Embed Size (px)
Citation preview
Large-scale flooding is a seriousproblem in many parts of the world. InAsia, countries such as China andBangladesh regularly suffer, while inEurope many countries are at risk. Thereis danger of loss of life and seriousdamage to the economies of the areasaffected.
For this reason, the ANFAS projecthas been set up to include partners fromEurope and China. The partners bringtogether expertise in remote sensing,scientific computing, computer vision,internet technology, geographicinformation systems, knowledge-basedsystems, soil sciences, and other areas.
When there is a risk of flooding,decision makers have to decide the mostappropriate actions to take: evacuation ofthe population, reinforcement of dikes,intentional breaking of dikes, etc. The aimof the ANFAS project is to develop asimulation and prevention tool to helpdecision-makers take decisions that willlimit flood damage.
The ANFAS project will use datafrom the most advanced acquisitiontechnologies. In particular, remote sensingimagery – optical radar, interferometryradar – will be incorporated in aconventional Geographical InformationSystem database. Advanced dataprocessing tools for scene modelling andflood simulation will be developed.Computer Vision and ScientificComputing will be used together in orderto take into account information extractedfrom real images for the calculation ofparameters of the simulation model. Theproject will have strong end-usersinvolvement through the definition of theobjectives, the design of the system andthe evaluation of the simulation results.It ensures that the System answers to‘terrain needs’ and is well adapted inpractical use. Industrial partners areinvolved for the implementation of the
final system. One of the maincharacteristics of the system will be itsmodularity, i.e., it will be composed ofeasily interchangeable blocks.
The functionality of the ANFASsystem will be:• to perform flood simulation, simulating
water flow propagation based onscenarios. The user will be able tointeract with the scene that models thereal site in order to add/remove dikesor other human constructions
• to assess flood damage. Floodassessment can be done either using thesimulation results or using the remotesensed images of a real flood event
• at a prospective level, to analysefloodplain morphology includingchanges in riverbed and subsurfaceproperties due to repeated floods at agiven site.
The output of the system will be:• visualisation: consisting of two-
dimensional maps of the extent of theflood. If simulation is performed, time
sequence maps tracing the propagationof the simulated flood will be provided
• direct impact assessment, evaluation ofthe total affected area
• regional and statistical analyses overthe affected area
• simple socio-economic assessment ofdamage directly caused by flooding.
Pilot Sites The ANFAS Information System
will be applied and validated on three pilotsites: the Vah river in Slovakia will beused to set up the method and the system;the Loire river in France and the JingJiangReach in China – along the Yangtze River,80km dowstream from the Three Gorges– will be used to validate and consolidateif necessary, the developed approach.
For these sites, the scene modelling(geometrical and photometric modelling)will be entirely achieved. This means thateach site will be computationallydescribed as a geometrical threedimensional surface, completed by a
When there is a risk of flooding, decision makers haveto decide what are the most appropriate reactions:evacuation of the population, reinforcement of dykes,etc. The ANFAS project aims to develop a decision
support system for flood prevention and protection,integrating the most advanced techniques in dataprocessing and management.
.ERCIM Project Fact Sheet.
ANFAS – Data Fusion for Flood Analysis and Decision Support
Floods regularly cause life loss and damage like here in France in 1999. Theaim of the ANFAS project is to develop a simulation and prevention tool tolimit flood damage.
multi-layer description of the underground(e.g., geology, hydraulic properties ofmajor aquifers and soil, units, urbanregions, forests, fields, reservoirs, dikepositioning) and the upper ground (rainfallrate). This description will be done atvarious scales, depending on the type ofthe feature, the exact localisation and theavailable data.
The choice of the sites is motivatedby the similarity between them: they arein a natural state where their beds havenot been entirely reconstructed. The majordifference between China and Europe isthe size of the sites: very large in the firstcase, much smaller for the second one.
Floods are the result of variousfactors including: short duration heavyrainfall upstream, long duration lowintensity rainfall, failure of dams or dikesor snowmelt. In Europe, the floods are atthe end of the winter season, generatedmostly by snowmelt; in China, on thecontrary, flood period is currently at thebeginning of the summer due to the heavyrainfall.
In Slovakia, flood protection andproblems of flood control are among themost important and also challenging tasksof hydrologists and river engineers. TheVar river has a mean discharge of 175m3.s-1 and the annual mean rainfall is926mm It is heavily influenced by rivertraining, dam construction, hydro-powerutilisation – mostly derivation schemeswith lateral canals and abandoned originalriver bed.
The Loire river in France will beused to test the developed IntegratedInformation System. A large part of theriver (450 km), called the middle Loire,is characterised by a high gradient, asandy and mobile bed, continuouslyreworked by the river and a man-madetransformed bed, 500 km of dikes beinglocated along this part of the river. Thesedikes delay flooding of the neighbouringvalleys and form an artificial boundarythreatened by the river in each high water-flow period.
These facts are responsible for highcurrents at high-water and an importanterosion and sediment transport capacity.A numerical model of the middle Loirewas developed in 1997 by an engineeringcompany. The model is based upon thedetailed topographic features of the mostimportant hydraulic structures includingdikes. The result of the ANFAS simulationwill be compared to the results from thisalready existing model. This test andvalidation will be done with Plan LoireGrandeur Nature who will provide the(Graphical Information System) GIS dataand the Digital Elevation Model (DEM)for the requirements of this applicationand who might run specific simulationsfor the assessment of the ANFASInformation System.
The sites in China is located alongthe Yangzi river which has manyimportant afluents and deposits largeamount of sediment. Its flow can reach75000m3/s (1931, Wuhan). Duringsummer, part of the excess water isdiverted into the Dongting and Poyanglakes, which play a natural reservoir role.But their capacity is most of the timeinsufficient.
The Three Gorges dam, to becompleted in 2009, will regulate the waterflow for flood control. The Three Gorgesis of particular interest to the project inorder to simulate the voluntary floodingof the filling phase of the reservoir. Thesize of the reservoir leads to manydifficulties that the system may be able topredict.
Expected resultsAt the end of the project, the final
result will be:• An Integrated Information Systems
adapted to end-user requirementsconsisting of three sub-systems asdescribed below.
• Three sub-systems: (i) a GeographicInformation System (GIS) (ii) Tools forScene Modelling from Images and (iii)a Flood Simulation Model.
The final system will be animmensely valuable tool in planningnational and regional responses to thethreat of flooding, and will be adaptableto other areas.
ANFAS is a research projectcoordinated by ERCIM
The Partners: • Bureau de Recherches
Geologiques et Minières, France• CLRC, United Kingdom• IACM-FORTH, Greece• INRIA, France• Matra Systemes & Information,
France• Slovak Academy of Sciences,
Institute of Informatics, Slovakia• University of Reading• Chinese Academy of Sciences,
Institute of Automation• Chinese Academy of Sciences,
Institute of Remote SensingApplications
• Chinese Academy of Sciences,Institute of Atmospheric Physics
• Wuhan University, China• Yangtse River Water Resources
Commission, China
Duration1 January 2000 - 31 December 2002
Budget: 4 280 000 Euro
Funding Agencies:• European Commission,
IST Programme: 1,600 000 Euro• Chinese Ministry of Science
and Technology: 1 000 000 Euro• World Bank: 245 000 USD
Scientific Coordination: Poulicos Prastacos, ICS-FORTHE-mail: [email protected]
Project Coordination:ERCIM EEIGE-mail: [email protected]://www.ercim.org
ANFAS project homepage:http://www.ercim.org/anfas/
Colour composition from radarimages, Poyang Lake, China, 1995.
