The catastrophic floods that struck Texas last month, claiming over 135 lives and leaving nearly 170 missing, sent shockwaves far beyond the United States. Despite Texas boasting one of the most advanced flood forecasting and monitoring networks, the region was overwhelmed by a so-called “1-in-1,000-year” rain event – one of three such storms to hit the country within a single week. It’s a pattern echoed globally, and a wake-up call for civil engineers and policymakers alike.
Peter Gabriš is a hydrologist and flash floods – early warning leader at GoSpace Labs
2024 was officially the wettest year on record. More than 8,700 people died in floods worldwide; 40M were displaced; financial losses soared past £400bn. Europe was hard hit: Valencia, Spain, saw more than 750mm of rain in just eight hours, resulting in 229 fatalities and £8bn in damages. Central Europe faced unprecedented flooding, with storms breaking rainfall records and dam failures amplifying destruction in Austria, the Czech Republic, and Poland. The UK’s 2023/24 winter brought 12 named storms, widespread inundation, and crumbling infrastructure.
What is driving these unprecedented events? At the root is the thermodynamic fingerprint of a warming planet. A warmer atmosphere can hold – and ultimately release – more moisture, resulting in more frequent and intense downpours. According to the Clausius-Clapeyron relationship, each degree Celsius of warming enables the atmosphere to retain about 7% more water vapor. This amplifies the potential for extreme rainfall and, combined with evolving storm dynamics, creates compound hazards: consecutive storms, saturated soils and rapid intensification, all of which test the resilience of existing infrastructure.
Jozef Pecho is climate/numerical weather prediction model analyst at GoSpace Labs
Hydrological systems are under unprecedented stress. Urbanisation, degraded soils, and poor landscape management across Europe and the UK exacerbate runoff and erosion, stripping natural flood buffers from watersheds. The consequence: faster, more dangerous floods and a sharp increase in systemic vulnerability.
Traditional flood defences – levees, channels and barriers – were designed for the “n-year” storm events of the past. Today, these standards are regularly outpaced by real-world extremes. Fixed, centralised protections, once considered reliable, are now overwhelmed or rendered obsolete by rainfall intensities and durations outside historical experience. Systemic risk is no longer theoretical but a lived reality for millions.
For civil engineers, the lesson is clear: we cannot rely solely on historic technical norms or the protection offered by major flood works. Instead, flood resilience must be engineered into every project – through local runoff management, landscape adaptation and new approaches to design that anticipate tomorrow’s extremes, not just yesterday’s averages.
Perhaps most critically, the speed and hyper-locality of flash floods expose gaps in even the best national warning systems. National agencies, like the UK Met Office and Environment Agency, provide high-quality risk mapping and forecasting. Yet their updates, often issued every one to three hours, cannot match the pace of rapidly developing convective storms, where minutes matter and local impacts vary widely.
What is needed are rapid, hyperlocal warning systems – networks of autonomous rain and water level sensors, real-time data transmission resilient to severe conditions and smart evaluation software capable of issuing targeted alerts to authorities, infrastructure operators and at-risk populations. These systems must seamlessly integrate with high-resolution nowcasting models, radar and satellite data and, increasingly, AI-driven weather prediction tools that promise further gains in lead time and precision.
In Slovakia, Gospace Labs is piloting the Floodar platform: an advanced, hyperlocal early warning system that fuses real-time data from autonomous rain and water-level sensors, radar, high-resolution numerical weather prediction and INCA nowcasting system. By integrating these streams with smart, resilient communications, FLOODAR can deliver actionable, site-specific alerts directly to local authorities, infrastructure operators, and at-risk residents. The vision is modular, interoperable, and designed to work alongside national systems—not to replace them, but to bridge the critical gap in speed and resolution.
Future iterations will go further, incorporating AI-driven forecasting models like GraphCast to boost both precision and lead time. Yet already, projects like Floodar demonstrate how municipal, regional and operator-level preparedness can be dramatically improved through digital transformation. Civil engineers must advocate for — and help implement — these new layers of resilience, embedding warning networks and real-time risk data into the fabric of our built environment.
Future resilience depends on a hybrid approach: combining the forecasting power of national agencies with the agility of local, smart warning networks and flood-aware engineering. Only by integrating top-down data with bottom-up, site-specific intelligence can we close the critical gap between early warning and early action.
Climate-driven rainfall extremes are here to stay, challenging both our infrastructure and our imagination. The next generation of civil engineers must lead the way – not only designing robust physical defences, but also embedding real-time risk awareness and adaptive management into the heart of every project. In a world where yesterday’s “record flood” is tomorrow’s baseline, the imperative is clear: we must prepare for extremes that are no longer rare, but routine.
- Peter Gabriš is a hydrologist and flash floods – early warning leader at GoSpace Labs
- Jozef Pecho is climate/numerical weather prediction model analyst at GoSpace Labs
Like what you’ve read? To receive New Civil Engineer’s daily and weekly newsletters click here.
