How do you decide how high to raise a transport infrastructure, e.g., a road across a flood-prone floodplain?
In practice, engineers often rely on return periods. A road embankment, culvert, or bridge may be designed using a 10-year or 50-year flood estimate to reduce the likelihood of overtopping, erosion, and costly disruption over its design life. What was once considered a T50 event under a historical baseline, however, may occur far more frequently under near-, mid-, or late-century socio-economic pathways as rainfall extremes intensify.
That means a flood event expected once in 50 years today could behave more like a 20-year or even 10-year event in the future, raising important questions about whether our infrastructure is being designed for the climate we remember or the climate we are moving into.
As Francesco Serinaldi reminds us in Dismissing return periods!, the real challenge is not simply estimating return periods but understanding the probability of failure across the lifetime of infrastructure in a changing climate.
I would be interested to hear your view: should infrastructure design still rely on static return periods, or should climate-adjusted risk now become the standard?
Figure: Projected shifts in T10 and T50 return periods under baseline, near-century, mid-century, and late-century socio-economic pathways for Western Kenya.
Friday read: Serinaldi, F. — Dismissing return periods (Springer)
- #GeoFront
- #Hydrology
- #ClimateModelling
- #Infrastructure
- #FloodRisk