Why predicting a flash flood is harder than predicting a storm
In short: Heavy rain doesn't always equal a disaster. The difference between a manageable downpour and a catastrophic flash flood often lies in the hidden geography of the land—how porous the soil is, how much it was scorched by wildfire, or even how many railroad-like storm echoes are moving over the same area.
Traditional flood management is struggling to keep pace with rising global water levels. Water resource engineer Virginia Smith argues that our current approaches are insufficient for the modern era of flooding, necessitating a dynamic shift in how we manage and stymie the flow of storm waters to ensure a sustainable future.
The complexity of flash floods stems from their non-meteorological nature; they occur when specific hydrologic conditions meet favorable weather. Hydrologists must account for a dizzying array of variables: terrain steepness, vegetation types, paved surfaces, and even the depth and condition of snowpacks. A critical factor is 'training echoes,' where lines of storms move repeatedly over the same area like railroad cars, compounding the water load.
A major hurdle in flood mitigation is the lack of data in developing countries, which often lack dense streamflow gauge networks. Standard hydrological models usually require long historical records to calibrate accurately for specific watersheds. However, new research shows that artificial intelligence can bridge this gap. AI-based forecasting has demonstrated the ability to predict extreme riverine events in ungauged watersheds with a five-day lead time—matching or exceeding the reliability of current state-of-the-art global systems that only provide 'nowcasts' (zero-day lead times).
These advancements are already being put into practice. An AI model was recently incorporated into an operational early warning system providing free, real-time forecasts in over 80 countries. By achieving accuracies on five-year return period events comparable to current one-year event models, this technology allows for earlier warnings of larger, more impactful disasters.
Opens the Intellect+ feed with this card first. Swipe up for the next one.
Sources
- Summary by Intellect · source: Virginia Smith: How we could better predict and stop floods
- Also: Severe Weather 101: Flood Forecasting
- Also: Global prediction of extreme floods in ungauged watersheds | Nature
To cite this card, link to https://intellectplus.co.uk/app/ideas-how-we-could-better-predict-and-stop-floods.
How we source, check and correct cards: Editorial standards.
General information, not professional advice. Spotted an error? Flag the card or see corrections and complaints.
Intellect Plus is a visual educational discovery platform for exploring fascinating facts and ideas. Brain Food, Not Brain Rot. What is Intellect Plus?
Related
More in Artificial intelligence · More in Ideas & Philosophy · Latest · A–Z · All topics
