Table 1 An overview of the four lines of evidence to identify unprecedented weather hazards discussed here, and their key benefits and limitations

From: How to stop being surprised by unprecedented weather

Lines of evidence

Benefit

Limitation

Conventional statistical methods using observations

Represent events that actually occurred

Are inherently limited in identifying unseen events

Past events from historical observations, documentary evidence, oral history and proxy data

Help understand events that may have happened before the modern observational records

Are rare, may be challenging to compare to present conditions, and (for the proxy data) have limited resolution

Event-based storylines

Provide a physically plausible unfolding of a single event

Only applies to the single event

Weather and climate-model data exploration

Produce information outside what is possible from observations

Rely on the realism of these models