Earth's wetlands are havens for wildlife and vital to thewater cycle, but they are also under threat. An ESA-led initiative aimsto develop a global wetland information service based on EarthObservation for conservation efforts. The Globwetland project has nowentered a new phase, with prototype products being developed based onsites across four continents.
Abundant water makes wetlands themost biologically diverse ecosystems on Earth, more productive eventhan tropical rainforests. Unlike rainforests, they are scatteredacross the world, providing regional flood and erosion prevention,water purification and nutrient recycling.
For much of the 20thCentury, wetlands were drained or otherwise degraded. However growingunderstanding of the vital importance of wetlands led to the signing in1971 of the Ramsar Convention on Wetlands. Today more than 142930wetlands have been designated as Wetlands of International Importance,a total area of 125 million hectares. The Convention's 145 nationalsignatories commit to maintaining the ecological character and areobliged to reporting on the state of listed wetlands they havedesignated.
Taking place as part of ESA's Earth Observation DataUser Element, the aim of Globwetland is to utilise satellite imagery toprovide detailed wide-area views of individual wetlands and theirsurrounding catchment areas, and how they are changing. This willassist with Ramsar reporting, and also aid national and localconservation efforts – the success of which ultimately comes down toindividual wetland managers.
For this reason Globwetland isuser-oriented, developing and demonstrating a space-based informationservice based on the specific requests of local and national usersacross 50 sites in 21 countries worldwide.
"Our aim is to providewhat the wetland managers think is useful," explained Kevin Jones ofVexcel Canada, the company managing Globwetland. "So the first phasewas carrying out a detailed survey of user requirements, to know whatsites should be acquired, in what style.
"We also need to ensurethat products that can be employed locally are also standardised to beof use nationally and internationally.
"Now comes the stage ofprototyping products, with specific products being developed based onuser requirements. In turn, wetland managers are helping us with groundtruthing our products – verifying that what is shown in an image isreally there."
Based on the user requirements, Globwetland coreproducts include base maps, land use-land cover (LULC) maps and changedetection maps – with historical satellite images being compared withcurrent acquisitions to see what changes have occurred during the lastten years or more. Water cycle regime monitoring maps are also beingcrated using Envisat and Radarsat radar data to show flood and retreatpatterns.
Specialised products available if requested includedigital elevation models (DEMs) and biophysical data acquired frommultispectral satellite sensors, showing vegetation health based onchlorophyll levels or the sediment contents of wetland water bodies.
Cooperation in Canada
Canada'sCreston Valley Wildlife Management Area is a Ramsar site in BritishColumbia, whose management team is working with Globwetland on thedeveloping of prototype products.
Its area manager is BrianStushnoff, who explains: "The Wildlife Area covers 7 000 hectares. Itis a pretty wild area, and there are some difficult and waterloggedparts. What we need is an effective, economic means of knowing how theentire area responds to changing water levels.
"Globwetland haspresented our team with a collective opportunity to learn more aboutsatellites, what they can and cannot do. Before the project began, wenever realised there were so many different types of data available."
Managinga wetland such as the one in Creston Valley is a complex task. Toensure continued high biodiversity, the water levels of its marshesmust be periodically drawn down. Such cycles of wet and dryness mimicnature's droughts, providing oxygen to decompose dead plant materialand release nutrients.
In addition, this practice helps stopinvasive plant species, such as cattails or reed canary grass, fillingopen water and pushing out other plants.
"We try to keep up anearly stage of succession, and a high level of wetland vegetation cantend to reduce diversity," Stushnoff explained. "We have torehabilitate areas that get choked with vegetation, drying them up andthen ploughing up the ground to get rid of the seed bank that developsover time.
"What satellite multispectral or radar images can do is allow us tomap vegetation species, and see where there are concentrations ofinvasive plants to give us problems. Reed canary grass can stand two tothree metres high, so shows up well in a radar image. Radar also showsup the extent of flooded areas well, so we can see how water levels arechanging.
"Historical satellite images are useful to see patterns over time –a lot can happen in 30 years, or just a decade. Annual satelliteupdates will help us into the future."
Classifying wetlands from space
Another advantage ofthe project is that it should increase our ability to recognisedifferent types of wetlands using Earth Observation, and by extensionmore accurately classify them within land cover maps.
"Wetlands have been very difficult to identify from space,"explained Doug Taylor of Wetlands International, a non-governmentalorganisation that maintains the Ramsar Sites Database and is aGlobwetland partner. "The Millennium Ecosystem Assessment remotesensing group identified this ecosystem as the least well recognisedfrom space.
"Although sensors detecting water are useful, mostwetlands are also associated with vegetation, so characteristic wetlandsurfaces may be anything from forested or floating plants to ricepaddies or wet grassland. Building up a catalogue of characteristicsignatures for all latitudes has only just started, so the GlobWetlandproject, by virtue of its wide geographical and temporal reach, is avery useful platform."
Globwetland should mean that, for thefirst time, wetlands should be capable of being recognised solelythrough Earth Observation imagery.
In preparation, WetlandsInternational has worked with the European Environment Agency todevelop an extension to their Corine Land Cover classification system,developed for the 300-metre-resolution Corine European land cover mapderived from satellites. This extension is based on the globallyauthoritative 'Ramsar Classification for Wetland Type' system, coveringall known wetland types.
Kampala COP 9 will present progress
TheNinth Meeting of the Conference to the Parties (COP 9) of the RamsarConvention is due to take place in Kampala, Uganda from 8 November2005. The meeting will include a presentation from the Globwetland teamon the project's progress so far.
Globwetland involves sitesacross four continents: North and South America, Europe (includingEuropean Russia) and Africa, the countries involved here being Algeria,Egypt, Cameroon, Central African Republic, Chad, Niger, Kenya, Nigeria,Senegal and South Africa.
"This will be an important opportunityto demonstrate the innovative work of the GLobWetland project, showinghow Earth Observation can be used fruitfully to support wetlandmanagers on the ground in their management planning for Ramsar sites"says Nick Davidson, the Ramsar Convention's Deputy Secretary General."Many of us recognise the huge potential for satellite-derivedinformation to help,but for the non-expert it's much harder to find outjust what Earth Observation tools and sources can help us for eachsite-specific management issue.
"The experience of the EarthObservation analysts and site managers at these 50 Ramsar sites, manyof them in Africa, will help all of us involved in the conservation andwise use of wetlands, wherever we are."
VexcelCanada Inc is leading the Globwetland consortium for ESA's DUE, withsubcontractors Remote Sensing Solutions GmbH of Germany and SynopticsRemote Sensing and GIS Applications of the Netherlands. TheNetherlands-based Wetlands International organisation is also takingpart, both as a sub-contractor and also as a link to the global wetlandscience community.
Cite This Page: