What does a hundred-year flood mean?
A “hundred-year flood” means that a specified streamflow at a specified location has about a one-percent chance of being equaled or exceeded in any given year under the statistical assumptions used. It does not mean that such flooding can occur only once every century, that another event must wait one hundred years, or that an entire watershed will experience the same water level.
Probability, location, and measured flow
1. What is the one-percent probability? USGS explains a recurrence interval in terms of the chance that a given discharge threshold is equaled or exceeded within a year. A threshold with a one-percent annual exceedance probability is traditionally called the hundred-year flow. The probability applies afresh to each year; two exceedances can occur in consecutive years. It refers to a particular river reach, record period, and statistical method. It is not automatically the probability that a particular home will be flooded, because depth, location, drainage, defenses, and other pathways matter. The threshold should always be tied to a named site. 2. How is a threshold estimated? Analysts gather annual peak flows from a streamgage, examine record quality and changes in channel conditions, then estimate rare discharges with a suitable flood-frequency method. A short measurement record produces uncertainty about rare events. Historic flood marks and neighboring basins may add information. Dams, channel works, urban surfaces, and a changing climate can complicate the use of past peaks as a stationary guide to the future. Read the observation years, confidence interval, and model assumptions beside any single quoted discharge. The one-percent value at one gage is not the same flow at another. 3. Is a hundred-year rainfall the same event? Rainfall-frequency estimates consider a particular duration, intensity, and area, while a river-flood estimate concerns streamflow at a site. Antecedent soil moisture, the distribution of rain, snowmelt, pavement, and channel capacity determine how rain turns into runoff. A nominal hundred-year rain does not automatically make a hundred-year river peak. A small steep basin may respond within minutes while a large watershed peaks later. A very intense short shower and a modest but long storm can produce different consequences even when their totals seem similar. Check exactly what a map calls its event variable.
Maps, risk, and changing conditions
4. How is a flood boundary mapped? After selecting a discharge, analysts use land elevation, channel geometry, bridges, and a hydraulic model to estimate water depth and extent. The drawn line is the output of a scenario and model. A clogged culvert, local storm, new development, or changed embankment may alter real water levels. A property just outside a line is not guaranteed safe. Hazard describes the chance and intensity of water; risk also depends on people, buildings, infrastructure, and vulnerability. Read the map’s issuing agency, date, and revision before using it for a parcel-level decision. 5. Is the probability constant forever? Independent, stationary years make the definition easy to explain, but real river channels, land use, and climate can change. New observations revise estimates of rare flows. USGS notes that recurrence estimates can change as more data become available or flow patterns change. An old one-percent line does not permanently enclose all possible hazards. The same probability corresponds to different discharge and water depth in different places. Experiencing one rare flood does not create immunity from another; trend analysis and local engineering remain necessary. 6. How should the public use this term during a warning? The recurrence phrase is for long-term planning, not a forecast of exactly when floodwater will arrive. Current official weather and emergency bulletins, river stages, and local evacuation instructions are more urgent. Do not drive through floodwater or underestimate flow near bridges and culverts. General probability is insufficient for parcel safety, insurance, or building design; current local maps and qualified engineering review are needed. Stating the probability with its location and time basis avoids the false reassurance that a one-in-a-hundred event cannot happen again soon.