Why Do Some Cities Flood and Others Don't?

In September 2023, Storm Daniel dumped an estimated 750 mm of rain on parts of Thessaly in 48 hours — roughly a year's worth of rainfall in two days — turning the Thessalian plain into an inland sea, flooding the city of Volos, devastating agricultural land, and killing at least 17 people. Two months later, in November 2023, heavy rainfall caused severe flooding in the Athens suburb of Mandra for the second time in six years — a recurrence that raised the obvious and urgent question: why does this keep happening here and not elsewhere? The answer to why some cities flood and others don't is not simply "more rain" — it is a complex interaction of geography, geology, urban planning, infrastructure investment, and the long-term consequences of decisions made decades ago about how to build on a landscape that water was using first.

The Basics: How Urban Flooding Happens

To understand why cities flood, you first need to understand what happens to rain when it falls on a natural landscape versus an urban landscape. In a natural environment — forest, grassland, wetland — rainfall is managed through four processes: interception (rain caught by vegetation canopy), infiltration (water soaking into soil), evapotranspiration (water returned to the atmosphere through evaporation and plant transpiration), and slow runoff (water flowing overland gradually, filtered and slowed by vegetation and soil). In a healthy natural landscape, only 10-30% of rainfall becomes surface runoff — the rest is absorbed, stored, or returned to the atmosphere.

In an urban environment, the equation is reversed. Concrete, asphalt, rooftops, and paved surfaces are impervious — they do not allow water to infiltrate into the soil below. In a heavily urbanised area, 75-100% of rainfall becomes surface runoff — water that must be collected and transported by the storm drain system to rivers, the sea, or retention facilities. The volume of runoff from a given amount of rainfall is therefore 3-10 times higher in an urban area than in the natural landscape it replaced — and the speed of runoff is also dramatically higher, because paved surfaces are smoother than vegetated surfaces, and storm drains transport water faster than natural stream channels. The result is that urbanisation transforms a moderate rainfall event (manageable in a natural landscape) into a flood event (overwhelming the engineered drainage system), and a heavy rainfall event into a catastrophic flood.

Geography and Topography: The Cards the City Was Dealt

Some cities are geographically predisposed to flooding — built in locations where the natural water cycle concentrates flow, and where human development has intensified rather than mitigated that concentration. Basin cities (cities in topographic depressions surrounded by higher terrain) — like Volos (at the foot of Mount Pelion, receiving runoff from the entire Pagasetic catchment), or many Attic towns (in valleys between limestone mountains) — receive water from a much larger catchment area than their own footprint, making them vulnerable to floods whenever heavy rain falls on the surrounding hillsides.

Coastal cities on flat terrain face a different challenge: they have minimal elevation gradient to drive gravity-based drainage, and during storm surge events (wind-driven rises in sea level), the sea can actually push back against the drainage system, preventing stormwater from discharging. River cities — built along rivers that seemed perfectly manageable for 364 days a year but overwhelm their banks on the 365th — face the most common flood hazard worldwide: fluvial flooding, where upstream rainfall generates river flows that exceed the channel's capacity. The critical insight is that geography is not destiny — cities in flood-prone locations can (and do) manage their risk through engineering and planning — but geography determines the baseline risk that those engineering solutions must address. A city built in a basin at the confluence of two rivers (like many Greek cities) starts with a higher baseline flood risk than a city built on a hilltop — and must invest proportionally more in flood management to achieve the same level of protection.

Infrastructure: The Pipes Beneath the Streets

The storm drainage system — the network of gutters, inlets, pipes, channels, and outfalls that collects and transports rainwater from urban surfaces to receiving water bodies — is the primary engineered defence against urban flooding. The capacity of this system determines the rainfall intensity (mm/hr) that the city can handle without surface flooding — and in many cities, particularly those with older infrastructure, this capacity is woefully inadequate for the rainfall events that climate change is now producing.

Storm drainage systems are designed to handle rainfall events of a specified return period — typically the "10-year storm" (a rainfall intensity that has a 10% chance of occurring in any given year) for residential areas, and the "25-year" or "50-year storm" for critical infrastructure. The problem: many urban drainage systems in Mediterranean countries (including Greece) were designed in the 1960s-1980s using historical rainfall data from the mid-20th century — data that does not reflect the increased rainfall intensity that climate change has produced in subsequent decades. The result is that storms that would have been "50-year events" under historical climate patterns now occur with "10-year" or "20-year" frequency — overwhelming drainage systems that were adequate for the climate they were designed for but inadequate for the climate that now exists. Upgrading urban drainage systems is enormously expensive (it requires excavating and replacing pipes beneath existing streets and buildings) and politically difficult (the benefits are invisible — prevented floods — while the costs and disruption are immediately visible). This is why underinvestment in storm drainage is a near-universal feature of flood-prone cities: the cost of upgrading is visible today, while the cost of not upgrading is a hypothetical future event — until it isn't.

The Cities That Get It Right: Dutch and Japanese Models

Not all cities are equally vulnerable — and the cities that manage flooding most effectively share common strategies that flood-prone cities worldwide can learn from. The Netherlands — a country where 26% of the land area is below sea level and flooding is an existential threat — has developed the most comprehensive flood management system on Earth: a combination of massive engineered defences (dikes, storm surge barriers, pumping stations) and spatial planning policies (the "Room for the River" programme, which deliberately widens floodplains, creates water storage areas, and relocates development away from flood-prone zones). The Dutch approach is based on the principle that fighting water is futile — you must work with water, giving it space to go when it comes in volumes that exceed normal capacity.

Tokyo — a megacity of 14 million people built on a flat coastal plain at the confluence of multiple rivers — has invested over $2 billion in the Metropolitan Area Outer Underground Discharge Channel (G-Cans) — a system of five giant underground silos (each 65 metres deep and 32 metres in diameter) connected by 6.4 km of tunnels that can store and discharge 200 cubic metres of floodwater per second during typhoon events. The system — sometimes called an "underground cathedral" — represents the engineering extreme of flood protection, demonstrating that even the most flood-prone cities can be protected if the investment is sufficient. Copenhagen, after a catastrophic €800 million flood in 2011, developed a comprehensive cloudburst management plan that combines traditional grey infrastructure (enlarged storm drains, underground storage) with green infrastructure (rain gardens, permeable surfaces, urban wetlands, green roofs) to manage rainfall at the surface rather than channelling it all into pipes — a cheaper, more resilient, and more ecologically beneficial approach than pipe expansion alone.

Green Infrastructure: The Natural Solution

Green infrastructure — the use of vegetation, soil, and natural processes to manage stormwater — represents the most promising and cost-effective approach to reducing urban flood risk, particularly in cities that cannot afford the massive pipe expansion or underground storage systems of Tokyo or the Netherlands. The principle is simple: instead of collecting all stormwater and piping it to a central discharge point, manage it where it falls by creating surfaces and systems that absorb, store, and slowly release rainwater.

The toolkit includes: Permeable paving (paving materials that allow water to infiltrate through their surface into the soil below, reducing runoff by 50-80%), rain gardens (shallow planted depressions that collect and absorb runoff from roofs and paved areas), bioswales (vegetated channels that convey and filter stormwater while allowing infiltration), green roofs (vegetated roof surfaces that absorb and delay rainwater, reducing peak runoff by 50-90%), urban trees (a mature tree can intercept 15,000-20,000 litres of rainfall per year through canopy interception and root absorption), retention and detention basins (parks and open spaces designed to temporarily store floodwater during heavy rainfall events), and restored urban streams (daylighting streams that were previously channelled underground, restoring their natural capacity to store and convey floodwater). The economic case for green infrastructure is compelling: studies consistently show that green infrastructure costs 30-60% less than equivalent grey infrastructure expansion, while providing co-benefits (improved air quality, urban cooling, biodiversity habitat, recreational space, property value increases) that pipes and tunnels cannot provide.

Climate Change and the Future of Urban Flooding

Climate change is increasing flood risk in cities worldwide — not primarily through increased total annual rainfall (though this is occurring in some regions) but through increased rainfall intensity: the same amount of rain falling in shorter, more intense bursts that overwhelm drainage systems designed for gentler rainfall patterns. In the Mediterranean region specifically, climate projections indicate that while total annual rainfall may decrease (the region is expected to become drier on average), the intensity of individual rainfall events is expected to increase — meaning fewer but heavier storms, each of which poses greater flood risk than the more frequent but gentler storms of the historical climate.

This pattern is already visible in Greece: Storm Daniel (2023), the Mandra floods (2017 and 2023), and the repeated flooding of urban areas across Attica, Thessaly, and the Peloponnese in recent years follow a pattern of increasingly intense rainfall events producing flooding in areas that historical experience suggested were safe. The implication for urban planning is clear: cities can no longer design flood protection based on historical rainfall data — they must design for the future rainfall patterns that climate models project, which in most regions means designing for storms 20-40% more intense than the historical record suggests. Cities that plan for the past will flood in the future. Cities that plan for the future — investing in green infrastructure, oversized drainage, floodplain protection, and rainfall-responsive design — have the best chance of remaining functional as the climate that built them continues to change.