Creating a landscape atlas: a demonstration and discussion.
Landscape maps are formalised relational representations of data and meaning. They portray understandings of environmental and social information across space. Landscape maps have been around for a long time (Dorling and Fairbairn, 1997, p.6). In Eurasia, they date back to at least 6200 BC, with the still extant map of Catal Hyuk, Anatolia (O'Connor and Robertson, 2004). In Australia there is a long tradition of songlines, probably developed over many millennia, which are maps that can take the form of rock engravings, verses and rituals (Nanou, 2003; Rose, 1996, p.7). Where such examples describe multiple landscape characteristics, they could be described as landscape atlases.
By nature, landscape atlases have multiple information stakeholders. Environmental professionals, farmers, students, tourists, amongst many other people, all might want to access landscape information, possibly even want to contribute to a package; these are people with different backgrounds and different motivations.
Add to that the perspectives and contexts which do not necessarily have a human voice (e.g. a koala, a wetland, or a river), and it is apparent that issues such as interpretation and flexibility are important, as well as inclusion and integration of information. Whilst an atlas can be pitched to suit potential viewer preferences, what should set a landscape atlas apart from other publications (including more technical atlases or data portals) is that it treats a space from many different vantage points, but most importantly, with a relatively open and balanced style.
Hillman et al. (2003) argue that holistic and balanced approaches to landscape management are as much a consequence of how management is undertaken as of what management is about. They suggest that it is a positive thing for management processes to be adaptive and participatory; these ideals are upheld in this paper. Hillman et al. (2003, p.229), observing increases in social capital (e.g. better organisation and trust), write that:
these changes allowed for a more open, 'non-defensive' use of information and a consequent increased engagement ... with the full range of information rather than identifying individual information 'parcels' as supportive or otherwise of a stakeholder interest. The capacity of stakeholders to move beyond a focus on single or 'pet' issues was seen as critical.
The research reported in this paper starts from the premiss that just as it is important to look both at the processes and forms of management, so too is it important to analyse the processes of information development, as well as the forms of it. The paper aims to undertake an analysis of this sort. There are many methods and ideas for conveying landscape information, influences that can have a large bearing on what information is available and how it is used. There has been a trend in recent years to see landscape atlas developers as mere technicians (Casey, M., pers. comm., 29 January 2004; Dorling and Fairbairn, 1997, p.vii). This paper, following on from Dorling and Fairbairn (1997), argues that landscape
atlas developers should function as information facilitators, managers, technicians and artists, amongst other roles.
The paper recounts and discusses the development of a demonstration landscape atlas, one developed by the author that deals with the Capertee Valley of NSW. This atlas is developed in several different, but related, versions: as hard copy, as a PDF file for CDROM (and also downloadable from www.laurel.ocs.mq.edu.au/~pbatten/pbatten.htm), and as a website (also available from the same site), and behind them all is GIS, text and graphic material. Buckley (2003, p.150) notes that technological transformations in cartography--with printed, CD-ROM and website atlases being common--allow a reconceptualisation of what is an atlas. All of the versions of the Capertee Valley Landscape Atlas (CVLA) relate in particular to landscape topography and biodiversity themes, but the concepts raised have relevance to a wide range of environmental and socio-cultural domains, to the mapping of landscapes, even to information management in general.
The structure of this paper incorporates a background section before outlining some of the important procedures used in developing the demonstration landscape atlas. These sections are then followed by an overview of some other mapping packages (especially a NSW landscape atlas, CANRI, 2003; and a Victorian landscape atlas, Imhof, 2003) in light of the demonstration landscape atlas. Finally, discussion and concluding statements synthesise the salient points of the paper. These discussions, including the outline and overview, are valuable in their own right, as they comment on current mapping practice. They also provide an important foundation for discussing conceptual principles for the mapping of landscapes.
One of the most important and yet simplest decisions in the development of a landscape mapping package is to have multiple maps rather than just one layer. As was argued in Batten (1999a, 2001) for landscape-topography representation, and in Batten and Aplin (2002) for biodiversity understanding, multiple information layers reflect the diverse nature of landscape. Multiple interpretations are meaningful and useful for these separate information domains, topography and biodiversity; they are certainly meaningful and useful for overall landscape appreciation. It is important that a landscape atlas, across its multiple maps, strives to convey the complexity of a landscape, to respect it, but also to describe organization.
Two other mapping principles encouraged in Batten (1999a, 2001) are rigour and meaningfulness. A technique used to work on these principles in the landscape-topography domain was to regulate the data brought into a package (morphological data itself as opposed to surrogate data), and then to generate other information from that based on theory and application. However, it is not necessarily appropriate to regulate the data input into an open-ended landscape atlas. Both researched and perspectival information, from methodical and innovative data sources, are suitable. Whilst rigorous physical meaning is not necessarily required in all landscape maps, explanation is important, and putting layers in their context is valuable. These messages were carried in Batten and Aplin (2002). The four biodiversity domains--history and health, counting and valuing -are filters which provide some context, and are a part of a framework which encourages even more explanation of context, through the use of textual support.
Some form of written explanation is traditionally associated with maps in atlas productions. Key information, and more recently metadata are examples of it; but these do not necessarily allow commentary on meaning, beyond that of representation or background. Map support material can include natural language statements that undertake analysis and synthesis. This commentary should be about leading out and, importantly, opening up discussion of the mapped information. The text in national and international atlases (e.g. The Macquarie World Atlas, 1994, p.71) and guide books that run parallel to some single map productions (e.g. Porteners et al., 1997) are excellent demonstrations that there is a capacity to fulfil this point. It seems appropriate, given the inclusive and integrative philosophies of landscape appreciation, that such commentary be a pivotal part of landscape atlases (in contrast, the developers of CANRI, 2003, for example, do not use commentary). The role of atlas developers is obviously central to the character of a landscape atlas. They oversee content and quality. The role is even more central, though, should commentary be a part of atlas development, as argued for in this paper.
It is appropriate to precede a description of the CVLA by briefly recounting a previous mapping application from its developer. The Hastings Catchment Information Package (Batten, 1999c) was developed as a part of a project that researched two aspects of commentary: integration and interpretation (Batten, 1999b). Each map (some are presented in figure 1) was value-added with written commentaries of the visual information. Moreover, the different layers were linked thematically, both within the message of the interpretations and in PDF using hot-links (areas which, if mouse-clicked, change the page location). Finally, an overall synthesis of the layers, a Hastings Basin landscape summary, was created, including the provision of some quick-grab statistics (figure 1b). Whilst this integration was rudimentary, it did form--together with the interpretation--part of the conceptual platform from which the CVLA developed.
[FIGURE 1 OMITTED]
OUTLINE OF KEY PROCEDURES IN ATLAS DEVELOPMENT
The information contained in this document has been produced to assist public knowledge and discussion and to contribute to good landscape management. This 'Landscape Atlas' has been set up as a living database. The information package is intended to be expandable and changeable, with a mixture of researched and perspectival influences. All presentations should be viewed in a balanced way. (from 'intentions' page at beginning of CVLA)
In addition to the details explained in the quote above, the Capertee Valley Landscape Atlas (CVLA) has been developed as a concept package for a PhD project of the author at Macquarie University. As a part of this process, topography and biodiversity were used as demonstration theme areas. Whilst these theme areas are closely linked to landscape character, other information, such as that contained in social layers, is also relevant to landscape character, and can fit into this version of the package in its 'general section'. As the CVLA explains, though, 'it is not out of the question for other major theme areas to be added in the future. In fact, it is a good idea for other theme areas to be added'.
Information frameworks for topography (after Batten, 1999a, 2001) and biodiversity (Batten and Aplin, 2002) were accessed and, whilst the collection of map layers in the CVLA is not as developed as it could be, each of the core components of those frameworks is covered. In the case of biodiversity, map layers demonstrate the four major filters discussed in Batten and Aplin (2002): biological indices, historical integrity, biotic health, and landscape valuing. In the case of topography, map layers include: general shape measures, process-zone measures, catchment measures, and a synthetic characterisation layer, as outlined in Batten (2001).
The general section of the CVLA includes a regional map (see figure 2), and, similar to the Hastings Catchment Information Package, an outline map and a geology map. Also presented is an initial climate layer--of average annual rainfall--and a demonstration perspectival layer that deals with heritage. This latter reference, to a perspectival layer, is of particular interest, as the following paragraphs explain.
[FIGURE 2 OMITTED]
As Dorling and Fairbairn (1997, p.80) evocatively account, an important task in atlas development is to acknowledge subjectivities within information provision. This idea is heightened even more within the Internet, for example, given the immediate nature of multiple stakeholders / wide uses. This paper argues that it is useful to acknowledge the context of information, giving background to subjective information, objective information, or information that is a combination of both, which most layers are (after Dorling and Fairbairn, 1997, p.160). Maps should not be treated as absolutely authoritative, as the last word on a topic. The following statement is included in each of the theme areas of the CVLA:
Other information layers that are available for addition to the Capertee Valley Landscape Atlas, either currently available (with Paul Batten) or ideas for production (in italics), are: ...
Even more important in the context of openness is the message contained in the CVLA that encourages other parties to contribute to a regularly updated (living) information package:
You are invited to share any of your ideas about the Capertee Valley Landscape Atlas, be it a suggestion for presentation or an idea for another layer. In addition, there are hardcopy pro-forma for both mapping and providing associated textual information.
The inclusion--in the atlas itself--of facilities that encourage stakeholder involvement in atlas development (see pro-formas in figure 3) is one of the most important contributions of this paper. The idea that different peoples' (not just experts') perspectives are of interest is an important manifestation of the landscape principles of inclusion and flexibility. This idea also promotes balance in an atlas. The expanded scope can offer a diversity of perspectives and interpretations.
[FIGURE 3 OMITTED]
Different Delivery Media
The option of having different delivery media is also about better connecting to potential atlas users. The CVLA is more accessible because it caters for different styles of readership. During the Capertee Valley fieldwork, many stakeholders requested a hard copy version, whilst many others requested electronic versions, including both online and CD-ROM. The CVLA is available in each of these three delivery media.
Microsoft Publisher was used for the hard copy CVLA. Publisher pages are relatively stable and efficient at including and placing both text and images (especially in comparison with Microsoft Word or equivalent applications). This style of file format (Adobe Illustrator is a similar application) is therefore effective for organising the atlas pages. Whilst images available for the CVLA are also stored as separate files, it was not deemed necessary to do the same with text at this stage. The base CVLA maps are stored as GIS--in the case of the CVLA, this was ESRI ArcMap files (which organise data layers including shape files and grids). Through a process of trial and error, it was decided that it was better to not transfer the maps to the Publisher files (by saving as image files, such as EPS, JPEG or TIF, and then embedding into Publisher, or even storing as those image files or PDF). The balance between quality and file size means that it is better to print straight from ArcMap for the CVLA hard-copy production.
PDF is suited to CD-ROM because it balances print-out quality (layouts are stable, including text and vector lines) and smaller file sizes, and has the added bonus of some capacity for document navigation. In this way PDF sits between file formats designed especially for printout (ArcMap and Publisher in this case) and those designed for online access (HTML, JPEG, and other browser-readable and download-efficient media). Both the CVLA pages stored as Publisher files and those stored as ArcMap files are translated to PDFs by 'printing' to the Adobe Distiller (note: the export to PDF option in ArcMap is not preferred to the Distiller for quality and file-size reasons). If a file is to be changed in any way, the publisher or ArcMap file is altered, and the PDF reprinted.
[FIGURE 4 OMITTED]
Information Synthesis and Package Structure
A pyramid structure was used in the CVLA (see figure 5). The base level of the pyramid is made up of individual maps and their support information, whilst the upper levels of the pyramid contain organisational material and commentary. Introduction, summary, and suggestions are used in these upper levels to explain and collate the layers (base level) of each theme area (middle level) to provide transparency on decision-making and a quick grab on the subject matter of the atlas material and where it was sourced. The top level of the pyramid, which deals with the CVLA as a whole, synthesises through from the theme areas (the middle level of the pyramid). In addition to introduction, summary, and suggestions, the top level also explains the intentions of the package (as referenced at the beginning of the 'outline' section).
[FIGURE 5 OMITTED]
The base level of the pyramid has object-oriented information parcels. 'Object-oriented' means here a structure that includes autonomous information units or layers (after Macquarie Essential Dictionary, 1999). The latter term--layers--is more appropriate for the base level given that the same geographical space is being treated in each object. So as to provide some context for the maps, each layer is structured so that there are many opportunities for providing background and commentary (see figure 2 and figure 3). At the same time as being able to be viewed independently from the other atlas pages, each layer profits from being a part of an atlas. This is attributable to two inter-related concepts: context and perspective. The other material in the atlas provides further information on the contexts and perspectives involved within a map and its immediate support material. The atlas is more that the sum of its parts, particularly in terms of context and perspective.
The final decision about what themes to use, and what layers to include within them, is up to the landscape atlas developers themselves. Atlas developers make more than just technical decisions, as Buckley (2003, p.155) explains:
Atlas mapping involves the presentation of coherent information about selected themes. This does not mean that the atlas maker does not need to deal with collecting and compiling, sifting and sorting, retaining and removing, and assembling and arranging data.
Developers manage a landscape atlas. Management involves controlling the character of a project, and bringing it through to presentation (after Macquarie Essential Dictionary, 1999). Atlas developers make major choices based on their appreciation of the subject matter and how it can be balanced. Technical decisions, such as those discussed above for delivery media, are important; choices regarding aesthetics also heavily influence viewing experience (Robinson, 1989)--atlas design, from backgrounds to headings to support images, were considered at length in the CVLA development. But probably the most significant of all the roles that atlas developers have is the shaping of the material in an atlas, all the way from theme and layer choice through to the meaning and pitch of each piece of commentary.
Different packages can even be created to suit different purposes. For example, the hard copy version of the CVLA demonstrates how a subset of layers can be collated--in this case the regional map, elevation map, and satellite image--and compiled into a tailored product--in this case an overview (of 16 pages) of the longer CVLA (currently 64 pages). As well as selecting the subset of maps, the development here involved, amongst other tasks, the tailoring of upper-level syntheses.
As inferred above, the principle of sourcing landscape information from multiple points is upheld in this paper. Individual CVLA information layers were developed as a result of both internal and external inspiration. The CVLA includes layers created in the process of its development (particularly the topography layers), but also layers sourced from other parties. Skills of facilitation are required in atlas development, where legal and other organisational issues need to be negotiated. Further, the process of tailoring such externally sourced information involves cartographic skills such as those related to scale, resolution, accuracy, projection, generalisation, and description (after Dorling and Fairbairn, 1997). For example, the CVLA geology layer is a generalised version (where classifications are less technical and grouped together) of externally sourced information. The development of this layer also threw up challenges, such as different data accuracies in different areas of the map, including some data being absent. It is a truism that challenges like these are negotiated on a case-by-case situation. In spite of this, or even because of it, atlas developers benefit from accessing principles such as those outlined in this paper.
OVERVIEW AND COMPARISON WITH OTHER PACKAGES
Various issues encountered in the development of a landscape atlas relate to the political ecology of mapping and GIS (Harley, 1988, p.277). The capacity of the CVLA to be further influenced by viewers' feedback and contributions is a particular, positive example of such an issue. The following paragraphs overview various GIS projects and their 'power' implications. Before embarking on that overview, it is important to note that an atlas is not about actually doing on-ground management itself. An atlas is a tool that can help contribute to good landscape management, but it does not make field decisions. Decision-support systems are often misconstrued as decision-making systems.
High-end, online GIS projects, such as those using ESRI's ArcIMS are proliferating (for examples, see links off Mahoney, 2002). There is little doubt that there are positive features associated with these projects, and they are certainly getting better all the time. However, it is healthy to critique high-end projects, especially by asking how accessible they are, both for viewing and creating information.
High-end online GIS software packages are so resource-intensive that there can be a power imbalance between those who are able to be involved, and those who are excluded (after Dorling and Fairbairn, 1997, p.80). The development and distribution of landscape maps, however, is by no means necessarily immoral. As the MapCruzin (Clary-Meuser Research Network, 2003) and Inova projects (Goncalves, 2003) demonstrate, shareware such as ESRI's ArcExplorer and other generic software is available for the wide exchange of rich information. As the sub-heading on the MapCruzin homepage asserts, their network has a goal of 'democratizing the production and consumption of information and knowledge' (Clary-Meuser Research Network, 2003). In this spirit, the CVLA attempts to circumvent technology-dependence by offering a range of transfer formats (e.g. paper, CD-ROM, and online delivery) and facilitating stakeholder involvement (e.g. a pro-forma can be filled out and sent in to the atlas developer). The central role of the atlas developer is a position of power, but one that can be negotiated with a landscape ethic of inclusion, integration, and positive engagement (see also Aplin and Batten, in press).
The Community Access to Natural Resources Information (CANRI, 2003) is an online landscape atlas that uses high-end mapping and database technology (including a product developed in-house, 'ICMISS'). There are glimpses of exciting things here, but some fundamental principles should still be taken into account. In terms of friendliness of use, a question that immediately strikes the viewer of the CANRI mapping sites is whether they are overly technical, and whether the delivery process is time-efficient. It can take several minutes, even on a good computer and network, for CANRI maps to load. Not only is information output often burdensome, but getting layers into high-end landscape atlases, particularly in large organisations, can be ponderous.
An example of a simple but effective landscape map is one that was linked straight off the NSW Rural Fire Service website front page (NSW Rural Fire Service, 2003). It was a single JPEG image showing the spatial distribution of the then recent and ongoing NSW bushfires. As well as being interesting, importantly, this landscape map was accessible (both in terms of download time and map presentation); in other words, it was user-friendly. The map could have an option of higher resolution and zoom in, so that there was better access to locations and spatial relationships. This review of technological depth was taken into account in the CVLA development, in particular, with the dual use of JPEG and PDF maps in the online version. The option of pre-creating the online maps rather than have them created at the time of use (such as in high-end online GIS) is another important decision. Not only does this suit the limited bandwidths available now, but it also allows the different delivery formats (hardcopy, PDF, online) to correspond, and, more importantly, it eased the path of meaningful interpretation and synthesis (as maps and their support material could be easily matched).
Interactivity, where viewers are able to select and transform displays (Robinson et al., 1995), is, according to Buckley (2003, p.150), 'the most important technological change in atlas making'. As was implied earlier, CANRI (2003) has strengths; and on the score of manipulating multiple layers, its mapping capabilities are impressive. High-end online GISs offer the opportunity for end-users to compare layers by overlaying them. An important question, though, is whether interactive display options can distract from the information transfer process. As was also implied earlier, the development of the CVLA treated interpretation and synthesis as more important than dynamic overlay. Perhaps there is a middle ground, where all three online processes could be used, and where the dynamic overlay would not compromise the commentary or synthesis.
A Swedish ArcIMS site, the Naturatlas for Soderhamn! (Kartorna, 2003) is an example of a high-end online landscape atlas that does incorporate commentary--a generic commentary is available as a side-panel option. An even better example of a major landscape atlas production that does include commentary is the Victorian Resources Online or VRO (Imhof, 2003). Whilst the VRO is linked to a dynamic mapping capability, the majority of the site uses pre-made maps. More importantly, the VRO and the associated Regional Matters: Atlas of Regional Victoria (Information Victoria, 2003) augment pre-made maps with a brilliant style of layer summation and interpretation of information.
One example of this capacity is the use of a brief but poignant style of upper level summation, as demonstrated in the following quote: 'Climate: Victoria has a moderate climate which is generally favourable to plant growth' (Imhof, 2003, Natural Resources Homepage). Similar to the CVLA, the VRO uses a 3-layer pyramid structure with an object-oriented base level, and a middle level split into an expandable set of theme areas (Climate, Landform, Landuse, Soil, Water, Biodiversity, Vegetation, Land & Water Management, Investing in Land & Water).
DISCUSSION AND CONCLUSIONS
Even though the CVLA has a focus on biodiversity and topography information, the concepts raised by its development are relevant to other applications, with different focuses and in different places and data realms. A major thread developed in the outline and comparison above is that discussions of context and perspective are important. All information layers are a mixture of subjectivity and objectivity (after Dorling and Fairbairn, 1997, p.160). Irrespective of whether atlas developers create a map themselves or incorporate an externally developed layer, they are in a position to use commentary to discuss the nuances of the information, its background, and its potential uses.
In this spirit, not only researched but also perspectival information is available for atlas compilation. From the end-users' point of view, one of the most important outcomes of the research reported here is the encouragement of their involvement. The CVLA has prominent and recurrent calls for end-users to submit their ideas or feedback, and even for them to contribute information layers that could be included in the atlas. Another concept intended to promote a culture of involvement is the pitching of the landscape atlas as a 'living' database; that is, one that has a capacity to change regularly. Open atlas development can be manifest right from the character of map layers, how they are presented, with what they are arranged, and how the overall atlas is compiled, integrated, and synthesised.
The task of creating landscape atlases can be an involved one, but it need not be a heavily resourced one. High-end landscape mapping has the potential of providing impressive mapping products, particularly with updating, high resolution, zoom and, if used well, overlaying options. But there should not be a situation where low-end maps are considered of no consequence. Simple maps, as well as being accessible, can also be beautiful, meaningful, informing, and balanced. This paper argues that creativity is more important than technology.
Atlas development is not mere technical work; it is information facilitation and management, involving technical, but also artistic, development. A professional culture of seeing mappers as only technicians is a poor one. It creates workplace tension, and adds to a culture of deterministic products. If the level of respect and acknowledgement of landscape mappers were to increase, then better productivity should follow for many reasons, including there being more freedom to develop creative and imaginative mapping products, and a mandate to access material in an inclusive, integrative and flexible manner. The mappers in turn should match this respect by delivering on these principles of application.
Principles such as accessibility and balance are important to both high-end and low-end landscape mapping. The Capertee Valley Landscape Atlas demonstrates that simplicity and information depth can go together. The development of the CVLA reveals an expandable / changeable nature, and an object-oriented design. The latest, long version in PDF and HTML, and a hard-copy short version, amongst others, are testament to the changeable nature of the overall design of the information package. The object-oriented information template means that maps and associated information can be added efficiently and effectively.
Aplin G. and Batten, P.E., in press, "Open-minded Geographers: Their Potential Role in Integrated Adaptive Environmental Management", Australian Geographer.
Batten, P.E. and Aplin, G., 2002, An Expanded Conceptual Framework for Biodiversity, IAG Conference, Canberra, ANU, July.
Batten, P.E., 1999a, The Development of a Landscape Mapping Framework, Honours thesis, Sydney, Macquarie University.
Batten, P.E, 1999b, "Catchment information packages: a format for the distribution of environmental information", RipRap, vol. 13, pp. 8-9.
Batten, P.E., 1999c, Hastings Catchment Information Package, (Online). Available: http://www.ocs.mq.edu.au/~pbatten/phd/ appendix3.htm Accessed 15/2/03.
Batten, P.E., 2001, A New Approach For Landscape Mapping, 6th International Conference on Geocomputation, Brisbane, Australia.
Buckley, A., 2003, "Atlas mapping in the 21st century", Cartography and Geographic Information Science, vol 30, no. 2, pp. 149-159.
CANRI, 2003, Community Access to Natural Resources Information, (Online). Accessed: http://www.canri.nsw.gov.au Accessed 30/1/03.
Clary-Meuser Research Network, 2003, MapCruzin, (Online). Available: http://www.mapcruzin.com Accessed 30/1/03.
Dorling, D. and Fairbairn, D., 1997, Mapping: Ways of Representing the World, London, Longman.
Goncalves, P.P., 2003, Inova GIS, (Online). Available: http://www.inovaGIS.org Accessed 30/1/03.
Harley, J.B., 1988, Maps, knowledge and power, in Cosgrove, D. and Daniels, S. (eds), The Iconography of Landscape, Cambridge: Cambridge University Press.
Hillman, M., Aplin, G. and Brierley, G., 2003, "The importance of process in ecosystem management: lessons from the Lachlan catchment, New South Wales, Australia", Journal of Environmental Planning and Management, vol. 46, no. 2, pp. 219-237.
Imhof, M., 2003, Victorian Resources Online, (Online). Available: http://www.nre.vic.gov.au/web/root/Domino/ vro/vrosite.nsf/pages/vrohome Accessed 5/9/03.
Information Victoria, 2003, Regional Matters: Atlas of Regional Victoria, Melbourne: Victorian Government.
Kartorna, 2003, Naturatlas for Soderhamn!, (Online). Available: http://www.kartorna.nu/website/natur/viewer.htm Accessed 30/1/03.
Macquarie Essential Dictionary, 1999, Sydney: The Macquarie Library.
Mahoney, N.C., 2002, Websites Powered by ESRI's Internet Solutions, (online). Available: http://www.esri.com/software/internetmaps/ visit_sites.html Accessed 30/1/03.
Nanou, 2003, Songlines, (Online). Available: http://www.nanou.com.au/songlines/ Accessed 7/7/03.
NSW Rural Fire Service, 2003, Frontpage (Online). Available: http://www.bushfire.nsw.gov.au Accessed 27/1/03.
O'Connor, J.J. and Robertson, E.F., 2004, The History of Cartography, (Online). Available: http://www-gap.dcs.st-and.ac.uk/ ~history/HistTopics/Cartography.html Accessed 8/1/03.
Porteners, M.F., Ashby, E.M., and Benson, J.S., 1997, "The Natural Vegetation of the Pooncarie 1:250 000 Map", Cunninghamia, vol. 5, no. 1, pp. 139-231.
Robinson, A.H., 1989, Cartography as an art, in Rhind, D.W. and Taylor, D.R. (eds), Cartography Past, Present and Future, London: Elsevier Applied Science for the ICA.
Robinson, A.H., Morrison, J.L., Muerhcke, P.C., Kimerling, A.J. and Guptil, S.C., 1995, Elements of Cartography, New York: Wiley.
Rose, D.B., 1996, Nourishing Terrains: Australian Aboriginal Views of Landscape and Wilderness, Canberra, Australian Heritage Commission.
The Macquarie World Atlas, 1994, Revised Edition, Sydney: The Macquarie Library.
Paul Batten, PhD Researcher, Department of Human Geography, Division of Environmental and Life Sciences, Macquarie University, NSW 2109, Australia. Email: email@example.com, Phone: (61 2) 9850 6372; Fax: (61 2) 9850 6052.
|Printer friendly Cite/link Email Feedback|
|Date:||Nov 1, 2004|
|Previous Article:||Mapping a new colony: the geographical construction of Queensland 1860-1880.|
|Next Article:||Bits of real estate: digitizing the church map collection of heritage maps at Auckland City Libraries.|