Emergency mapping guidelines
Floods
Contents
1. INFORMATION CONTENT
2. Rapid Mode - REFERENCE/PRE-EVENT MAP
3. Rapid Mode - IMPACT/DELINEATION/GRADING/DAMAGE LEVEL/POST-EVENT MAP
4. Rapid mode - SITUATION UPDATE, EVENT MONITORING MAP
5. Risk & Recovery mode – hazard, vulnerability and risk mapping
1. INFORMATION CONTENT
Damaging flood events occur both over very large areas (e.g., major river valleys in eastern Germany and Poland; South Sudan) and, for damage assessment purposes, small areas (e.g., urban flooding). After Hurricane Katrina (2005) along the southern Gulf Coast, USA, there was a need for very detailed, neighbourhood-level mapping in New Orleans of still-standing water over weeks of time (as also recently the case for Srinigar, India, 2014). Yet also there was a need for an overview of the entire central Gulf coastline, as the storm had caused flooding of many small communities that were difficult to access by local authorities. Another example is the Kashmir flooding that affected Srinigar, India: the event also extended along the Chenab River and affected settlements far into Pakistan. In both cases, sustained observation was also necessary, as the event over the whole area-affected lasted weeks. Experience thus indicates that flood emergency mapping must proceed at two spatial scales: 1) moderate spatial resolution (100 – 300 m pixels), but wide area: so that an “event map” can cover the entire area affected; and 2) high spatial resolution “detail” maps (1-100 m pixels), so that the necessary detail for damage assessment can be obtained.
Use of sensors such as MODIS and VIIRS, with free public data access, offer clear opportunity for many organizations to produce and disseminate the first, and such products can also provide the temporal coverage needed for sustained events. A wide variety of other sensors, some with free access and others restricted, can provide the detailed coverage, which will normally also be much more restricted in temporal sampling. Both types of products are often necessary and they are complementary: an isolated SPOT-based flood map, for example, lacks the information needed to determine where in the course of a flood event it was obtained. That is, does the map show maximum flood extent, or not? However, when integrated with sustained mapping at a coarser spatial resolution (e.g., at 250 m, via MODIS), then the SPOT coverage and detailed mapping can now be accurately interpreted as covering pre-flood peak, or peak, or post –flood peak.
The recommended information content for the two types of maps differs somewhat. The challenge is to move towards similar mapping methods so that the end users can be most efficiently served.
Flood Event Overview maps can be at a variety of map scales, depending on the event, but should provide in a single view (when possible) coverage of the entire event. In some cases, when the event is over an exceptionally large area, several overview maps may be needed. These maps can show current conditions, as they are updated with latest-available sensing data, but can also accumulate the record of flooding and so serve as maximum flood extent maps, thereby providing a useful permanent record of the event for hazard assessment purposes (see below). They can also provide the locations of the Flood Detail Maps.
Flood Detail Maps are at high spatial resolution, cover much smaller areas of land, and also commonly cover particular, widely spaced, moments in time. Their geographic locations, as for the Overview maps, can be coded via GeoRSS, but it is also helpful if their coverage is illustrated directly in the Overview maps.
For both types of maps, the flood inundation limits should be provided. A variety of geospatial data formats are available (e.g., .shp files are commonly shared already; other formats are coming online). Such flood outlines (delineations) are also commonly associated with a specific calendar date and time). Meta-data or GIS attribute data should include sensor name, sensor spatial resolution, and specific sensor data products and water classification algorithms employed, by name.
For Overview maps, and in some cases for detail maps, the flood inundation extent is assembled from more than one image obtained at more than one time. Thus, cloud cover interferes with optical flood mapping. Especially for overview maps, and, less often for detailed maps, the map-maker may wish to show composite flood extent information in order to provide coverage over areas intermittently obscured in one image or another. This situation should be described in a map legend, which at a minimum should provide: map scale, map projection, title of map, producing organization and/or author, base map source, flood inundation extents (one or more colors, coded by date), and where composite “maximum flood extent” is shown, the dates and sources of satellite images used to produce this composite product.
Flood categories (river-, pluvial-, flash-floods), precursors and early warning; potential data resources – satellites, in-situ; users
2. Rapid Mode - REFERENCE/PRE-EVENT MAP
The aim of a pre-event map is to quickly provide knowledge on the territory and assets prior to the emergency. The content consists of selected topographic features of the area affected by the disaster, in particular exposed assets and other available information that can assist the users in their specific crisis management tasks.
The reference map is based on available reference data and the pre-event images, when available. If pre-event images are not available, the reference map will be based on reference data, the post-event image and ancillary information from other resources.
Potential exposure: critical infrastructure to be emphasized.
3. Rapid Mode - IMPACT/DELINEATION/GRADING/DAMAGE LEVEL/POST-EVENT MAP
Delineation maps provide an assessment of the event impact and extent. Delineation maps are directly derived from satellite images acquired immediately after the emergency event.
Flood extent as fast as possible.
When relevant, they may be combined with digital modelling and compared with archive information of similar event occurrences.
Flood extent + normal water bodies (reference map data) = grading map
Additionally, damage level maps provide an assessment of the damage (and eventually of its evolution). Damage level maps are directly derived from satellite images acquired immediately after the emergency event. When relevant, they may be combined with digital modelling and compared with archive information of similar event occurrences. Damage level maps include the extent, type and damage specific to the event. They may also provide relevant and up-to-date information that is specific to critical infrastructures, transportation systems, aid and reconstruction logistics, government and community buildings, hazard exposure, displaced population, etc.
Potential/real exposure (critical infrastructure, population, agriculture, etc.) + flood extent = impact/damage level map
4. Rapid mode - SITUATION UPDATE, EVENT MONITORING MAP
Impact/Delineation/Grading/Damage level/Post-Event maps can be updated to provide an assessment of the evolution of the event impact and extent.
Peak-flood level, rate of retreat, dam breakage, etc. = essential components for monitoring.
Monitoring maps: need clear indication of date/time of data capture (instead of production) and highlight change (+/-)
5. Risk & Recovery mode – hazard, vulnerability and risk mapping
To be included here – specifically important for floods - or as specific section under a multi-hazard Risk & Recovery module?