TUHH > Research and Transfer > TUHH Societal Impact > Preventing Disaster with Models

Preventing Disaster with Models

Researcher: Prof. Dr. Simon Michael Papalexiou

What can be done when “once-in-a-century events” in weather occur far more frequently? Professor Simon Michael Papalexiou is establishing a new institute at TUHH that aims to help calculate the future more accurately—so that we are well prepared for extremes.

What begins as typical Hamburg drizzle escalates within minutes into a torrential downpour, with such massive rainfall that the water can no longer drain away. At Baumwall, water presses against the quay wall. In HafenCity, storm drains overflow and shop doors are sealed with towels. The air smells of wet concrete and a hint of diesel. A suburban train comes to an abrupt halt between two stations; the loudspeakers crackle, the lights go out.

It could happen. Such disaster-movie scenarios are part of the daily work of Professor Simon Michael Papalexiou. He explains: “Risks are often underestimated. That’s due to how our brains function. We orient ourselves toward what we see day after day—normal everyday life.” Yet policymakers, urban planners, emergency services and healthcare systems, as well as businesses and insurers, must be prepared for the real dangers. After positions in California and Canada, the Greek-German hydrologist Papalexiou is now establishing the Institute for Global Water Security at Hamburg University of Technology (TUHH). His goal: a significantly improved assessment of potential hydrological extremes—very rare, particularly severe rainfall, flood, or drought events. This is not about improving next Tuesday’s weather forecast, but about risk assessment: Where will water enter basements first? Subway tracks? Where is backflow likely? His methods are applicable worldwide and can be transferred to different regions and climatic conditions.

Stress Tests for the City of Tomorrow

Autos fahren über eine überflutete Straße — Flooding severely disrupts traffic.

The methods commonly used to date have evolved historically and insufficiently account for very large, rare values—such as in the opening Hamburg example. The city is prepared for storm surges and also for heavy rainfall, but not for both occurring simultaneously. Papalexiou cannot predict exactly when such a situation will arise. But he can unleash artificial storms on digital city models and use simulations to depict flooded basements, risks to underpasses, and the limits of sewer systems. “For emergency planning and the resilience of critical infrastructure, it is not enough to rely on the past,” Papalexiou emphasizes. “We must run through all plausible scenarios.”

And what does this mean for everyday life? “The calculations and stress tests we conduct here,” explains the risk researcher, “are relevant for decision-makers. They make it possible to plan the city of tomorrow correctly.” Bridges, dikes, and protective walls were long designed under the assumption that the climate would remain as it was. That is no longer the case. The climate is changing. Many structures were designed for a service life of 100 years but are now exposed to greater strain due to increasing weather-related stress. Moreover, commonly used global climate models, with grids of 100 by 100 kilometers, are far too coarse for practical use in policy and planning. Within such a grid, the entire city of Hamburg fits easily into a single square—far too imprecise for sound decision-making.

Why the Bridge Held

In addition, measurement data are incomplete and not of equal quality everywhere. Some rain gauges are spaced far apart. Some data series are short or interrupted. All of this must be taken into account by a robust risk model. One of the new institute’s first goals is therefore to scale down results to the local level (“downscaling”). This will make it possible to provide concrete assessments for individual streets or structures throughout Germany. The ground floor of a new residential tower might then be designed to withstand storm surges after all—or an insurer might adjust premiums according to the thickness of basement walls.

In a sense, Papalexiou operates in the B2B sphere—“business to business”—like wholesalers who sell only to retailers, not to private individuals. He works behind the scenes of everyday life. If he does his job well and subsequent authorities incorporate the data, private citizens will never know why the bridge held when their train was caught in a storm, or why grain supplies remained stable during a summer drought.

His findings emerge at the intersection of mathematics and hydrology—the science of water on and beneath the Earth’s surface. Hydrology examines, for example: How much rain falls, when, and where? Where does the water flow—to streams, rivers, sewers? What happens in soil, lakes, groundwater, snow, and ice? Why do floods, heavy rainfall, or droughts occur—and how often? How do cities, agriculture, dams, drainage systems, and soil sealing alter the water cycle? While meteorologists explain what comes from the sky, hydrologists study what the water does afterward.

The End of Illusory Certainty

Papalexiou aims to “make the unpredictable predictable.” Widely used hydrological models represent averages well but fail when it comes to extremes. They may show what a typical river level looks like but are poor at capturing exceptions—the days when everything tips over completely. Yet these are precisely the events that matter for risk assessment and early warning systems. Otherwise, a false sense of security arises—until it is too late. As in Germany’s Ahr Valley in 2021. Or in Derna, Libya, where in 2023 a dam failure killed 4,000 people and left 8,000 still missing.

More generally, Papalexiou advises: “We will have to come to terms with uncertainty. Uncertainty is increasing. If we accept that, we can prepare for it and handle it better.” His models do not seek to conceal uncertainty but to quantify it. He does not pretend that everything can be known precisely. Instead, he asks: How wide is the range? And which decision remains correct even if events turn out worse than we hope?

He is inspired by the possibility of designing better models using complex statistical systems: “I am delighted to see the complexity of the world reflected in mathematical formulas. In doing so, I make a small contribution to ensuring that people can live more safely in the future—that cities are better planned and structures endure longer, that fewer people are put at risk. I am proud of that.”

Despite Everything, Simon Papalexiou is an Optimist

Professor Papalexiou steht in der Hamburger Speicherstadt — Prof. Simon Papalexiou

After many years abroad, he enjoys being back in Europe. Cities here have a soul, he says; it is wonderful to stroll among historic buildings. The European university system also offers greater creative freedom. This allows the hydrologist to focus on what truly matters.

Although he deals daily with disaster scenarios, Papalexiou remains optimistic: “We have just moved here to Hamburg with our ten-year-old twins and are getting to know the city.” Perhaps, many years from now, his children will return to these places and think of their father. When the rain sets in, they might observe how the city responds: beneath HafenCity, retention basins open; pumps are controlled according to a plan based on thousands of calculated scenarios. Streets are designed to temporarily absorb and then release water. At an underpass, a warning light flashes: “Closed – risk of flooding!” Later, the news reports only: “Heavy rain front over Hamburg – traffic disruptions, no major damage.” All this would unfold because, based on Papalexiou’s scenarios, the necessary measures had been incorporated in time to address potential risks.

Then the rain continues for a while longer. And no one notices how close things might have come without these new ways of thinking.