untitled Cartographic traditions and the future of digital maps: A Finnish perspective NIINA VUORELA, CHARLES BURNETT AND RISTO KALLIOLA Vuorela, Niina, Charles Burnett & Risto Kalliola (2002). Cartographic tradi- tions and the future of digital maps: A Finnish perspective. Fennia 180: 1–2, pp. 111–121. Helsinki. ISSN 0015-0010. In this article, we examine changes and trends in Finnish cartography using examples from the history of map-making and from our own empirical land- scape research. Three major issues of data sources, combined uses of spatial data, and interactive spatial information services are emphasised. The Finnish map-making tradition illustrates well the close linkage between cartography, society, and surveying techniques. Sweden and Russia influenced the map- ping traditions until the twentieth century. Geographers published the world’s first National Atlas in 1899. Comprehensive state-run mapping has character- ised Finnish cartography from the early twentieth century onwards. New tech- niques and data have modernised map-making, but the recent convergence of automated positioning, wireless communication, and digital spatial informa- tion handling are currently creating previously non-existing ways of cartograph- ic communication. Spatial or geographic information systems enable the col- lection, management, and analyses of spatial information for a wide variety of uses. The contents and means of delivering cartographic information to the public is broadening and new digital spatial information services are designed for resource management, retailing, and tourism. A critical examination of data quality, visualisation, and the ethical and sociological consequences of these trends are needed to sustain the Finnish cartographic tradition in the Informa- tion Age. Niina Vuorela, Department of Geography, University of Turku, FIN-20014 Turku, Finland. E-mail: niina.vuorela@utu.fi Introduction Map users in the twenty-first century will witness changes in the visualisation and application of spa- tial information. The convergence of enabling tech- nologies, such as automated positioning, wireless communication, and digital spatial information handling, will permit the development of previous- ly non-existing implementations of cartographic communication (Burnett & Kalliola 2000). The adoption of cartographic developments in main- stream society must not, however, be viewed as a simple deterministic cycle of technological won- ders begetting societal benefits. Much of the pub- lic discourse regarding future cartographic tools is filled with uncritical enthusiasm. Graham (1998) cautions against such stary-eyed utopianism. The role of the scientific community, including cartog- raphers and those working with the perception of space and the human context, is to critically in- vestigate the interactions between digital spatial information and society (Kalliola et al. 2001). The evolution of cartography as spatial informa- tion communication is driven by changes in so- cietal need and in the state of measurement and presentation techniques (Robinson et al. 1995; Keates 1996). Throughout their evolution, maps have increased their agency for conveying infor- mation and expanded into new societal uses that include – but are not restricted to – teaching, geo- politics, military needs, resource exploitation, lo- gistics, marketing, and tourism. While the neces- sity to communicate cartographically in these fo- rums will remain, developments in digital carto- graphic communication will add new rings (so- cio-technical artefacts) to an already existing chain of past advances in map-making (Mellaart 1963; Bagrow & Skelton 1985; Campbell 1993; 112 FENNIA 180: 1–2 (2002)Niina Vuorela, Charles Burnett and Risto Kalliola Dorling 1997). In this article, we take the perspec- tive that these new artefacts will draw heavily from conventions established through the history of cartography while adding unforeseen models of cartographic information exchange. Undoubtedly, digitalisation has already induced important changes in cartography. In the late twentieth century, the introduction of remote sensing and Geographic Information Systems (GIS) were quickly adopted for map production tasks worldwide (Morrison 1989; Artimo 1994). GIS has made it easier to reproduce, update, cus- tomise, cross-analyse, and perform a variety of other spatial operations and analyses (Aronoff 1991; Laurini & Thompson 1992; Jones 1997; Clarke 2001). In the majority of these applica- tions, however, GIS only help users gain speed and efficiency in their work. Thus, the view of GIS as a revolutionary approach to cartography is part- ly a myth. More specifically, it is a myth, if devel- opments in society are ignored in favour of one- sided views of technology ‘impacting’ society. The work of the Finnish geographer J. G. Granö (1930, 1997) is an illustrative example of a pre- digital cartography, where a highly organised landscape analysis system was developed half a century before the introduction of the GIS tech- nology. It is worthy of note that even with the ‘ad- vanced’ GIS tools, good geographic analysis has remained true to the Granö tradition. Granö’s syn- thesis was a delicate process of data combination, fulfilled with a careful synthesis and analysis of landscape regions. The presence of this care and attention in contemporary cartography, whether in creating tourist information systems or landscape- ecological assessments, lies at the core of bal- anced use of modern geographic information. In this article, we examine cartography as a co- evolutionary process of society and technology. We use examples from the history of cartography in Finland and from our own empirical landscape re- search on the island of Ruissalo in southwestern Finland (Vuorela 2000, 2001; Burnett & Toivonen 2001). We examine five themes: the history of car- tography in Finland; new digital data sources; is- sues arising from data combination; new digital cartographic services; and digital map reliability. Landmarks of cartography in Finland Finland can be considered one of the frontline countries in the evolution of map-making over the past few centuries. This situation results partly from a fruitful mixture of versatile map produc- tion efforts and traditions drawn from two neigh- bouring countries, Sweden and Russia. They im- pacted Finnish map-making tradition from the late seventeenth to the twentieth century. More pre- cisely, the national trends in map-making illustrate well the close linkage between cartography, so- ciety, and surveying techniques. The centuries-long influence of Sweden result- ed in a strong map-making tradition in Finland. The introduction of triangulation and plane table meth- ods to map-making in the seventeenth century im- proved the accuracy of maps greatly (Niemelä 1984: 2). As Finland was economically and strate- gically important to Sweden, tax collection and implementation of land reforms were planned with the aid of the early maps. The Swedish King initi- ated mapping in Finland during the 1630s. These mappings were targeted to measure the extent and quality of arable land and meadows and became known as geometrical or geographical maps (Lön- borg 1901; Johnsson 1965). During the period when Finland was a Grand Duchy under the Russian Empire (1809–1917), new trends and styles were adopted in map pro- duction and design. The Land Survey administra- tion still developed along the lines of the Swed- ish system, however. Overall, the progress of car- tography slowed down during these years, but military mapping was carried out both by Finns and Russians (Niemelä 1984: 2). The previous mapping by the Russian Topographic Service (in eastern Finland, part of Russia before 1809) was extended to the rest of the country, mainly south and west. Between 1870 and 1917, a series of maps was created to cover large parts of Finland at scales of 1:21 000 and 1:42 000 (Gustafsson 1932, 1933: 86–88). Finland thus became topo- graphically described relatively early compared to other European nations (Paulaharju 1947; Niemelä 1984, 1998). These Russian Topograph- ic Maps constitute important documents of the time, as they include typical features of the present-day basic maps, such as arable land, meadows, roads, and buildings in addition to landforms. Maps also became a valuable means of char- acterising Finnish identity and nationality at the end of the nineteenth and beginning of the twen- tieth century. As an effort of geographers, the first edition of the National Atlas of Finland was pub- lished in 1899. This first national atlas in the world FENNIA 180: 1–2 (2002) 113Cartographic traditions and the future of digital maps represented Finland through a variety of themat- ic maps showing characteristic features of the country (Niemelä 1982: 170). Immediately fol- lowing Finland’s independence in 1917, map pro- duction developed into an organised, state-run mapping system, where the Military Topographic Service carried out military mapping and the Na- tional Land Survey of Finland (NLS) other map- pings (Lyytikäinen 1983: 448). Until the 1940s, two main products, the Parish Map and the Topo- graphic Map, were produced. Production of the former had started already in 1825. By the time the program terminated in 1950, these maps had covered approximately 27 percent of Finland (Niemelä 1998: 27). Due to their long production history the parish maps vary, reflecting changes in printing techniques and visualisation. Topo- graphic maps were produced between 1917 and 1947 (Niemelä 1998: 40). A significant change in the underlying mapping techniques occurred in the 1930s. Instead of the plane table measurements, mapping was now based on aerial photography. The role of remote sensing has been central ever since. As remote sensing gained wider use in map production, the National Land Survey renewed its maps during the 1940s. The parish map and the topographic map were combined to produce the Basic Map of Finland (1:20 000) (Niemelä 1984). Basic maps included general information about the landscape and were designed for a large audience of both professional and accustomed users (Niemelä 1998: 54). The mapping was based on aerial pho- tography acquired at a scale of 1:10 000 and car- tographic visualisation was based on a six-colour printing. The social and technical change that drives car- tographic evolution is propelling changes in car- tographic models, methods, and maps, and putting pressure on large governmental organisa- tions such as the National Land Survey of Finland. The implementation of computer technology and new storage systems for geographical object lo- cations and attributes have diversified the possi- bilities of designing and making maps. Instead of storing spatial information statically on a printed map, location-related data is increasingly stored in digital Geographic Information (GI) databases and further visualised in printed form or as com- puter display maps (Artimo 1996). The 50-year-old Finnish Basic Map production system has had to adjust to the new developments (Artimo 1996; Niemelä 1998). From the 1970s onwards, the map production chain has been changed into a GI-based storage and visualisation system (Niemelä 1998). Currently, the production of the Basic Maps relies on digital databases (top- ographic database, map database) with thematic or attribute information attached to vector format geographic objects. The accuracy of the topo- graphic database varies between 1:5 000 and 1:10 000. They are updated every 5–10 years, ex- cept for roads, which are updated every year (Niemelä 1998: 134). Due to these technical changes, the NLS is pro- viding customers with new product types and map visualisation options. The Basic Map can be purchased either as a printed map (in Finnish, maastokartta) that introduces a different set of col- ours and symbols compared to the previous Ba- sic Map, or as a digital vector format Basic Map Dataset (maastotietokanta), which can be adjust- ed to a customer’s needs. These data are availa- ble for manipulation in the end-user’s computer system. Similarly, Basic Maps can be purchased in raster format (Perus-CD), where different the- matic information layers can be explored. New cartographic data sources and processing Developments in the technology of new data sources and processing methods tend to be adopt- ed into mainstream use approximately a decade after the first prototypes are built. Recent improve- ments in data collection technology in the fields of surveying and remote sensing are changing car- tography. In survey science, satellite-based global positioning systems (GPS) and hand-held infra-red (IR) electronic measuring devices have increased the accuracy of traditional surveying methods. Expensive ground-based GPS survey systems that make use of differential GPS with kinematic and ionispheric corrections can claim centimetre ac- curacies. GPS systems are also used in the crea- tion of digital orthophoto mosaics. The accuracy of consumer GPS systems improved in May of 2000, when the US government removed the sig- nal accuracy degrading “selective availability.” GPS instrument manufacturers now produce in- struments that track both the American GPS satel- lites and the Russian GLONASS satellites. Future reliance on these foreign systems may decrease at least in Europe, where the national governments within the European Union have recently pledged 114 FENNIA 180: 1–2 (2002)Niina Vuorela, Charles Burnett and Risto Kalliola support for a European satellite navigation system called Galileo (Ochieng & Chen 2000). In the field of remote sensing, several new dig- ital satellite, airborne, and terrestrial systems have been, or will soon be, deployed. There are tens of satellite-based systems already in operation that can provide useful data for cartographers. These include Landsat and SPOT with nominal ground resolutions between 10 and 30 metres, and IKONOS and EROS-1 at 1 metre. Some 30 satel- lite launches are planned for 2001. They focus on earth observation missions, most being military or weather satellites. Several, however, will collect data of interest to the civilian and governmental mapping markets. These include the first sub-me- tre resolution satellite-based sensor (Quickbird2, 61 cm) and the first satellite-based imaging spec- trometer (ARIES-1). A full summary of space and remote sensing research and industry projects conducted under the recent five-year TEKES (Na- tional Technology Agency) Globe 2000 and Space 2000 programs is given in Fabis and Gudmand- sen (2001). Fabis and Gudmandsen (2001) esti- mate that approximately 20 research organisations and groups, as well as several companies, partic- ipate in research activities that focus on increas- ing Finnish space-technological capacity and re- mote sensing expertise. Airborne data acquisition systems that are being developed or are in opera- tion in Finland include digital cameras/semi-au- tomated image mosaicking (Holm et al. 1999), the AISA Airborne Imaging Spectrometer (Okkonen et al. 1997; Mäkisara et al. 1997), an airborne SAR (Kallio et al. 2000; Martinez et al. 2000), and a number of companies and government agencies that produce digital orthophotos (e.g., FM-Kartta Oy, NLS of Finland). In the terrestrial mode, vid- eo cameras are being installed and connected via telecommunications networks to provide real- time geographic information via the Internet (Trav- el… 2001). The range of data processing tools and meth- odologies being developed in Finland is enor- mous and only a sample is mentioned here. The modern analogue to the topographic map is the digital elevation model (DEM), created either by stereo-photogrammetry (using digitised aerial photography) or, more recently, directly from LIDAR (LIght Detection And Ranging) height measurements and scanning laser (see Samberg 1997). DEMs have many uses, from modelling landscape processes to image drapes, but perhaps the most useful, cartographically, is the creation of geometrically corrected aerial photography, or digital orthophotography. Orthophotos can be used like maps, but they contain more informa- tion as they have not been generalised. The NLS of Finland creates its DEM from the contour lines and coastline elements of the Basic Map by tri- angulated network interpolation into a grid mod- el where the grid cell size is 25 x 25 metres. Mosaics of aerial photographs and video or dig- ital camera data are increasingly used as GIS base maps. The mosaicking of images and subsequent atmospheric correction procedures are challeng- ing data processing steps, which Finnish scientists have explored thoroughly (Holm et al. 1999; Pel- likka 1998; Pellikka et al. 2000). Holm (1995) has taken the processing of stereo imagery a step fur- ther, reconstructing complete image models (buildings and trees as well as the elevation mod- el). Elevation models may also be produced by using scanning LIDAR (scanning laser) data. Us- ing a combination of aircraft XY and altitude po- sition (GPS) and attitude (GPS or attitude sensor), LIDAR data can be accurately corrected to ground heights above sea level. The difference between the NLS DEM and the LIDAR DEMs is mainly in regard to resolution: the latter is able to provide sub-metre 3D surveys. Scanning LIDAR data has been used in the modelling of buildings, utility line corridors, and trees (Ziegler et al. 2000). A great deal of mapping research in Finland has focused on delimiting the boundaries and deriv- ing the attributes of trees and forest stands from remotely sensed data. Examples of this type of analysis include the k nearest neighbor (KNN) method of forest inventory developed at METLA (Finnish Forest Research Institute) (Tomppo 1996, 1998), individual tree crown mapping, and auto- mated stand mapping using segmentation. CD- Figure 1 shows two examples of these applica- tions. In the first example, the mapping of indi- vidual tree crowns is intended as the first of sev- eral semi-automated steps that will take combi- nations of remotely-sensed and field data and pro- duce forest habitat compartment maps (Burnett & Toivonen 2001). The second example of forest mapping shows 1.6-metre ground resolution AISA data of a forest scene segmented into polygons in a semi-automated fashion using algorithms devel- oped at METLA (Haapanen & Pekkarinen 2000). In summary, several trends in new data sourc- es and processing can be identified in Finland. Firstly, improvements in technology (optics, elec- tronics, computers, etc.) are producing new geo- FENNIA 180: 1–2 (2002) 115Cartographic traditions and the future of digital maps graphic information data sources. Depending upon their cost, some of these sources are becom- ing available to the general public (e.g., terrestri- al video data and digital map data from the NLS). Secondly, remotely-sensed images are used in- creasingly in combination with vector-form map data. Most of these data sets, however, are either too expensive or require a large amount of expert processing. Thirdly, development of data process- ing methods and applications are related to in- creases in the number and quality of data sourc- es. This development is cyclical in nature: appli- cations create markets (demand), which drive the development of new data sources. These, in turn, create application potential. Finnish scientists are leading the way in many processing research themes, especially in those related to mapping. Possibilities for flexible data combination and visualisation It would probably have been unimaginable to a seventeenth-century surveyor to see the potential that digital systems give to the usage of old maps or to realise the value of these maps as informa- tion sources to present-day researchers. Especial- ly for landscape researchers, old maps are the pri- mary representations of the past features and char- acteristics of the landscape. Maps can be used in a broad spectrum of landscape research, from ec- ological and geomorphological to political, so- cial, and cultural topics (Dickinson 1979; Spor- rong 1990; Roeck Hansen 1996). Maps have been a major source of inspiration to study landscape evolution from the historical times to the present-day. Even though maps are widely used in landscape research, in few coun- tries the variety and availability of maps are as good as in Finland. Due to the rich history of map- making in Finland, the possibilities for landscape change research are manifold. Further, combined uses of maps have expanded, since digital tools enable transformation of old maps into digital spa- tial data. Similarly, demands for a careful assess- ment of their information contents have increased. Most often, the transformation of printed maps into geographic information requires unique so- lutions that differ from those used in remotely sensed data transformations (Vuorela et al. 2002). Visualisations in CD-Figure 2 illustrate how com- binations of multi-temporal GI, originating from different sources (e.g., remotely sensed images, old maps, written records), can reveal interesting links between valuable landscape features and other environmental factors, such as topography, soils, and land uses (see Vuorela 2000, 2001). Combining landscape information to explain distribution of oaks The current distribution (1998) of old oaks on Ruissalo Island in southwestern Finland (CD-Fig. 2A) is fragmented within the woodland areas. This pattern is very difficult to explain unless both land use history and natural edaphic factors are taken into account. Most of the oaks seem to be con- centrated on slopes (CD-Fig. 2B) and near edges of the wooded areas (CD-Fig. 2A). The lack of oak both from the upper parts of the hills and from low clay areas can be explained in the following manner: soil conditions affect species distribution profoundly. Oaks do not grow on the tops of the hills because they are too rugged and nutrient- poor. Edaphic conditions thus determine the up- per margin of the oaks. Similarly, oaks do not ex- ist in clay areas, because these areas have been in agricultural use (arable land, meadows) con- tinuously after their exposure from the sea (land uplift). Therefore, the lower margin of the oaks shows adaptation primarily to land use changes. The key to explaining the distribution patterns of oaks is related to the natural site conditions and land use changes of the slope areas. Slopes can be identified both as edaphic and land use tran- sition zones. Here, clay deposits change into rug- ged bedrock areas, but the change is not always abrupt and different types of shallow moraines can be identified within the bedrock areas. These often provide suitable sites for an oak to grow. Slope areas have, however, experienced fluctua- tion and shifts in land uses over time. Oak was a typical tree on meadows and pastures that were characteristic to gentle slopes (CD-Fig. 2C) (Ekqvist 1846). Since the end of the animal hus- bandry agriculture, some of the wooded mead- ows were cleared for agricultural land. Others re- mained wooded and changed gradually into dif- ferent types of woodlands, which have been light- ly managed over the last century (CD-Fig. 2D). Altogether, the present-day distribution of Ruis- salo’s large oak individuals reflects the natural potential of the sites convolved with land use var- iations at work over hundreds of years. In other words, oak woodlands reflect the diversity of nat- ural and human induced environmental factors 116 FENNIA 180: 1–2 (2002)Niina Vuorela, Charles Burnett and Risto Kalliola that have evolved and changed through time (see Vuorela 2000, 2001). The Ruissalo example shows how maps provide a unique information source for reconstructing, describing, and understanding the development of landscapes through time (Tollin 1991; Skånes 1996; Cousins 2001). For many applications, there are needs to combine data originating from differ- ent sources (e.g., information services). Visualisa- tion of cartographic (spatial) information is based on the combination of data within the shared co- ordinate reference system. The digital data format makes combining data technically relatively easy. Many data types (like old maps), however, require specific solutions to data format transfer and spa- tial adjusting (accuracy). Similarly, different digit- al media (e.g., mobile phones, laptops) require unique solutions to visualise data. Further, as in- formation content and its representation are so var- iable in different data sets, conversion into GI should be a selective process where spatial, tem- poral, and thematic data properties are evaluated (Faiz & Boursier 1996; Richardson 1996). An il- lustrative example of the possibilities and restric- tions of data combination can be found from Vuo- rela et al. (2002), where a collection of nine maps were evaluated from the perspective of combin- ing them in a GIS to perform landscape change analysis. It is possible to improve the usability of GI originating from many data sets, if an attempt is made to control and adjust the spatial, themat- ic, and temporal accuracy of data sets. Only when data combination is properly made, the provision of good services that are based on digital location- al multi-source data is possible. Expansion of the digital media: interactive spatial information services Digital map production has influenced the con- tents and means of delivering information servic- es to the public. Geographic information servic- es are being promoted as ‘useful’ in several fields of the information society, including resource management, retailing, and tourism (Kasvio 1997). There is also an increasing demand for flex- ible access to digital geographic information with- in Finnish government and corporations. As maps often form a backbone to an information system, printed maps have been either accompanied or replaced by digital spatial information accessed through several types of digital interfaces (e.g., mobile phones, the Internet, information kiosks). By augmenting or replacing paper versions, the model of map use has changed from static to in- teractive. This means that geographic attribute data (including spreadsheets, 3D animations, and video) are accessed through a dynamic interface. Contemporary developments in Finnish map- making show an increase in the number and type of digital spatial registers and customised map services. Many municipalities and marketing companies have established their respective map databases on the Internet, with specifically de- signed search, query, and multimedia functions. New data is now collected directly into digital form (e.g., high-resolution airborne remotely sensed data) and existing analogue data sets are converted into digital data formats. For example, the seventeenth-century geometric maps of Fin- land were digitised in a shared effort by the De- partment of History of the University of Jyväskylä and the Center for Scientific Computing (CSC). The result can be accessed through the Internet (www.csc.fi/maakirjakartat/). Further, the “Kartta- paikka” searchable Internet map database, imple- mented by the National Land Survey in 1996 (www.kartta.nls.fi/), enables a user to explore maps from any region in the country up to the scale of 1:50 000. Finer-scale maps are available upon a prepaid session fee (Niemelä 1998: 92). Digital GI is also distributed in form of digital at- lases on compact disks and sold to clients with map-reading software. Digital atlases are also made GPS compatible. Among the first products to appear in the market in 1998 and 1999 for dig- ital navigation were two raster-based road maps (GT-Suomi, CD-Tiekartasto) and two vector-based route planning systems (Genimap AT and YT). An Internet-based search engine for environmental literature in the Archipelago Sea Region, using a map interface, has been established recently (Tolvanen 2001). When access to the Internet through wireless devices comes into mainstream use, data is not only explored in an interactive manner, as with stationary Internet access devices, but the user’s location becomes an input variable. Cartograph- ic services that are keyed to this location variable have been dubbed “Location Based Services” (LBS). The key to LBS is then to facilitate the ac- cess to selected information contents, appropri- ate in context (users requirements), location, and time. For example: While the traditional use of city maps is based on reading their information, FENNIA 180: 1–2 (2002) 117Cartographic traditions and the future of digital maps future mobile devices can transfer the same mes- sage in many different ways. It is often unneces- sary to see the information plotted over a map if the same message could be delivered otherwise, for example, by spelling out: “Just continue walk- ing along this road.” Even in these connections, however, the primary function of the automated system is based on digital GI and coordinates, maps, and algorithms concerning their relations. If LBS systems, such as personal navigation, penetrate into everyday use, future travellers may load their portable communication platforms with map information and, having set their individual content preferences, interact with a four-dimen- sional data set (geographic dimensions and time) as they move through the physical world. While moving, navigation systems with real-time GPS linkages will then activate the relevant informa- tion based on the position of the user and the pre- defined sites of interest. With the eventual intro- duction of the third-generation (3G) mobile phones and hand-held computers, all the neces- sary equipment will be carried along easily (Kal- liola 2000a). Working LBS prototypes already ex- ist. Much development is still needed on techni- cal issues (such as data, software, and communi- cation standards), but some of the most essential research concerns the sociological, ethical, and legal issues accompanying LBS systems. We outline the following model for a trav- eller’s information service of the early twenty-first century. The system envisioned is a wireless nav- igation and information assistant running on a hand-held computer and linked to a GPS, using third-generation mobile phone networks for GI data transfer (CD-Fig. 3). This concept is based on linking the real-time position of the user with re- mote dynamic GI databases that contain contex- tual data on the user’s environment. For the sys- tem to function, all relevant data sources relating to the same area (ranging from historical maps to accurate satellite imageries) are first made com- patible with each other and stored in a distribut- ed computer environment (Krisp 2000; Lähde 2001). Their substance contents can be linked with each other interactively to fit ideally to the actual needs of information, reaching hundreds of years backwards in some areas. Query results can be visualised in a number of ways, including cartographic representations, virtual landscapes, and multimedia solutions. The interface provides access to common mapping queries, such as search, location finding, zoom, pan, and help. In the first example (CD-Fig. 3A), our hypothetical guide shows the position of the user vis-à-vis a three-dimensional visualisation of the landscape from 1690 and 1998 data sources. In the second example (CD-Fig. 3B), the guide shows three vis- ualisations of park management effects in the for- est at the user’s position (Tahvanainen et al. 2001). In the light of these trends, the ways that spa- tially organised geographic information is stored and delivered to the people has expanded during recent years. The traditional cartographic commu- nication model is becoming more dynamic. GI systems that enhance a user’s experience in reali- ty are being developed. In these systems, real-time data from sensors are incorporated into maps (Raper 2000) and reality is supplanted by letting a user explore GI-based virtual realities (Rak- kolainen et al. 1998). Digital databases have in- troduced a multifaceted model of cartographic reality, where information is stored both as loca- tional and non-locational data using either vec- tor or raster data structures (formats). Locational data that refers to coordinate information is used in drawing and displaying geographic features, while non-locational information (i.e., attribute data) of the mapped features forms ‘the message’ on the interface (Robinson et al. 1995: 169–177). As a result, the three basic map attributes of scale, projection, and symbolisation (Monmonier 1996) are being changed and adjusted by the map us- ers when creating their own maps. Maps are also becoming not only abstractions of a physical re- ality, but spaces that the users can explore. As an example, mapping based on remote sensing may employ image and sound (and eventually smell and touch) recorders with real-time links to the map interface. With such a system of sensors con- nected to a location-aware mapping system, the ‘map’ becomes, in actuality, an image (or sensu- al representation, if other senses are evoked) of the user in the ‘mapped’ environment (Burnett & Kalliola 2000). Such maps may update in real time and will include over-laid directional and velocity vectors or other data. Digital maps – reliable maps? In parallel with the trends of expanding carto- graphic communication in the digital realm, pres- sure is increasing for (1) a critical examination of methods for increasing and communicating data quality; (2) an improved understanding of the in- 118 FENNIA 180: 1–2 (2002)Niina Vuorela, Charles Burnett and Risto Kalliola teractive spatial information exchange model; and (3) ideas on how to address issues of privacy and data ownership. Burnett and Kalliola (2000) ar- gued that the changes we now engender and wit- ness are so unique that they herald a new era in spatial information communication. These chang- es relate to how our societal changes are gener- ating demand for digital technologies and to sci- ence and technological actors that respond with rapid artefact development. Digital maps, created from GI databases, suffer from the same malaise as printed maps: they may be as inaccurate and misleading as any printed map despite being products of the most recent technology. These problems do not necessarily relate to their digital format as such, but, rather, to our weakness in handling and using digital data and – often – lack of product-related information. Some fundamental principles in mapping, such as geodesy and survey technique, run through all historic cartographic production modes and have remained unchanged despite the rally of space- borne remote sensing and automatic positioning techniques. Three major causes for inaccuracy and poor quality in digital maps can be ad- dressed. These problems arise with the combina- tion of digital data, the visualisation of geograph- ic data, and the production of digital maps in a market economy. Firstly, since GI is increasingly stored in digital and geo-referenced formats, data combinations can be created in a very flexible manner, at times creating poor-quality maps. Even though much of the confusion in data properties can be tested, modelled, and documented, there will be increas- ing heterogeneity and inconsistency between GI datasets. This is due to combining spatial infor- mation produced by different disciplines and or- ganisations. Thus, there is a need to assess the nature of spatial information in a broader sense to be able to understand spatial and temporal scales and relations of GI. Much discussion has been involved around metadata (structured infor- mation on the content and format of datasets) standards and GI data structures (ISO/TC 211 Ge- ographic Information/Geomatics Metadata). Well- documented and described metadata about GI is essential and should provide the user with rele- vant knowledge of the data characteristics (such as its processing history) and help in the task of evaluating and combining GI from different sourc- es. Metadata, however, is not necessarily effec- tive enough to overcome the ontological differ- ences of spatial information coming from differ- ent sources (Bowker 2000). Secondly, a variety of cartographic visualisation techniques, such as the use of colors, is directly applicable in the digital realm. Yet, novel solutions are needed to better meet the restrictions and pos- sibilities computer and GSM phone displays and new printing tools bring about (Kraak & Ormeling 1996). To design a map layout requires knowledge of how colour combinations can best be related to the message the map is supposed to deliver (Monmonier 1996; Nurmi 1999). As non-profes- sional map makers (who may be unfamiliar with basic visualisation rules) can now access digital data (e.g., from a digital topographic database purchased from the NLS), there is a risk that map products will perform poorly. On the other hand, when professionals have access to a digital GI database, incorporation of new information and changes to the visual appearance of the map can be tested and implemented efficiently. Of course, society must accept or reject these changes. It was no surprise that when the visuali- sation of the Basic Map of the Finnish NLS was changed during the 1990s, much criticism was presented towards accuracy and usability of these maps compared to the previous map versions. The critics focused especially on cartographic visual- isation and applicability of the natural landscape features (e.g., topography, soils, forest cover). For example, no longer was it possible to distinguish pine from a spruce forest, one of the most impor- tant tree species indicating differences in soils and moisture (Shemeikka 1998; Kallio 1999). As these visualisation changes were made simultaneously with digitalisation, much of the blame was put on these production changes. Digitalisation general- ly widened cartographic information, however, since no longer are the printed and the digital product the same. In other words, much ancillary information for each Basic Map lies in the digital database and only a selection is visualised on a printed map. These differences between digital and analogue maps are not necessarily clear to map users. Once a single digital product has been criticised, the criticism is misdirected to digitali- sation in general. The third major source of map quality problems is the effect of GI, GI tools, and digital communi- cation modes (Internet, CD) on marketed prod- ucts. Some alarmingly unsatisfactory digital map products and freeware base maps have been dis- tributed. They have already damaged the ‘image’ FENNIA 180: 1–2 (2002) 119Cartographic traditions and the future of digital maps of, or public trust in, digital products in general. If maps contain obvious location, naming, and content errors, like in the case of some Internet map sites, false data get distributed widely. Also, many commercial GIS packages contain example data sets, which may not be meant for any seri- ous use. They are widely used, however, and also silently accepted (Kalliola 2000b). Government mapping institutions have traditionally set the standard for cartographic products. Partly as a cost-recovery method and partly to control the quality of associated products, these institutions charge significant sums of money for digital data. Companies that create new products have access to fewer resources than government agencies and are usually not able to finance purchases of gov- ernment databases. In addition, these companies are often in a competitive situation, which priori- tises rapid development over map quality. This problem is thus a juxtaposition of two broader technology issues set within a digital cartography context: ‘data quality versus cost’ and ‘state ver- sus market-determined standards’. We believe that this quality problem is a part of normal so- cio-technical artefact development and that the criticism regarding the dubious quality of early- to-market products and the lack of standards is an important part of this evolution. Conclusions For present-day scientists, the Finnish cartographic legacy is a windfall as archives burgeon with val- uable representations of the country’s landscape. These historical maps reveal important changes in the landscape and document the ways land was used and maps were made at different times. We believe that this cartographic tradition is an im- portant guide to the development of new and in- novative, but reliable and sustainable, ways of producing and using GI in the coming decades. New data sources and processing methodology continue to test the skills of GI experts and open up interesting new cartographic ‘views’ on human and natural worlds. Geographic information can now be combined to produce a myriad of user- defined thematic maps. This utility only arises, however, with careful attention to the spatial ac- curacy and thematic content of the created digit- al data layers. All use of combined data must be conducted with reference to data individuality and intricacy. Once digital GI data is combined, new servic- es may be developed quickly, especially when connected to automated locational systems and portable computing devices. The challenge here is to understand the human–computer interaction factors so that larger societal issues of usability, privacy, sustainability, and copyright are account- ed for. Finally, issues of data quality must be fore- most in the minds of cartographers. Whereas tra- ditionally the requirements of the government dic- tated standards, now many parties, both public and private, create maps. The speed with which digital geographic data can be collected, proc- essed, and distributed to the public has increased enormously. Changes both in cartographic design and conventions of distribution (including stand- ards of acceptable accuracy for different levels of digital map data) will come about. Cartographers and geographers will play an active role in these technical and social changes, helping to guide the future of maps in Finland. ACKNOWLEDGEMENTS The authors wish to thank Dr. Petri Pellikka for valu- able comments on the manuscript. Many thanks also to the University of Turku Laboratory of Computer Cartography (UTU-LCC) for facilitating the study. The study is part of the project “Landscapes of the past, present and future” (SA 37121), funded by The In- formation Research Programme of the Academy of Finland. REFERENCES Aronoff S (1991). Geographic Information Systems: A management perspective. 2nd ed. WDL Publi- cations, Ottawa. Artimo K (1994). The bridge between cartographic and geographic information systems. In MacEachren AM & DRF Taylor (eds). Visualiza- tion in modern cartography, 45–61. Pergamon, Oxford. Artimo K (1996). Kartta – kulttuurin kuva. Maankäyt- tö 2, 8–11. Bagrow L & RA Skelton (1985). History of cartogra- phy. 2nd ed. Precedent, Chicago. Bowker GC (2000). Mapping biodiversity. Interna- tional Journal of Geographic Information Systems 14, 739–754. Burnett C & R Kalliola (2000). Maps in the Informa- tion Society. Fennia 178, 81–96. Burnett C & T Toivonen (2001). Applicability of semi- automated stand mapping to ecological mapping. In Belward A, E Binaghi, PA Brivio, GA Lanzar- 120 FENNIA 180: 1–2 (2002)Niina Vuorela, Charles Burnett and Risto Kalliola one & G Tosi (eds). Proceedings of the interna- tional workshop Geospatial Knowledge Process- ing for Natural Resource Management, Varese, Italy, June 28–29, 2001, 209–212. European Commission Joint Research Centre, Ispra. Campbell J (1993). Map use and analysis. 2nd ed. Wm C Brown, Dubuque. Clarke KC (2001). Getting started with Geographic Information Systems. Prentice Hall, New Jersey. Cousins SAO (2001). Analysis of land-cover transi- tions based on 17th and 18th century cadastral maps and aerial photographs. Landscape Ecolo- gy 16, 41–54. Dickinson GC (1979). Maps and air photographs. 2nd ed. Edward Arnold, London. Dorling D (1997). Mapping: ways of representing the world. Longman, Harlow. Ekqvist B (1846). Karta öfver Runsala med derunder lydande holmar. Turun kaupunki, Mittaus- ja tonttiososton arkisto, Turku. Fabis B & P Gudmandsen (2001). Space technology programmes 1995–2000. TEKES Evaluation Re- port 11. Faiz S & P Boursier (1996). Geographic data quali- ty: from assessment to exploitation. Cartographi- ca 33, 33–40. Graham S (1998). The end of geography or the ex- plosion of place? Conceptualizing space, place and information technology. Progress in Human Geography 22, 165–185. Granö JG (1930). Puhdas maantiede. WSOY, Porvoo. Granö JG (1997). Pure geography. Eds. O Granö & A Paasi. The Johns Hopkins University Press, Bal- timore & London. Gustafsson AA (1932). Suomen topografiset kartat. Maanmittaushallituksen julkaisuja 20. Gustafsson AA (1933). Maanmittarikunta ja mit- taustyöt Ruotsinvallan aikana. Suomen maanmit- tauksen historia I. WSOY, Porvoo. Haapanen R & A Pekkarinen (2000). Utilising satel- lite imagery and digital detection of clear cuttings for timber supply management. In International Archives of Photogrammetry and Remote Sens- ing XXXIII, Part B7, 481–488. Geomatics Infor- mation & Trading Centre, Amsterdam. Holm M (1995). Global image matching and surface reconstruction in object space using three aerial images. The Photogrammetric Journal of Finland 14, 9–23. Holm M, A Lohi, M Rantasuo, S Väätäinen, T Höyh- tyä, J Puumalainen, J Sarkeala & F Sedano (1999). Creation of large mosaics of airborne digital cam- era imagery. In Proceedings of the 4th Internation- al Airborne Remote Sensing Conference and Ex- hibition, II, Ottawa, Canada, 21–24 June 1999, 520–526. ERIM International, Ann Arbor. Hyyppä J, U Pyysalo, H Hyyppä & A Samberg (2000). Elevation accuracy of laser scanning-de- rived digital terrain and target models in forest environment. In Proceedings of the 4th EARSeL Workshop on Lidar Remote Sensing of Land and Sea, Dresden, 16–17 June 2000. Johnsson B (1965). Synpunkter på 1600-talets tidiga geometriska kartering. Ymer 1–2, 9–84. Jones C (1997). Geographical Information Systems and computer cartography. Addison Wesley Longman, London. Kallio P (1999). Vähemmän karttoja, kunhan on tark- koja. Helsingin Sanomat 13 October 1999, A19. Kallio K, T Kutser, T Hannonen, S Koponen, J Pul- liainen, J Vepsäläinen & T Pyhälahti (2000). Re- trieval of water quality by airborne imaging spec- trometer in various lake types at different seasons. The Science of the Total Environment 268, 59– 77. Kalliola R (2000a). Quality, information and maps in the information age. Foresight 2, 223–229. Kalliola R (2000b). Digitaalinen paikkatieto – tietoi- suus sijainnista muuttaa arkea. Reports of the In- formation Research Programme of the Academy of Finland 4, 245–251. Tietoyhteiskunnan tut- kimuskeskus, Tampereen yliopisto, Tampere. Kalliola R, C Burnett & N Vuorela (2001). Needs for critical interdisciplinary research in the use of digital geographic information. In Pantzar E, R Sa- volainen & P Tynjälä (eds). In search for a human- centered information society, 235–251. Tampere University Press & The Information Research Pro- gramme, Tampere. Kalliola R, N Vuorela & I Suojanen (1999). Digitaa- liset kartat ja paikkatieto. In Westerholm J & P Raento (eds). Suomen kartasto, 26–27. Suomen Maantieteellinen Seura & WSOY, Helsinki. Kasvio A (1997). Digitaalinen kumous, työn murros ja tietoyhteiskunta. In Stachon K (ed). Näkökul- mia tietoyhteiskuntaan, 89–108. Gaudeamus, Helsinki. Keates JS (1996). Understanding maps. 2nd ed. Long- man, Harlow. Kraak M-J & FE Ormeling (1996). Cartography. Visu- alization of spatial data. Addison Wesley Long- man, London. Krisp J (2000). Multimedia cartography and informa- tion kiosks. Unpublished Master’s thesis. Depart- ment of Geography, Ruhr University, Bochum. Lähde P (2001). Henkilökohtainen navigointi ja kar- tografia matkailusovelluksissa – tutkimusesimerk- keinä Ruissalo ja Vienan Karjala. Unpublished Master’s thesis. Department of Geography, Uni- versity of Turku. Laurini R & D Thompson (1992). Fundamentals of spa- tial information systems. Academic Press, London. Lönborg S (1901). Geografiska och kartografiska ar- beten i Sverige under 1600-talet. Ymer 2, 113–144. Lyytikäinen HE (1983). Valtakunnalliset karttatyöt. In Maanmittaus Suomessa 1633–1983, 448–518. Maanmittaushallitus, Helsinki. Mäkelä H & A Pekkarinen (2001). Estimation of tim- ber volume at the sample plot level by means of image segmentation and Landsat TM imagery. Remote Sensing of Environment 77, 66–75. Mäkisara K, J Heikkinen, H Henttonen, T Tuomai- FENNIA 180: 1–2 (2002) 121Cartographic traditions and the future of digital maps nen T & E Tomppo (1997). Experiments with im- aging spectrometer data in large-area forest in- ventory context. In Proceedings of the 3rd Inter- national Airborne Remote Sensing Conference and Exhibition II, Copenhagen, Denmark, 7–10 July 1997, 420–427. ERIM International, Ann Ar- bor. Martínez JM, N Floury, T Le Toan, M Hallikainen & M Mäkynen (2000). Measurements and model- ling of vertical backscatter distribution in forest canopy. IEEE Transactions on Geoscience and Remote Sensing 38, 710–719. Mellaart J (1963). Excavations of Catal Hyük. Ana- tolian Studies, Journal of the British Institute at Ankara 19, 17–177. Monmonier M (1996). How to lie with maps. 2nd ed. The University of Chicago Press, Chicago. Morrison J (1989). The revolution in cartography in the 1980s. In Rhind, DW & DRF Taylor (eds). Car- tography: past present and future, 169–185. El- sevier Applied Science, London. Niemelä O (1982). Thematic maps – important car- tographic products. Maankäyttö 4, 169–172. Niemelä O (1984). Early mapping of Finland. In Niemelä O (ed). Atlas of Finland, Folio 112: Map- ping of Finland, 2–3. National Board of Survey & Geographical Society of Finland, Helsinki. Niemelä O (1998). Suomen karttojen tarina 1633– 1997. Maanmittauslaitos, Karttakeskus & Suomen Kartografinen Seura, Helsinki. Nurmi J (1999). Kartan suunnittelun muuttujat ja te- kijät. Teknillinen korkeakoulu, maanmittausosas- to, kartografia ja geoinformatiikka 5. Ochieng WY & W Chen (2000). High level archi- tectural and programmatic issues of the Galileo System. Journal of Geospatial Engineering 2: 2, 16–20. Okkonen J, T Hyvärinen & E Herrala (1997). Aisa Airborne Imaging Spectrometer – on its way from hyperspectral research to operative use. In Pro- ceedings of the 3rd International Airborne Remote Sensing Conference and Exhibition I, Copenha- gen, Denmark, 7–10 July 1997, 189–196. ERIM International, Ann Arbor. Paulaharju A (1947). Maanpinnan muotojen ku- vaaminen sotilastopografisissa kartoissa. Maan- mittaus 3–4, 71–81. Pellikka P (1998). Development of correction chain for multispectral airborne video camera data for natural resource assessment. Fennia 176, 1–110. Pellikka P, DJ King & SG Leblanc (2000). Quantifi- cation and reduction of bidirectional effects in deciduous forest in aerial CIR imagery using two reference land surface types. Remote Sensing Reviews 19, 259–291. Rakkolainen I, S Pulkkinen & A Heinonen (1998). Visualizing real-time GPS data with internet’s VRMLworlds. In Proceedings of the 6th interna- tional symposium on Advances in geographic in- formation systems, 52–56. Washington DC. Raper J (2000). Multidimensional Geographic Infor- mation Science. Taylor & Francis, London. Richardson DE (1996). Spatial, semantic and tem- poral data integration for application in remote sensing and Geographic Information Systems. Cartographica 33, 1–2. Robinson AH, JL Morrison, PC Muehrcke, AJ Kimer- ling & SC Guptill (1995). Elements of cartogra- phy. 6th ed. John Wiley & Sons, New York. Roeck Hansen B (1996). The agrarian landscape in Finland circa 1700. Stockholm University, De- partment of Geography, Meddelanden 97. Samberg A 1997. What laser scanning can do today? In Proceedings of the 3rd EARSeL workshop on Lidar Remote Sensing of land and sea, Tallinn, Estonia, 17–19.7.1997, 119–124. Shemeikka P (1998). Jos metsään haluat mennä nyt: maantieteilijät ja maastokartat. Terra 110, 129– 130. Skånes H (1996). Landscape change and grassland dynamics. Naturgeografiska Institutionen, Stock- holms Universitet, Avhandlingsserie 8. Sporrong U (1990). Land survey maps as historical sources. In Sporrong U & H-F Wennström (eds). Maps and mapping. National atlas of Sweden, 136–145. SNA Publishing. Tahvanainen L, L Tyrväinen, M Ihalainen, N Vuorela & O Kolehmainen (2001). Forest management and public perceptions – visual versus verbal in- formation. Landscape and Urban Planning 53, 53–70. Tollin C (1991). Ättebackar och ödegärden. Riksan- tikvarieämbetet, Uppsala. Tolvanen H (ed) (2001). Lounais-Suomen paikka- tietoyhteistyön kehittämishanke, 2. vaiheen ra- portti. UTU-LCC Publications 2. Tomppo E (1996). Multi-source national forest inven- tory of Finland. New thrusts in forest inventory. European Forest Institute Proceedings 7. Tomppo E (1998). Standwise forest variate estima- tion by means of satellite images. IUFRO Re- search Notes 21, 103–111. Travel and traffic information (2001). Finnish Road Administration (Finra), May 2001. Vuorela N (2000). Can data combination help to ex- plain the existence of diverse landscapes? Fen- nia 178, 55–80. Vuorela N (2001). Combined use of spatial data: implications for landscape dynamics in an oak woodland site in southwest Finland. Annales Universitatis Turkuensis AII 150, Biologica–Geo- graphica–Geologica. Vuorela N, P Alho & R Kalliola (2002). Systematic assessment of maps as source information in landscape research. Landscape Research 27, 141–166. Ziegler M, H Konrad, J Hofrichter, A Wimmer, G Ruppert, M Schardt & J Hyyppä (2000). Assess- ment of forest attributes and single-tree segmen- tation by means of laser scanning. Laser Radar Technology and Applications V 4035.