Extreme precipitation

Extreme precipitation

Drainage is built for the rain a place usually gets. When a storm exceeds that capacity, water ponds, enters buildings, erodes and scours. The loss comes from rain beyond what the local drainage was designed for, not from rain as such: a wet climate with matching drainage reads no loss. Extreme precipitation prices both Property damage (% of asset value) and Business interruption (% of annual revenue).

What the map shows

• Today: the mean annual maximum 5-day rainfall (RX5day), 1995-2014, observed on a ~9 km grid (sheet). This band stands for the drainage capacity the networks were sized for.
• Projected: one band per scenario at 2035, 2050 and 2080, under the low, intermediate and high emissions settings. Each band is the observed baseline scaled by the IPCC Atlas CMIP6 change for that time slice and scenario.

How the loss is computed

The engine sets the site's drainage capacity at the baseline storm of the EN 752 drainage design standard. At each return period it compares the storm with that capacity; only the excess causes damage. A cubic curve on the excess, scaled by the site type's pluvial exposure tier, gives the share of asset value lost, and business interruption is a loss-of-use relation on the same ratio. The losses are integrated over return periods into an annual loss. Because the capacity is fixed at the baseline, intensification is what drives the figure. The hazard is tagged Proposal. See Damage functions: Extreme precipitation and Climate risks.

Sources

• ERA5-Land (Copernicus Climate Change Service / ECMWF), total precipitation, 0.1° global land grid, 1995-2014. CC BY 4.0.
• IPCC WGI Interactive Atlas CMIP6 RX5day change ratios per time slice and scenario; the SSP2-4.5 ratios come from the IPCC AR6 Atlas dataset served by the Copernicus CDS. CC BY 4.0.
• Damage curve: cubic excess over the local drainage design storm (EN 752); ceiling from JRC Huizinga (2017) at 0.5 m ponding. Shape and tiers are Darwin's.

Limits

• The curve shape, the exposure tiers and the rainfall distribution's shape parameter are Darwin choices.
• The change signal comes from a ~115 km ratio, so it stays coarse even though the baseline is observed at ~9 km.
• ERA5-Land is land only, and 9 km still smooths convective storms. The layer describes a landscape, not a rooftop.
• The loss grade is withheld on both channels, because the figure is not considered plausible enough yet.