|
|
Opening up the map
By Peter Ferne, Technical Consultant, Futurelab |
For years the field of Geographic Information Systems, or GIS for short, has been the exclusive preserve of professional geographers and cartographers. GIS brings the power of information technology to bear on the making of maps.
Once the base map data is available digitally it can be displayed and manipulated in any number of interesting and enlightening ways. At the simplest level it becomes easy to pan around large maps and to zoom smoothly between different scales. But GIS offers much more - it allows you to add virtually any kind of data with a spatial or geographic component in its own layer to build composite views.
A council planning department, or a residents group, wanting to determine where to plant trees in a public square might start with a base layer containing a street map and add separate layers to represent the locations of drains, gas pipes, electricity cables and the phone network. Politicians or civil servants determining the boundaries of electoral constituencies might look at combined data from the census and other sources to map population density and economic activity.
There have been two main reasons why this incredibly useful technology has not, until very recently, become more widespread - the cost and complexity of the software, and the difficulty in obtaining the base layer of geographical data.
Commercial GIS software costs many thousands of pounds and is geared very much towards departmental use, often requiring dedicated servers and specially trained technical staff. Although at least one full featured GIS package - the Geographic Resources Analysis Support System, or GRASS for short - is available for free under an open source licence, it is still a sufficiently complex and daunting collection of tools to discourage experimentation by all but the most dedicated of amateur cartographers.
Since 2001 the World Wide Web Consortium (W3C) has been working with a large number of researchers and industrial partners on a collaborative effort called the Semantic Web. By extending the way that web pages are written to convey the meaning of the information as well as its presentation, and to do it in a way which makes that meaning accessible to machines as well as to people, it hopes to provide a common framework that allows data to be shared and reused across application, enterprise, and community boundaries.
One strand of semantic web development which has seen some early success has been the practice of tagging web pages with location data - known as geotagging. For example the author of a website can tag a web page with the location of their home city or tag a photograph with the location where it was taken. These pages can then be automatically added to maps at aggregating websites such as geourl.org or geobloggers.com which make explicit the hidden relationships between websites and physical sites. These aggregators also allow people to search for websites using geographical criteria, such as within 10km of a particular website.
This practice is giving rise to what has become known as the geospatial web. In contrast to fully fledged GIS the emphasis is on ease of adoption and integration, meaning that thousands of people are gradually adding geospatial metadata which, as the network effect grows, becomes progressively more valuable.
Two recent events which have given the whole area a huge boost, and raised the bar in terms of what is easily achievable, have been the opening up by Google and Yahoo of their online mapping services. You can now use either service to plot any data which you have access to on their maps on your website. Both services are currently free, although they do reserve the right to display adverts alongside the map. Importantly, using these services allows you to avoid the potentially complex business of installing and configuring your own mapping software. You can have a real map-based application up and running in a matter of hours.
Two of the more notable examples of people spontaneously adding value to data from public sources are HousingMaps and the Chicago Crime Map. HousingMaps pulls in houses and flats for sale and to rent across the USA from the classified ads at craigslist.com and plots them on maps so that you can see exactly where they are and what the area is like. In fact if the property is in Chicago you can use the Chicago Crime Map to see what the crime rate is in the area. The site uses data on reported crimes from the Chicago Police Department's website and plots them on maps of the city. It allows you to see how many of what types of crime have been reported in a given area or even along a given route, perhaps the one you take to work or to school.
Most topically these tools are allowing distributed data gathering and collation in situations where more formal channels are not functioning fast enough. The Katrina Information Map has been built for people affected by Hurricane Katrina to both provide and find real-time updates of information about the status of specific locations, down to individual houses, affected by the storm and its aftermath. This site and others like it are enabling vital information to be shared more widely and more quickly than via more established mechanisms.
Within days Google maps already included satellite imagery taken after the flood and this has been combined with other data to develop the New Orleans Flood Map which shows the depth of floodwater at any given location where the data is available.
If you don't want to rely on external service providers such as Google or Yahoo, and you want to develop your own standalone services, there are open source tools available, such as MapServer and ka-Maps, which provide similar functionality. However, aside from the need to set up and run your own server the main drawback with this approach, at least in the UK, is getting access to the geographic data (geodata).
In the USA geodata which has been gathered with public money is placed in the public domain and is available for anybody to make use of as they see fit. This has enabled an explosion of activity in this area. In contrast UK geodata gathered by the Ordnance Survey is subject to Crown Copyright for 50 years from the date of first publication, meaning that up-to-date data is only available by paying hefty commercial licensing fees to the OS. This of course puts it out of the reach of most organisations and certainly the vast majority of social enterprises and individuals. Crucially it makes experimenting with such services prohibitively expensive, severely inhibiting the creative public service innovations coming from grass roots developers.
Some people have become so frustrated by this that they have started regathering data through other channels. Perhaps the most radical of these is OpenStreetMap, where people are actually walking, or cycling, round the streets with GPS units, uploading their data and making it available under a Creative Commons licence for others to share and exploit, collaboratively building afresh the street maps already created with public funding by the OS.
In parallel with the practical approach is the political - there is now an online petition to lobby for open access to state-collected geospatial data. Although getting volunteers to walk along every street in the country in order to generate up-to-date high quality open geodata might seem like a lengthy process, there are some that think it will be quicker than persuading the powers that be to make the existing data more readily available - let's hope they are wrong.
By Peter Ferne (www.petef.com)
Links
Open mapping resources: petef.org/wp/open-mapping/
Support open access to state-collected geospatial data: www.pledgebank.com/geodata
September 2005
|
|
