From extreme to extreme: Scientists unlock clues about major weather events (2025)

The storm came as France's poorest territory was still reeling from a deadly cyclone from mid-December - Copyright AFP Niharika KULKARNI

With regularity, around the world, gaps in disaster preparedness systems and adaptation actions have left communities underprepared and unequipped to handle “surprising” weather events. To move this forward, a new model intended to help communities to predict extreme weather events, by assessing past events that have never been recorded in modern history, is the focus of a new study. This involved seeking to move beyond the limitations of conventional weather records, which typically only cover the last century.

University of Reading (UK) scientists have revealed how nature’s own archives — like tree rings — combined with forgotten historical documents can help to unlock centuries of climate data that modern instruments have missed.

Lead author Timo Kelder states: “We’ve been limited by thinking extreme weather is only as bad as what we’ve measured since weather stations were invented. But our research shows we can use weather models to look back hundreds or even thousands of years to discover what’s truly possible in our climate system.”

From the assessment, the researchers identified four approaches that together create a more complete picture of possible extreme weather:

1. Analysing conventional records.
2. Studying historical and natural archives like tree rings (together with reconstructing information from ice cores, coral reefs, speleothems, pollen, ocean sediments, and diatoms).
3. Creating “what-if” scenarios based on past events.
4. Using climate models to simulate physically possible extremes.

Of the four aspects listed, tree rings proved especially valuable. The researchers charted how each ring preserves a year’s worth of climate history.

Researchers next deployed these natural time capsules to reconstruct 850 years of drought patterns in northwestern China. This model revealed several extreme events that would have been invisible in modern records.

There was an element of verification of the model by pinpointing some forgotten weather extremes from historical archives. This showed, for example, that June 1846 in Durham, UK was significantly hotter than any modern June temperature. Similarly, September 1774 in Oxford was wetter than any month recorded in the 250 years since.

These historical archives cross-matched the findings from the new analytical approach, demonstrating the accuracy of the new nature-based assessment approach.

The model also aids the prediction of future weather events, providing improved early warning systems. Consequently, this would allow for upgraded infrastructure as well as facilitating transformative social changes to reduce vulnerabilities. The researchers recommend the adoption of a so-termed ‘adaptation pyramid’. This consists of reactive (short-term disaster preparedness and disaster response), incremental (long-term disaster prevention), and transformative adaptation strategies.

The aim is to create a future state whereby societies can take “responsibility to build resilience rather than being surprised by unprecedented weather.” In other words, shifting from the reactive to the proactive.

The research appears in the journal Nature Communications, titled “How to stop being surprised by unprecedented weather.”