Riverine Flood risk (Aqueduct 3.0)

Source: WRI Aqueduct 3.0 Year: 2019 Category: Physical risks · Mitigating services · Climate Acute Risks Coverage: Global Format: Boundary polygons (sub-national units) Used in risk analysis: No (contextual layer; not used in materiality scoring)

What it shows

This layer estimates the share of population expected to be affected by river flooding in an average year, after accounting for existing flood-protection standards. It captures the ongoing, expected burden of riverine flooding rather than a single worst-case event, making it a useful indicator of chronic flood exposure for sites and supply chains located near rivers and floodplains.

How it is built

The risk is assessed by combining three elements: hazard (inundation caused by rivers overflowing), exposure (the population located in flood zones) and vulnerability, with the prevailing level of flood protection also folded into the calculation. Crucially, the indicator represents average annual impact: the consequences of rare, extreme flood years are averaged together with more frequent, milder years to produce an "expected annual affected population". Values are reported per sub-national reporting unit and delivered as mapped polygons.

How to read it

Higher values mean a greater proportion of the local population is expected to be affected by river flooding in a typical year, indicating higher chronic flood risk. Lower values indicate limited expected impact. Because the figure is an annual average, it smooths out the difference between calm years and catastrophic ones; it is best read as a measure of persistent exposure rather than peak severity.

Class thresholds

The underlying quantity is a physical exposure rate: the expected annual share of the local population affected by riverine flooding (the Aqueduct rfr_raw field, a fraction of population per year, ranging from 0 to roughly 0.35, i.e. up to ~35%), after accounting for existing flood-protection standards. The layer is symbolised on the discretised rfr_cat field, which places each sub-national unit into one of five risk categories on that rate:

(A separate No Data value of −1 marks units without a computed estimate.)

This is a contextual layer and is not used in materiality scoring, so the categories serve to read chronic flood exposure rather than to drive a live risk flag. Higher classes mean a larger share of the population is expected to be affected by river flooding in a typical year.

How these thresholds were set. The expected-annual-affected-population rate and the five-class break structure follow the WRI Aqueduct flood-risk methodology (World Resources Institute); the population-fraction ranges shown above are taken directly from Aqueduct's published category definitions and are not a Darwin in-house classification.

Reference: WRI Aqueduct — wri.org/aqueduct.

Source

WRI Aqueduct (Aqueduct 4.0 flood-risk methodology), World Resources Institute.

Comparison with the WWF Risk Filter Suite

This layer maps to WWF Biodiversity Risk Filter indicator S3_7 Flooding and the WWF Water Risk Filter B3 Flood category. WWF expresses flooding as a basin-level composite of occurrence and hazard. Darwin instead provides an exposure-weighted, sub-national riverine flood-risk surface (expected annual affected population), so the two share the same hazard theme but differ in how risk is aggregated and expressed.

Legend

Symbolised field: Riverine Flood risk category

Generated from darwin/layers/layer-riverine-flood-risk-aqueduct-3-0.toml and risk_indicator_pairs.toml (develop).