Digital mapping is vital to everything from crisis rescue missions to car navigation, writes Karlin Lillington

Marrying digital maps with other geographically-relevant data might seem at first glance a sleepy corner of the computing world, but nothing could be further from the case.

GIS - geographic information systems - can contribute to or help resolve political and cultural disputes, has been central to global rescue efforts, and raises pressing new issues, arising out of the ability to digitise and distribute mapped information, says Prof Michael Goodchild of the University of California, Santa Barbara. The distinguished academic, whose year as visiting professor at NUI Maynooth draws to a close this month, sees GIS as perhaps the most multidisciplinary area of computer science, an area of computing intertwined with how we see ourselves and the world.

"GIS is very pluralistic and has a very legitimate claim to being a common ground between the human and the physical world," he says. Both NUI Maynooth and UC Santa Barbara are recognised internationally for research work in GIS. Goodchild's background exemplifies how GIS weaves together many interests and areas of study. A basic degree in physics from Cambridge led to a PhD in geography at Canada's McMaster University.

He went into his career in academia just as computers were emerging as tools in geography, for both topographical and data mapping. He found the overlap between technology and geography - aspects of which are referred to as remote sensing - a fascinating discipline that has entranced him ever since.

But the niche area had a hard time establishing legitimacy. "It's been controversial as, within computer science, people that taught remote sensing [ and GIS] were seen as second-class citizens. At first, we couldn't have PhD students, for example," he says.

Initially, "it wasn't really clear that you could use a computer to store a map. You had to take a two dimensional object and somehow stick it on a piece of tape as a sequence of 0s and 1s", he says. The area really started to take off in the 1980s as the PC industry emerged and computing became more accessible.

Using computers for mapping was initially the realm of military and intelligence agencies and GIS came loaded with potential agendas, he says. Mapmaking has historically always "been used to power the empowered" - to name territories and ownership of them, to describe new lands and claim them as colonies. Identifying and bringing such political, economic and social dimensions into computer science departments raised eyebrows. "At first, the reaction was, 'oh, come on'. But the debate that emerged in the 1990s was very similar to that around the Manhattan Project," Goodchild says. GIS certainly hasn't moved away from controversy in this decade. Researchers are deeply frustrated that, since 9/11, "in the name of national security, many things once intended as public and open" - satellite imagery, topographical data and so on - are now considered classified information.

At the same time, he says state-run scientific organisations, that once funded the gathering and management of geographical data, are having their funding slashed or are being closed down. For example, US representative Newt Gingrich tried to close down the US Geological Survey in the 1990s and NASA, which gathers satellite data essential to GIS, has had its budget reduced repeatedly.

The US Geological Survey no longer updates topographical maps - traditional maps have been in decline since the 1960s, says Goodchild - so street maps and car navigation systems are mostly created by the private sector now. These maps are pieced together from people driving around with a GPS system to work out co-ordinates of streets, or by flying over the area, he says.

"It's the fastest and cheapest way to make maps," he notes, but it isn't particularly accurate. And maps tend to be made only for regions where populations can afford to buy them, with sometimes disastrous effect. He says that, after the horrific Stephen's Day tsunami in 2004, "in Banda Aceh, the only map rescuers had was drawn on the back of a napkin".

As more mapping data moves from print to online only, such situations are bound to repeat themselves and must be addressed, says Goodchild.

His own research has focused on the question of accuracy in geographical data. Some of this is a physical phenomenon: "In reality, we can't measure accurately to less than five metres because, in essence, everything slops around on the earth's surface within a five-metre error. But we get around this because everything is relative, and we can position something by knowing it lies near something else."

Then there is human error. Some of this is on purpose: "There's always been a deliberate distortion of maps in the interest of security, to hide things. But there are also cartographic traps: deliberate errors in printed maps to protect content. The curvature of a road might be exaggerated, or a kink put in that doesn't really matter but is distinctive. Often it's a feature that looks distinctive but doesn't exist."

US publisher Rand McNally for years included a non-existent suburb in its maps of Chicago. Why? If anyone tries to issue maps that include any of these fake elements, the publishers know the map data have been copied and has grounds for a copyright case. Other errors are the result of a lack of standards - data are collected in more than 300 formats - and little co-ordination among international bodies. The situation is exacerbated by the decline in the availability of reliable data from scientific bodies, as mapmaking is privatised. That means there can be great variability in the data sets one wants to layer into a GIS.

"GIS allows you to integrate and layer data from different sources, but in reality, they never actually fit, with some interesting consequences," says Goodchild. The Ordnance Survey of Northern Ireland and the Republic's Ordnance Survey use different methods of gathering data and hence, the borders do not line up when the maps from each are brought together, he says.

For similar reasons there's a bizarre virtual "step" of two metres between Belgium and the Netherlands. And emergency services responding to GPS co-ordinates of an accident may find they are actually a street over from the incident, he says.

All these elements fold into his particular interest, the "spatial web", which Goodchild describes as "a set of issues and relationships relating to how information is presented on the web using a spatial infrastructure."

That infrastructure includes the technical (data formats, standards, systems of classification) but also the cultural (terminology, languages and characters, different organisations).

"Integrating this data across Europe is incredibly difficult. Every state, every county, every municipality gathers data. Farmers even map their own soils. How do you integrate all of this? It's like a patchwork quilt," he says. The same is true everywhere, he notes. Because the internet is how we communicate and share geographic data, solving this problem is crucial, he says.

"Ireland is the right place to be right now because of these issues. Ireland is a microcosm of these issues and Europe is where the most is happening to address them."

The Republic interests a geographer because "many areas of mapping are very underdeveloped" such as placename geography, postcodes, "and the street names are a mess", he laughs. "I find a fascinating contrast here between a State focus on IT and the reality on the ground."

So how can we tackle the challenges offered by the spatial web, the erosion of reliable sources of data and the lack of awareness? "It really is a massive problem ... Geography is infrastructure and infrastructure is what people find most difficult to pay for."

But he adds that, for the first time, there at least is a wider public understanding of the uses of GIS, thanks to resources like Google Earth. He hopes this might stimulate debate on the issues so that disaster relief organisations are not lacking for maps in the future.

"It's absolutely essential - but it is on nobody's radar as an immediate need. It's very hard to find anybody who understands the seriousness of the situation."

He says researchers and organisations need to work out standards, need to share information internationally, need to get out from under ossified institutions with career administrators disconnected from such issues. The European Science Foundation offers a good model for how international collaboration might work, he says.

Technology, he says, "can let humans see the problems more clearly and produce the solutions more efficiently. But humans have had many of these problems for 5,000 years."

"Technology isn't going to supply a magic bullet."