveals Rising Climate Risks

Data Centers Face Growing Water Resource Limitations

A comprehensive global analysis published in Environmental Science & Technology has uncovered a pressing challenge: one-quarter of the world's data centers are located in areas where water demand exceeds supply for at least one month each year1

1

. The research, led by Lorenzo Rosa and colleagues, mapped 9,347 operational data centers worldwide against local water availability and current consumption patterns2

2

. Their findings reveal that approximately 6% of these facilities operate in regions experiencing water stress most of the year, raising concerns about the AI infrastructure environmental impact on communities and ecosystems1

1

.

The urgency stems from how these facilities function. Thousands of data centers globally provide the computing power and storage backbone for AI technologies, consuming large volumes of water to cool servers and prevent overheating2

2

. Unlike electricity that can be transmitted across long distances, water for these facilities must come from nearby rivers, lakes, and aquifers—the same sources that local communities, agriculture, and manufacturing depend upon1

1

.

Water Scarcity Measurement Reveals Critical Gaps

"The energy footprint of data centers has been quantified in detail because electricity is metered and reported," says Lorenzo Rosa, the study's lead author. "Water scarcity is measured locally, inconsistently, and often not at all. So, we built the map because it did not exist"2

2

. This lack of comprehensive tracking has obscured the true scale of water resource limitations facing the industry.

The researchers' analysis of data centers in water-stressed regions revealed competing demands. Cities and municipalities represent the largest water users near 41% of data centers, while agriculture serves as the primary consumer near another 28%1

1

. Rosa emphasizes the conflict: "These facilities are not landing in empty landscapes. They are landing on top of the water supply of cities and farms"2

2

.

Future Climate Scenarios Paint Troubling Picture

The research team projected forward using climate models that assume global temperatures averaging 5.4 degrees Fahrenheit (3 degrees Celsius) warmer than 1900 levels—a scenario that could materialize anywhere from 10 to 40 years from now1

1

. Under these future climate scenarios, the proportion of data centers facing water stress could jump to 33%2

2

. The United States faces particularly acute risks, as 39% of the global total of data centers are already located there, and most additional facilities projected to face water stress are American1

1

.

This escalation matters because local water availability fluctuates seasonally and annually, unlike the steady demand from data centers2

2

. As climate change intensifies droughts and alters precipitation patterns, the mismatch between supply and demand will likely worsen.

Environmental Impacts Demand Strategic Response

The findings highlight environmental impacts that extend beyond immediate water consumption. Data centers could intensify pressure on water resources in areas already experiencing—or projected to experience—water stress1

1

. Whether this scarcity translates into actual consequences for water users remains uncertain and requires further investigation. However, the research makes clear that long-term planning for data center operations must incorporate local water management and infrastructure needs2

2

.

Watch for industry responses to this research, particularly whether major technology companies will shift site selection criteria or invest in alternative cooling technologies. The study's comprehensive mapping provides a baseline that regulators and communities can use to assess proposals for new facilities. As AI technologies continue expanding, the tension between computational needs and water resources will likely intensify, making strategic location decisions and water-efficient technologies essential for sustainable growth.