Artificial intelligence (AI) infrastructure is reshaping urban thermal environments, yet concentrated digital facilities remain largely absent from urban climate analytical frameworks. This study introduces the Digital Heat Island (DHI) — a localized thermal anomaly associated with continuously operating digital infrastructure — and presents an exploratory spatial analysis of 46 data center buildings across Phoenix, Arizona. Using ECOSTRESS land surface temperature (LST) imagery, Landsat-derived NDVI, and American Community Survey sociodemographic data, we examine thermal, vegetation, and demographic patterns within 0-1, 1-2, and 2-3km buffer zones around data centers. Results reveal modestly elevated surface temperature (39.74°C within 0-1 km versus 39.26°C citywide), vegetation deficits (NDVI 0.15 within 0-1km versus 0.17 citywide), and disproportionate proximity to low-income and minority populations. While the analysis is exploratory and does not establish causality, the consistency of these patterns across buffer zones motivates a new interdisciplinary research agenda integrating urban climate science, digital geography, and environmental justice.
Alamin Molla (Mon,) studied this question.