New AI tool monitors Arctic permafrost thaw by spotting disappearing lakes
Researchers have launched an interactive website that uses AI to track permafrost thaw across the Arctic in near real time. The tool has already identified 30 additional drained lakes in northwestern Alaska since June 2026, using satellite imagery to detect sudden lake disappearances that signal thawing ice-rich ground. This monitoring could help communities prepare for land subsidence and property damage caused by permafrost melt.
The tool leverages decades of aerial and satellite imagery, allowing AI to compare historical lake positions with current scans. By flagging sudden drainage events, it distinguishes between gradual landscape evolution and abrupt collapses tied to ice-rich ground. This near-real-time capability is especially valuable in remote regions where ground surveys are impractical, offering a cost-effective early warning system for infrastructure planners.
Permafrost thaw not only destabilizes buildings and roads but also reshapes entire ecosystems, as drained lake beds become new terrain for vegetation and wildlife. The monitoring effort, part of the Permafrost Discovery Gateway, already tracks millions of lakes and ponds, providing a baseline for detecting future changes. Communities in Alaska and beyond can use these maps to assess land stability before construction, reducing the risk of costly subsidence damage.
This tool could significantly aid Arctic communities facing escalating property risks from thawing ground, potentially informing zoning decisions and infrastructure maintenance. By making permafrost dynamics visible in near real time, it may also help insurers and local governments anticipate liability and adaptation costs. However, its effectiveness depends on sustained satellite coverage and community access, and it cannot prevent the underlying warming driving these changes.