We got one part of the latest Monday Data edition wrong and I want to be clear about the correction.

We highlighted new research suggesting that data centres create a local “heat island” effect via their waste heat. A careful critique by Andy Masley shows that this conclusion goes too far. The underlying preprint uses satellite data to show that land surface temperatures around data centres rise after construction, but it does not convincingly separate the impact of waste heat from the more mundane effect of land‑use change – for example, replacing grassland with buildings and tarmac, which read as hotter from space regardless of what’s happening inside.

We have corrected the post: