As mapping becomes ubiquitous, cartography as a craft needs to take into account the vast amounts of constantly updated information. While this presents an entirely new set of challenges for the cartographer, it also provides new opportunities for visualising the world around us, explains Managing Editor Prof Arup Dasgupta
Hans-Peter Brondmo, Vice-President and head of new product innovation for Nokia HERE, spoke at this year’s GigaOM Mobilize conference about making sense of the real world through mapping. He said a map is really just a way to guide one through his or her life; so Nokia is interested in creating maps that would model the real world, one that can actually show where one is in an accurate 3D representation.
The evolution of cartography and visualisation from its beginning to this day is exemplified by this statement. Early map makers, who charted the known world through adventurous voyages, through explorers’ accounts or even by the ‘lifting’ of maps prepared by others, would paint unknown areas in dark shades with the warning “Here be monsters”, a colourful way of visualising the statement “I know no more of this region”.
When and where did cartography begin? That is a difficult question to answer. However, the foundations of modern cartography were laid by the Greeks. Aristotle visualised a spherical earth and it is known that the Greek philosopher Eratosthenes determined the diameter of the earth and divided it into meridians of latitude and longitude to enable placement of map features on a globe. Ptolemy brought in the concept of celestial measurements to locate features accurately on the globe. His eight volume atlas, Geographia, was another first with place names, north orientation, scale, legends and symbology. These were the fundamental elements of visualisation that continue to be used till this day. From these beginnings cartography and visualisation have progressed, drawing on technologies as they evolved.
The humble chain and vernier theodolite have given way to total stations, DGPS and remote sensing systems. Log tables and slide rules have been replaced by computers. Scribing tools, mylars and four-colour printing presses have been replaced by high-speed large-format digital colour printers and paper maps have been superseded by Internet publishing. The goal however, remains the same: to be able to communicate to the user a relevant view of the world. This view has today become interactive with the end user deciding what they want to see, how and where.
Science & technology or art?
On one hand, there are aesthetic maps which convey nothing as a map, while on the other there are very informative maps presented crudely in clashing colours, unnecessarily bold fonts and clumsy patterns. Eminent cartographer Erwin Josephus Raisz had stated years back that the good cartographer was both a scientist and an artist. But with digitisation and automation taking over cartography, how much of it holds true today?
Ed Parsons, Geospatial Technologist, Google, agrees with Raisz but also feels that the balance now is towards the artist, as much of the science or at least application of appropriate algorithms to manipulate information is carried out by software. “Now the cartographer more than ever must use his or her judgment to choose the appropriate techniques to communicate their intended purpose. Of course today, designing for the dynamic medium of a display screen, the range of techniques has hugely expanded,” he adds.
Prof Menno Kraak of the University of Twente, Faculty of Geo-Information Science and Earth Observation, agrees: “The technology to make maps gets better and better, but the result are often many maps of the same kind. The art of design plays a key role in the appreciation and usability of the map”. Dr Georg Gartner, President of the International Cartographic Association (ICA), feels that Raisz’s statement still holds true. “However, automation and digital cartography have brought up new players, thus cartography primarily is done by those who are good in automation,” he says, adding that it is necessary to emphasise the triangle of science-technology-art/design in education, find opportunities to bring proponents of these different areas together through conferences, workshops, and webinars, and finally establish instruments like awards, competitions to highlight good examples of the application of this triangle concept.
Prof Milan Konecny, chair of the ICA Commission on Cartography on Early Warning and Crises Management, believes that the statement is perhaps more valid than before, “because combination of new scientific approaches based on IT in cartography provide many possibilities for map creation and inclusion of art not only make maps attractive but also enables new and imaginative forms in which maps could be shown and naturally understood”.
This idea of the synergy between art, science and technology is brought out graphically in the conceptual diagram of the visualisation cube developed by MacEachren (Fig 1). The traditional conventions of cartography are useful in producing maps for decision support. However, geospatial technologies allow greater flexibility and customisation through geovisualisation, which develops and extends old concepts in new and innovative ways. The exploration of data for its relevance to the theme and its presentation of the theme in an intelligible and user-friendly manner are the core concept of conventional cartographic visualisation.
Representing geospatial data
Maps are two dimensional representations of geospatial data frozen in time. They represent the situation when the data was gathered. Though useful in its way, it is limited in its scope of applications unless the features it represents are persistent over time. The third dimension of height is included through artifices like contours or shading or through derivatives like slope and aspect. If we add the fourth dimension of time, the conventional system of feature representation again resorts to artifices like time series and overlays using different colours for time slices. Fast-changing data streaming from sensors defeats these conventional techniques; for example rainfall, tides, seismic events and crowd behaviour.
MacEachren’s visualisation cube. From Geographical Information Systems and Science, Paul A. Longley et. al. John Wiley and Sons
These limitations can be overcome through techniques that are made possible through the medium of computer graphics. Animation is one of the best examples of the integration of time with spatial data. 3D visualisation using wireframes and rendering of textures on the wireframe is another useful technique. Over and above this we can also add elements of interactivity like fly-through, immersive visualisation through virtual reality systems, hyper-linking to other datasets and visualisation of dynamic object behaviour like cyclones and flooding. Geovisualisation allows us to explore a model of the real world much more effectively than we have ever had the ability to do in the past, says David Watkins, Cartography Product Manager, ESRI.
In his article on Art and Cartographic Communication, (Cartography and Art, Springer, 2009), former ICA President William Cartwright discusses that data is not just confined to what can be seen, imaged or drawn but extends to other sensory stimuli like word-scapes, sound-scapes, paintings, events and even olfactory and haptic data.
A tactile map of the City of Basel
However, a problem with conventional cartographic representations as well as its modern computerised version is the absence of standards for the representation of 3D objects on the terrain; for example buildings and plantations. The technology of geodesign does attempt to bring together GIS, CAD and BIM for the construction sector, but it lacks photorealism and the output is static. Photorealism can be achieved by using photographs, which can be added as texture to the 3D CAD models. But a better way is to use a technology for creating photorealistic true 3D representations which can be manipulated in 3D space, merged with other graphic elements and is amenable to precise measurement of sizes and distances. A terrestrial laser scanner along with a highresolution CCD camera is a solution. This creates photorealistic 3D objects that can be manipulated and merged with other data. An apt example is the boat and raft in the movie Life of Pi, in which the tiger and Pi are seen together in the boat. Such true 3D rendering is increasingly being used in areas like heritage mapping, crime scene mapping, forest mapping, surveying, and piping mapping etc. However, unlike GIS, CAD and BIM, there are no standards for such photorealistic data representation.
Integration with computational environment
The integration of so many types of data into the computational environment poses serious challenges. While data like GIS, CAD, BIM and GPS are very well structured and amenable for inclusion into systems like DBMS; data like the ones described above are not structured and therefore not easily amenable to the constraints of structured data handling systems. Streaming data also has similar problems and this has given rise to the technology of Big Data Analytics. Integration of geospatial systems with Big Data is still a “work in progress’. The issue is not so much the computational power but the way to use the power on large volumes of data and visualise the results of the computation to make it easily assimilated. Examples of this are the weather models and the specific models for cyclogenesis and subsequent track determination and landfall prediction.
One of the key factors in this is the ability to discern patterns and relationships in geospatial data through visualisation. This requires the integration of visual and computational tools for human machine collaboration for knowledge construction. At the engineering end it requires the integration of diverse systems, each with their own standards, data structures and interfaces. The real problem is to integrate databases with visualisation such that the functionalities are available at the visual interaction level itself. In fact, the problem is similar to the Big Data problem of being able to handle streaming, unstructured data to extract knowledge in real time.
Kraak points out that “part of the cartographic community, such as ICA’s Commission on Geovisualisation and the Commission on Cognitive Cartography, are active in the geo-visual analytics domain”. He adds that Big Data requires an even better integration of human perception via visual interfaces like maps and geo-computation in terms of both shape and visual analytics. Gartner is of the opinion that “visual analysis is an area where visual tools are used to understand complex datasets or to derive meaning out of complex datasets, detect patterns or trends etc.” However, he points out that often visual analysis tools are rather complex themselves, so there is a big demand to establish those tools in a way, that they are not restricted to expert users only, but allow access to any interested user. “As this involves knowledge and skills of ‘efficient communication’ this is clearly a cartographer’s task.”
Unfortunately, cartography had become a second class citizen throughout the 1990s and into the early 2000s, with GIS becoming much of the focus of traditional cartography programme. As analysis became the priority, the cartographic aspects of presenting the results got a little lost along the way. “But people have begun to realise how the results are presented is as important as the analysis itself. The map is how we communicate and visualise geography. This communication is even more critical when you consider the general public who may know very little about reading a map,” says Watkins, who believes that cartography as a discipline is making a strong comeback.
According to Gartner, the main challenge lies in context modelling and personalisation. “Cartography has missed some developments and time, as it was not ready to react on the big technological developments. But this has changed since some years now. Cartography has become attractive due to its relevance,” he adds. Modern cartographers are scientists, technology-savvy and designers, and contribute to the developments directly either in national mapping organisations, in the volunteer geographic information (VGI) arena, in the industry or academia. “We can witness most successful new education programmes like the International Master on Cartography, growing numbers of excellent maps and map services being produced on the Web and on mobile platforms, and the growing interest, for instance, in the International Cartographic Association with all its instruments,” says Gartner.
The proliferation of online maps has created a consumer market that demands intuitive and usable interfaces and almost instant feedback, points out Watkins, adding, “This has helped push the development efforts of everyone in the mapping software industry and has resulted in continuous improvements.” According to Konecny, many of the scalability issues in geovisualisation tools could be solved by addressing two issues: how to deal with huge amounts of data and how to tune cartographical tools for various groups of users with different skills and knowledge. But there are still problems like how to ensure quality of data, for instance in the case of so-called VGI. There is also the question of comparability of amount and quality of data from different places. Some towns are mapped more than others and their cartographic visualisation is different, for instance in the cases of OpenStreet maps. In fact the scalability in 3D maps and questions of the potential representation on a scale of 1:1 remains open.
Users, user interfaces
The World Wide Web has provided an enormous opportunity for interactive mapping. Opportunities provided by online mapping services like Google and Bing maps, where people can produce their own personalised products and share with others, are endless. But in the absence of any standards, these outputs may not always be visually aesthetic. Gartner, however, feels that the “quality” of a map is not only to be measured by its design but by its “fitness-to-use”, and in this sense Google Maps and similar cartographic information systems are very good.
Parsons also completely rejects the notion that Google Maps and VGI has filled the world with ‘ugly’ maps. “A great deal of care goes into the cartography of Google Maps and their widespread use, a billion users per month, would not have been possible if they were ugly!” He goes on to add that while the fact that now any individual using the Web can produce a map, publish it and potentially reach an audience of billions is truly groundbreaking, too often there are maps which either don’t communicate clearly the information intended or of more concern are misleading. At a basic level, cartographic knowledge of the relative merits of projections, the need to Rendering terrain in digital 3D maps Developing projections for rendering terrain in digital 3D maps is another popular technique. The experimental projections above have been created with prototype software showing (from left to right) curved projecting rays for a panoramic view along the shores of a lake; plan oblique rendering; and plan oblique rendering with viewing angle varying along the direction of view. The Cartography & Geovisualization Group at Oregon State University. normalise area-based statistics and colour theory may all be required to design a map, says Parsons, as he flags off the issue of education or more user friendly tools.
Konecny feels the widely accepted classical cartographical methods are now enhanced through the capability of combination of maps with pictures, other models, zooming etc, giving a new dimension to map use. They allow ‘neo-cartographers’ to explore new techniques of personalised map creation thus learning how to design, use and understand maps. All these processes create new added-value for cartography and open it out to many new users. Thus, there is a growing interest among these users about rules, methods and individual steps on how to create good, true, nice and attractive maps. “Yes, still many maps are poorly designed, but we can see that the situation is improving,” he adds.
“New developments always come with a little backlash in quality,” points out Kraak. “This was the case with the introduction of scribing materials in the beginning of the 20th century and again with the introduction of the computer plotters in the 1980s and even with the introduction of GIS. We are now witnessing a similar trend.”
Aesthetics apart, the cartographic community is gung ho about the ‘interactivity’ aspect of maps and believes it is an integral part of a good map. “Whenever in the past we had cartographic tools being so ubiquitously available and easy-touse, that literally everybody can get them in her or his hands they have used it to express their perspectives, feelings, emotions, issues, etc. VGI, Grassroot Mapping and the availability of maps on mobile devices are a change of paradigm when it comes to the role of cartography for society,” says Gartner.
Parsons believes interactive maps will be a key to the development of cartography in the next decade and it needs to be refined to take into account the new world of multimedia and sensory computing. “In many cases, the map of the future will not be a map. Already today we accept that navigation is often best visualised as abstract instructions such as left arrow, forward arrow, etc, rather than a traditional overview map, or perhaps not visualised in the first place. Spoken or haptic ‘touch’ based interfaces may be more appropriate for some applications,” he says. He, however, adds that cartographers still have not really come to terms with the true potential of multimedia-based communications, using animation, sound and personalisation to communicate geographic information in different ways.
So are cartographers becoming redundant?
According to Kraak, “cartographers have the skills to make ‘correct’ well designed maps. They will still play a significant role in the professional world of the map. When it comes to the mass market, their skills might be incorporated in mashup tools”. Gartner feels the relevance (of professional cartographers) is high if the big players want to start to make a difference. In that respect, the ‘best product’ in terms of design, efficiency in communicating geoinformation, allowing to use the map for querying, exploring, finding, ordering, positioning, and entertaining will become the selling points.
As Konecny rightly points out, specialists still invite cartographers when they want to present information to end users in an attractive way. This is because digital cartography enables many new features like context awareness, adaptive environment, sensor integration etc, which if not used correctly may result in misleading maps. “It is necessary to use basic cartographic rules, logical systems of map legends and symbols, methods of cartographic generalisation etc because maps should model reality and deliver quality information for solving problems,” he emphasises. Another task for cartographers is how to control and measure the quality of maps; from such understanding they can select and develop appropriate approaches for map presentation meeting the needs of the public.
The term ‘ubiquitous mapping’ is not homogeneously used globally, says Gartner. It is a term primarily used in East Asia and also partly in Europe. The term might be best to describe the ubiquitous availability and accessibility of maps and mapping services, and its development seems to be continuing. Cartographers are contributing to conferences and research under terms like location-based services, mobile cartography etc as well. However, the attribute ‘ubiquitous’ as such is one of the five main attributes of modern cartography, the others being real-time, media-adequate, personalised and well-designed. Parsons is of the opinion that it is hard to argue that maps are not ubiquitous now, but maps are possibly being used when their use is not actually useful. “If there is no geographic pattern to an activity or phenomena, and it does happen sometimes, you don’t need a map!”
The term ‘ubiquitous mapping’ comes from the idea of pervasive computing and pervasive maps, which is more oriented for mapping supported by sensors. Cartographers recognised that nowadays intelligent access to databases and interactive user support can be used not only for the location of suitable maps on the Internet, but also for map creation (art) and modification according to specific and individual requirements of users. Instead of just using maps created by someone else in advance, these new research technologies allow individuals to use cartography interactively, on the basis of individual user’s requirement, to study and present spatial information.
According to Konecny, ubiquitous mapping “is one of the biggest challenges of contemporary cartography and geoinformatics and arose because of the perceived advantages of an ‘information society’. Mapping should be done by everybody, anytime, everywhere with all possible technological tools.”
Konecny also refers to adaptive cartography as the automatic creation of correct geodata visualisation with regard to situation, purpose and the user. Adaptive maps are maps in the conventional sense — they are correct and well-readable medium for transfer of spatial information. The user controls map modifications indirectly via modification of context. The term context refers to the following set of characteristics:
- Who is the map reader: information on abilities of the users to read maps, their visual preferences, level of knowledge and/or education. This information forms the user profile.
- What is the purpose of the map: information on solved problems, spatial extent of the problem and information on hierarchy of content items depending on the given problem.
- Where is the map to be used: information on place, time, orientation and natural conditions influencing map perception e.g. light conditions.
- What is the device displaying the map: set of information related to parameters of the display, transmission capacity and software characteristics of client application.
The main objective of visualisation is to be able to present the salient features of a spatial context in a manner that is easily and quickly assimilated by a human being. The problem lies in the volume of information and the consequential information overload leading to the user, missing vital aspects of the analysis. In a dynamic situation, the problem is compounded by the real-time stream of data that has to be rapidly surveyed for patterns and anomalies. The task of the cartographer is to select and visualise the analysis results in a manner that holds the users attention and effectively highlights the findings. The visualisation designer has to take into account aspects of human cognition and decide on the quality and quantity of the interactivity allowed to a user such that he or she is not overwhelmed.
Gartner points out that cartography can be understood as “communicating spatial information”. Thus the communication paradigm — sender codes information, transmits it, receiver decodes information — is still applied in cartography. This has three dimensions. The syntactical dimension in which we define the code in terms of symbols, sizes, colours, etc. Cartographers are very good in this. The semantic dimension is one in which we try to incorporate the meaning of the information using either self-explaining codes or a legend. Cartographers are very good in this as well but, could do better in the arena of dynamic visualisation and multimedia. Finally, we have the pragmatic dimension which ensures that the communicated information is relevant to the receiver through context-modelling and personalisation.
Konecny states during a project on dynamic geovisualisation for creating adaptive and context-based map concepts, his team discovered and confirmed cognitive style of users with different skills, abilities, education and cultural background and different ages. “We found and confirmed the importance of the idea of cognitive style.” Cognitive style or ‘thinking style’ is a term used in cognitive psychology to describe the way individuals think, perceive and remember information, or their preferred approach to using such information to solve problems. Two main streams of cognitive style are holistic or analytic ones and could give practical inputs to cartographers. Some find use of maps better, others prefer orthoimages. This can play an important role especially in disaster situations. But research is in the early stages and needs more deep analysis, cooperation with more disciplines at an international level.
Education and research
As geospatial systems become ubiquitous, there is a clear need for the education and research field to gear up to meet the challenges. According to Kraak, geographic education at school levels gets less space and introducing new technology has a price attached. However, at universities it is in the process of getting integrated with the curriculum. Gartner points out that there are several steps taken by the ICA, such as liaising with organisations like IGU; an Education Commission, which develops new instruments; an Open Source Commission, which has helped to establish various Labs under the term ‘ICA-OSGeo Labs’ which offer webinars and free education programmes; outreach programmes and capacity building programmes to name a few.
Konecny mentions the TEMAP project — Technology for Discovering of Map Collections — which aims primarily to focus on laymen to help them catalogue old maps and to design tools enabling them to use geographic information contained in scanned raster files. One of the very popular results for wide publicity is crowd pilot which aims at georeferencing of old maps available at www.staremapy.cz.
In terms of research areas Kraak identifies the understanding how maps work in highly interactive environments as important. Gartner identifies developing human-centred cartographic theory and practice; the use of cartographic displays for spatio-temporal inference and decision-making; construction and use of cognitively adequate and perceptually salient visual displays of geographic information; empirical geovisualisation design research on 2 and 3D, static, animated and interactive, virtual and immersive and mobile displays; the application of cognitive theories and methods to understanding usage of visio-spatial displays and tools for inference and decision-making which includes mental maps, space-time behaviour, navigation, etc; the application of visio-spatial displays and tools to understanding spatial cognition — spatial reasoning, inference and decision making with visio-spatial displays and tools and cognitive principles supporting human-visualisation interaction research.
“There are also other areas of cartographic research like perception of graphic variables, improvement of understanding of space, the role of short-term memory, understanding of the uncertainties of cartographic communication, and the adaptation to the use of maps,” adds Konecny.
Interactive maps are now being widely used not only by the academia but also by common people through a range of media and technologies in a variety of disciplines and application scenarios across the globe. But interactivity is always in direct conflict with privacy. So how can these issues be addressed?
Some like Gartner feel when we are interested in personalisation, context-modelling or simply see cartography as a process of communication between sender and receiver we get into the question of privacy. “However, the privacy question is an issue of society. We need to make society aware and eventually offer opinions, but the decisions need to be taken by all of us.”
Parsons feels that customised or personal maps now often delivered to mobile devices require some guidelines or ‘best practices’ to be followed. There is widespread recognition that sharing a user’s location with a service provider is sensitive and must only be done with the users’ explicit consent and control. By sharing location numerous benefits are then of course available, and if the user is also willing to share other personal information with the service provider, applications that for example can automatically re-route a traveller based on current traffic or transit status and their calendar become possible.
Google Glass is an example of immersive visualisation, and is a new way of visualising cartographic information
The largest business opportunities lie in disruptive technologies, according to Mak Poh Fatt of Faro Singapore. For example in 3D imaging technologies, the concerns are affordability, complexity of use, integration of technology and insufficient information about the technology and its applications. He feels that the solution lies in innovation to provide low-cost, simple mobile solutions with a degree of automation. The Cloud is one of the ways forward as it offers minimal set up and maintenance effort, best possible security level, scalable infrastructure, persistent measurements and annotations and support for mobile devices.
Social media is a huge platform for user interactions. Therefore, is there a social media context in geovisualisation? Parsons feels that though social interactions often have a geographical context to them there is nothing really profound by way of visualisation. However, Gartner feels that cartography is already in to social media. There are numerous applications being developed out of Twitter, Flickr, Facebook or any other social media platform that offers APIs. He anticipates that this is the start of a massive development. Data mining, collaborative filtering, personalised recommendations or simply context-dependent personalised maps will be most common in years to come.
The future developments in cartography will continue in the areas of automation, visualisation, and display of real-time information, feels Watkins.
Automation has driven the proliferation of maps throughout the world, and created efficiencies in map and data production, making it possible to provide more current maps. Visualisation improvements will continue to take place, giving us better ways to graphically display map information. There will be continued progress in photo-realistic 3D, for instance. Hardware and processing speeds will continue to improve, creating possibilities for new ways to consume and interact with maps. The Google Glass, for example, provides a new way of visualising cartographic information while at the same time allowing the user to see the surrounding environment. Also, flexible, foldable, materials are being developed for electronic display that will make large format, portable interactive maps possible in the future.
More and more sensor data is becoming available in real time. Huge amounts of streaming data must be quickly consumed, summarised and displayed in understandable ways. Future developments in cartography will need to take into account these vast amounts of constantly updated information. This not only presents an entirely new set of challenges for the cartographer, but also provides new opportunities for visualising the world around us.
Geospatial technology has made it possible for maps to go beyond the descriptive to the predictive. As more and more data is available, it can be used to visualise trends, increasingly in real time, and this can be used to make better predictions. The idea of Geodesign follows this premise. Geodesign involves planning future development using predictive models, taking into account many factors to make suitability predictions, thus allowing us to make better decisions. Cartographic communication is essential in understanding and interpreting the results of this predictive analysis, making it possible for decision makers to act as effectively as possible as they tackle the challenges we face today.
Link of Interest: http://www.wimp.com/europemap/