Climate Hazard Diagnostics - Fluvial Floods

Emmi — version 1.3 (2026-03-27)

Description

This dataset quantifies fluvial flood hazard intensity, 100-year return period inundation depth (m), exceedance probability, and asset average annual loss at 0.0083° (~1 km) spatial resolution. The dataset has global river basin coverage for a historical baseline (1980) and the years 2030, 2050, and 2080 under IPCC climate scenarios RCP4.5 and RCP8.5.

Dataset ID
9e2f989c
Category
Risk
Type
Raster
CRS
EPSG:4326
Spatial coverage
Global
Spatial resolution
1 km
Temporal coverage
1980-2080
Temporal resolution
Irregular

Usage notes

Spatial coverage is limited to global river basins and does not include local flood defences or drainage infrastructure. The historical baseline represents a modelled pre-acceleration climate state, not actual conditions in 1980. Future years should be interpreted relative to the baseline, not in isolation as absolute hazard levels. Average annual loss is calibrated to present-day conditions (~2025), enabling assessments of current risk to assets alongside change over time. Meaningful differences among climate scenarios typically occur from 2050 onwards.

Variables

average_annual_loss_baseline

NoDataNaN

Description

Expected annual damage as a fraction of asset value for the historical baseline (1980). Data is the product of flood intensity, flood probability, and a damage calibration factor derived from 1990-2024 disaster losses across 27 countries.

Usage notes

Values range from 0 to 1, representing mean annual proportional loss for all building types and materials. Absolute asset-level losses can be calculated by multiplying this variable by an asset's value. Values < 0.001 typically indicate low physical risk exposure and values > 0.05 typically indicate high physical risk exposure. The damage calibration factor is anchored to present-day conditions (~2025), making data suitable for current-day risk screening alongside future scenario comparison. Average annual loss is intended for portfolio-scale analysis rather than precise asset-level loss estimation. This variable is only available for the historical baseline year (1980), with all other years containing null values.

average_annual_loss_rcp4p5

NoDataNaN

Description

Expected annual damage as a fraction of asset value for 2030, 2050, and 2080 under IPCC climate scenario RCP4.5. See average annual loss baseline variable for methodology.

Usage notes

Values range from 0 to 1. This variable is available for years 2030, 2050, and 2080. The historical baseline year (1980) contains null values.

average_annual_loss_rcp8p5

NoDataNaN

Description

Expected annual damage as a fraction of asset value for 2030, 2050, and 2080 under IPCC climate scenario RCP8.5. See average annual loss baseline variable for methodology.

Usage notes

Values range from 0 to 1. This variable is available for years 2030, 2050, and 2080. The historical baseline year (1980) contains null values.

depth_meters_baseline

NoDataNaNUnitsm

Description

Expected flood depth of a 100-year (1% probability) fluvial flood event for the historical baseline (1980). Data is derived from the GLOFRIS framework coupling the PCR-GLOBWB 2 hydrological model with CaMa-Flood inundation routing, as provided by the WRI Aqueduct Floods dataset.

Usage notes

Depth values can be used as inputs to custom depth-damage curves. This variable is only available for the historical baseline year (1980), with all other years containing null values.

depth_meters_rcp4p5

NoDataNaNUnitsm

Description

Expected flood depth of a 100-year (1% probability) fluvial flood event for 2030, 2050, and 2080 under IPCC climate scenario RCP4.5. See depth meters baseline variable for methodology.

Usage notes

This variable is available for years 2030, 2050, and 2080. The historical baseline year (1980) contains null values.

depth_meters_rcp8p5

NoDataNaNUnitsm

Description

Expected flood depth of a 100-year (1% probability) fluvial flood event for 2030, 2050, and 2080 under IPCC climate scenario RCP8.5. See depth meters baseline variable for methodology.

Usage notes

This variable is available for years 2030, 2050, and 2080. The historical baseline year (1980) contains null values.

intensity_baseline

NoDataNaN

Description

Intensity of a 100-year fluvial flood event for the historical baseline (1980). Data is generated by square-root normalising depth meters baseline variable into a dimensionless index using Joint Research Centre global depth-damage relationships.

Usage notes

Values range from 0 to 1. This variable is only available for the historical baseline year (1980), with all other years containing null values.

intensity_rcp4p5

NoDataNaN

Description

Intensity of a 100-year fluvial flood event for 2030, 2050, and 2080 under IPCC climate scenario RCP4.5. See intensity baseline variable for methodology.

Usage notes

Values range from 0 to 1. Values should be interpreted in comparison to intensity baseline variable. This variable is available for years 2030, 2050, and 2080. The historical baseline year (1980) contains null values.

intensity_rcp8p5

NoDataNaN

Description

Intensity of a 100-year fluvial flood event for 2030, 2050, and 2080 under IPCC climate scenario RCP8.5. See intensity baseline variable for methodology.

Usage notes

Values range from 0 to 1. Values should be interpreted in comparison to intensity baseline variable. This variable is available for years 2030, 2050, and 2080. The historical baseline year (1980) contains null values.

probability_baseline

NoDataNaN

Description

Annual probability of fluvial flood depths exceeding damage-relevant thresholds for the historical baseline (1980). Data is generated by calculating ensemble probabilities of exceeding thresholds across multiple inundation levels from WRI Aqueduct data.

Usage notes

Values range from 0 to 1. This variable is only available for the historical baseline year (1980), with all other years containing null values.

probability_rcp4p5

NoDataNaN

Description

Annual probability of fluvial flood depths exceeding damage-relevant thresholds for 2030, 2050, and 2080 under IPCC climate scenario RCP4.5. See probability baseline variable for methodology.

Usage notes

Values should be interpreted in comparison to probability baseline variable. This variable is available for years 2030, 2050, and 2080. The historical baseline year (1980) contains null values.

probability_rcp8p5

NoDataNaN

Description

Annual probability of fluvial flood depths exceeding damage-relevant thresholds for 2030, 2050, and 2080 under IPCC climate scenario RCP8.5. See probability baseline variable for methodology.

Usage notes

Values should be interpreted in comparison to probability baseline variable. This variable is available for years 2030, 2050, and 2080. The historical baseline year (1980) contains null values.

Pricing

Please see this document for more details.

VolumePrice
Any (Entry tier)$5/ha
Any (Subscription tier)$0.5/ha

Resources

Provider

Emmi

emmi.io

Emmi is a climate data provider based in Sydney (Australia). Emmi produces data across emissions, transition risk, and physical risk, including hazard modeling, temperature alignment, emissions targets and forecasts, and market-based physical risk across multiple scenarios.