How AI clouds the future of WA’s rivers
A study by University of Washington scientists, prompted by the Columbia River Inter-Tribal Fish Commission, found that winter temperatures in the Columbia River have been gradually increasing near AI data centers in Washington. The research detected localized warming of up to 2.5°C, suggesting a need for increased transparency in cooling water management and expanded ecological monitoring. The findings call for sustainable AI governance that includes water and ecosystem impacts.
The Seattle Times reports on a study by University of Washington scientists, initiated by a question from Elaine Harvey of the Columbia River Inter-Tribal Fish Commission, investigating whether data centers contribute to river warming. Researchers analyzed 40 years of satellite-derived river temperature data in conjunction with 37 AI data center locations and operational timelines in Washington state.
The analysis revealed localized river warming signals, particularly in the Columbia River during December, near clusters of AI data centers. Scientists detected an average warming of up to 2.5°C following the establishment of these facilities, with the warming signal dissipating within six kilometers downstream unless other data centers were present.
These findings highlight the urgent need to understand the environmental and ecological impacts of resource-intensive AI data centers, especially given Washington's role as a major hub for such facilities due to its hydroelectric power access. The research team advocates for four key actions: making river temperature a standard monitoring requirement for data center infrastructure, improving transparency in cooling-water withdrawals and discharge, extending sustainable AI governance frameworks beyond carbon accounting to include water and ecosystem impacts, and viewing ecological monitoring as a prerequisite for responsible AI development. The goal is to ensure AI infrastructure evolves with environmental safeguards.