Author: Jack Graham Megan Rowling

Can AI 'digital twins' help protect us from climate disasters?

Source(s): Context

What’s the context?

As climate risks rise, the race is on to use better predictions from supercomputers and AI tools to keep people and assets safe.

  • Computer simulations offer a new tool to plan climate adaptation
  • Some say data quality in models is more important than visuals
  • Experts worry AI will leave hard-hit poorer communities behind


Analysts say advancements in AI and computing in projects like Destination Earth have the potential to improve the speed and accuracy of climate and weather models, including producing more detailed information to pinpoint local impacts.

"Downscaling basically tries to define a relation of this coarse-scale climate information to what really happens in the city, or in the valley, or on the mountains," said Martin Schultz, a senior researcher at Germany's Jülich Supercomputing Centre.


Others are more sceptical of the need for 'digital twins'.

"The name can be the lipstick on a pig," said Dan Travers, co-founder of Open Climate Fix, a non-profit that applies machine learning to reduce greenhouse gas emissions, such as by improving weather forecasts to make solar energy more predictable.

AI tools can augment climate models, especially for understanding weather, Travers noted - but he believes the quality and reliability of data, and how it is used, are more important than 'digital twin' visualisations that can be hard to build.


More than 90% of the world's largest companies will have at least one asset highly exposed to the physical impacts of climate change by the 2050s, according to ratings provider S&P Global.

London-based Cervest is another company that aims to help deal with that threat.

It combines a range of data, including publicly available scientific models, then uses machine learning to analyse climate-related physical risks facing assets like factories, hospitals and dams, from heat stress and flooding to high winds.


Please note: Content is displayed as last posted by a PreventionWeb community member or editor. The views expressed therein are not necessarily those of UNDRR, PreventionWeb, or its sponsors. See our terms of use

Is this page useful?

Yes No Report an issue on this page

Thank you. If you have 2 minutes, we would benefit from additional feedback (link opens in a new window).