Why climate scientists need to talk more about the very worst‑case scenarios

London is under water. The Houses of Parliament, Buckingham Palace, St Paul’s Cathedral and the Bank of England are all submerged. Far away, the West Antarctic and Greenland ice sheets have collapsed, triggering accelerated sea level rise which, combined with a storm surge in the North Sea and a high spring tide, has led to water flooding over the Thames Barrier. Thousands of shops, offices, schools and houses are several feet deep in effluent-rich water.

Economists struggle to calculate the ongoing economic damage, but everyone knows it is epochal. The realisation dawns that the UK’s capital city is no longer defendable. The government has no choice but to abandon it permanently to the waves.

This dystopian vision of the next century is not the most likely outcome. But nor do current scientific projections rule it out.

As scientists learn more about how climate change could trigger catastrophic flooding, wholesale collapse of food supplies and millions of heat-related deaths, there is a growing realisation that such risks need to be more widely understood and acted on. Too many people remain unaware about what could actually happen. Disasters like these are possible. Crucially, they are not yet inevitable.

This lack of awareness is concerning, as I outlined with colleagues in an article published in February 2026 in the journal Nature. It fuels both climate denial and climate doomism. We have the evidence needed to help steer humanity towards greater safety, but scientists have not communicated that evidence effectively enough. We have tied ourselves too closely to a paradigm that focuses on best-known outcomes.