What weighs on the result, and why
The index combines five factors that flood hydrology treats as dominant, but they do not weigh equally. The first two are geometric and do not change: once the house is built, the distance to the watercourse and the height above it are fixed, and they are what most determines the result. The other three vary with the episode.
Height above the channel. This is the single most decisive factor. The water surface in an ordinary flood rarely exceeds two or three metres above the channel in an open valley, and every metre of elevation buys margin non-linearly. A basement or a ground floor below street level falls into the high-risk category almost regardless of everything else.
Distance to the watercourse. Spanish law reserves the first five metres from the bank as an easement strip and the first hundred as a policing zone, where any works need authorisation from the river basin authority. It is not a safety line, it is a warning: a river’s floodplain can be hundreds of metres wide.
Expected rainfall. What produces a flood is not the total of the episode but the intensity: a hundred litres over twenty-four hours largely infiltrates, and the same hundred litres in two hours go almost entirely to the channel. In small Mediterranean catchments the response time is measured in minutes, which is why a DANA produces flash floods.
Soil type and urbanisation. Ground sealed by asphalt turns between 70 and 95 % of what falls into runoff, against 10 or 20 % on forested terrain. Urbanising a catchment without drainage works multiplies the peak flow with no extra rain at all.
The return period, which is almost always misunderstood
Saying that an area floods with a return period of 100 years does not mean it happens once a century. It means each year carries a 1 % probability of it happening, and that probability is not used up: if it happened last year, this year is still 1 %. Over a thirty-year mortgage, the cumulative probability of at least one flood with a 100-year return period is 26 %, which is a rather different figure from the intuition of «once a century».
The 30-year column is the probability of at least one episode, computed as 1 − (1 − p)³⁰.| Return period | Annual probability | Over 30 years | Typical use in Spain |
|---|
| 10 years | 10 % | 96 % | Urban drainage, gullies |
| 50 years | 2 % | 45 % | Minor crossing structures |
| 100 years | 1 % | 26 % | Preferential flow zone, SNCZI |
| 500 years | 0.2 % | 6 % | Flood zone in official mapping |
What this index cannot know
It is indicative by construction, and three things lie beyond its reach. First, the real capacity of the drainage: a well-sized sewer completely changes the outcome on a particular street. Second, the operation of reservoirs upstream, which can attenuate a peak or, if they are already full, fail to. Third, the prior saturation of the soil: the same rain on ground soaked by the preceding weeks produces far more runoff, and that is exactly the mechanism that made storm Domingos worse three days after Ciarán.
Where the official information is
Spain’s flood-zone mapping is published in the National Flood Zone Mapping System (SNCZI) of the Ministry for the Ecological Transition, with inundation extents for return periods of 10, 100 and 500 years searchable by address. Real-time river levels are published by the SAIH systems of each river basin authority. Warnings are issued by AEMET and evacuation orders by Civil Protection. During an episode the number is 112, and the rule that saves most lives is the simplest: never cross a flooded stretch, on foot or by car. Thirty centimetres of moving water is enough to sweep away a car.
