Image credit: Hatfield Consultants
Flood extent products can be delivered from the project/license scale to the regional scale. A combination of radar and optical data for flood mapping is usually implemented, as complementary information can improve the overall products.
Radar images are not affected by cloud cover, meaning that a sequence of radar images can capture flood extent progression regardless of weather conditions. Flood extent products from radar are generated based on multi-temporal images and change detection between a dry or flood‑free image and an image of a flood event.
Historical flood extent mapping is possible depending on the archive of radar images from a number of different radar sensors. New acquisitions of radar images to map flood extent may be planned, depending on the sensor.
Processing of radar images to produce flood extent maps can be completed in near-real-time, if the baseline data and site characteristics are established before a flood event.
Multi-spectral optical images can be used to generate flood extent (based on the strong contrast between reflectance from water and land surfaces in near-infrared spectral bands). Depending on spatial resolution and vegetation density, optical images can also identify flooded vegetation.
The flood extent product provides the analysed spatial limit of flooding (raster or vector).
Known restrictions / limitations
A potential limitation with flood extent products is the presence and density of vegetation, infrastructure and assets (which may affect the ability and reliability of flood extent detection). Approaches to mitigate this issue include:
1) Using radar with longer wavelength (e.g., L-band) that provides better penetration of a vegetation canopy;
2) Mapping of baseline vegetation using optical earth observation data and adjusting flood extent detection procedures; and
3) Using existing infrastructure and buildings data to improve flood extent detection and identify areas of lower detection certainty. Elevation and derived slope information can also improve flood extent mapping in mountainous landscapes, depending on the accuracy of the elevation data.
Historical flood extent mapping depends on archived images. Data may not be available for all flood events.
The mapping resolution of radar data and products may be reduced due to the effects of image speckle and the need to filter image and output flood extent products.
Similar limitations with optical flood extent products may exist, such as density of vegetation, infrastructure and assets. Cloud cover that may be associated with a flood event can prevent the acquisition of images.
Lifecycle stage and demand
Pre-license: Information on historical flood extent to support decision-making on a prospect.
Exploration: Critical historical information to support effective and safe land seismic survey. Information on flood extent may be provided in near real-time depending on risks within the operating area.
Development: Critical historical information for planning and design of infrastructure, to support understanding of floodplain areas and food hazards and risks in a proposed development area.
Production: Flood forecasts to mitigate risk and to improve H&S of personnel as well as reduce damage to equipment. Information on flood extent in near real-time depending on risks within the operating area to quantify the potential damage extent.
Decommissioning: Information on historical flood extent to support safe decommissioning of the site.
Geographic coverage and demand
Demand is global, focusing on tropical countries and any area susceptible to flooding (including flash floods).
Demand is focused on floodplains and low-lying areas.
High demand for information on flooding in forested lowland areas.
OTM:036 Geohazard exposure analysis
OTM:065 Flood plain mapping
OTM:072 Monitoring flash floods
Input data sources
Optical: HR1 and HR2
Radar: HR1, HR2, MR1
Spatial resolution and coverage
Spatial resolution: 4–100 m.
Varies depending on input imagery used and client needs. In case of radar images speckle effects may reduce the final spatial resolution of images.
Minimum Mapping Unit (MMU)
Variable, depending on source data resolution, assumed scale of the final product and expected minimum size of the flooded area. An MMU as small as 0.25 ha is possible.
Accuracy / constraints
Thematic accuracy: 80-90% in areas of low vegetation cover and density.
Spatial accuracy: The goal would be 1 pixel, but depends on reference data.
Accuracy assessment approach & quality control measures
Cross-validation based on observations of the flooded area on cloud-free days using optical data and radar observations.
Existing databases and river level gauges. Ground-based observations or validation survey. Statements of national or local institutions and local knowledge.
Frequency / timeliness
Observation frequency: Depending on sensor or radar beam mode (resolution/extent) selected, the frequency for new acquisitions can be as low as 3-5 days from the same satellite. Frequency of historical maps is highly variable depending on the archive.
Timeliness of delivery: radar and optical processing can be completed in near real time (< 24 hours) depending on the processing chain and availability of base images.
On-demand availability from commercial suppliers.
New acquisitions can be requested globally.
Archived products available for public search. Availability may be limited for specific dates and locations.
Delivery / output format
# of Pages:
Internal – Project consortium and science partners
External – ESA
This page has no comments.