Introduction

Landscape classification is driven by a need for practical solutions for the management of landscape change. The NZ Landscape Classification provides a macro, generalised description of landscape character that is intended for use alongside manual detailed landscape assessments that capture greater complexity. Although an inventory of natural and unnatural landscapes can be generated, the classification does not provide information on aesthetic quality. The NZ Landscape Classification was created using GIS technology and national spatial databases, which has generated a nationally consistent classification. This classification can be used to provide information on landscapes at a range of spatial scales (national, regional, and local) and at a range of generalisation levels.

New Zealand 's landscape is a multi-billion dollar tourism resource that has considerable quality of life value for ordinary New Zealanders. To manage this resource effectively, it is essential that it be classified so that planners have an inventory, and a frame of reference for communication and research. Classification is important for communication because it provides a consistent frame of reference. All the major sciences work with classifications (plant classifications, soil classification, chemical classifications, etc) and it is difficult to see how these disciplines could function without such classifications. A classification shows regions that are similar thereby allowing results to be inferred from one region to another. In order to advance our theoretical understanding of landscapes and the management of landscapes it is important to have a frame of reference for communicating about landscapes.

The classification of landscapes is complicated by the fact that it involves both human perception and physical reality, while many of the science classifications tend to be based on just the physical. This makes landscape classification particularly difficult because human perception generates a wide range of meanings. Jones (1991) describes this complexity as the "elusive reality of landscape" (p. 229). He adds that in the past, the lack of recognition that landscapes are a physical reality and a social or cultural construct has led to an "academic battlefield" with different disciplines and schools concentrating exclusively on either the physical landscapes or on the observer.

The uses of GIS-based classifications are becoming an important part of landscape assessment procedures and are regarded as complimenting manual field assessments. The Countryside Agency ( England ) use GIS as a tool for broad scale assessment of landscape character types (2004). Within the character types, fine scale landscape areas are identified manually using field visits and surveys. The combination of GIS and manual techniques appears to be a practical solution for landscape characterisation. The GIS approach ensures national consistency while the manual techniques capture the local subtleties. This hybrid approach therefore addresses many of the critiques of classification. The NZ Landscape Classification should be seen in the context that it can be complimented by manual techniques.

Landscape research methods are characterised by a wide range of different and sometimes conflicting perspectives. Palka (1995) has shown that there are a wide range of definitions of landscape, and there have been some excellent reviews of landscape literature by Zube et al. (1982) and Daniel and Vining (1983) that show the range of research perspectives used. For the purposes of the NZ Landscape Classification, landscape is defined as the appearance of the land. Swaffield (1991) showed that this is a common definition used by land resource planners. It is also consistent with Appleton 's (1980) definition - "the environment perceived, especially visually perceived" (p.14), and the Countryside Commission's (1970) definition - "the spectacle presented by the countryside"(p.2). Landscape is an overall impression presented by the land, and involves generalisation and combination (Brabyn, 1996b). The concept of visual landscape is different to the concept of landscape used by Landscape Ecologists who tend to focus on the physical environment. The NZ Landscape Classification compliments the Land Environments NZ (LENZ) classification (Leathwick et al. 2003) developed by Landcare Research. LENZ classifies pre-human ecosystems and does not include current human modifications such as agriculture, urban areas, and infrastructure. As a concept, an ecosystem is different to visual landscape character.

The NZ Landscape classification is explicit and repeatable, which are characteristics of all the major scientific classifications. These characteristics are important because it means that the procedures can be critiqued and evolve with new insights. Classification is based on understanding through research, but to research a classification is required. Classification therefore evolves with new understandings and all the major scientific classifications are developing in this way. GIS requires explicit instructions and the NZ Landscape Classification was generated using automated analysis functions applied to existing national spatial datasets. If these datasets are updated, or a particular step in the classification is improved, the automated process can be reapplied and the classification is updated.

The first version of the New Zealand landscape classification developed in 1996, was based on Lars Brabyn 's Ph.D. (Brabyn, 1996a). This work has subsequently been published in academic journals (Brabyn 1996b, 1997, and 1998). Since 1996, there have been major releases of national spatial data sets and improvements in the performance of software and hardware functionality. New classification techniques have been researched (Brabyn, 2005). The second version of the NZ Landscape Classification utilises the recently released Landcover 2 dataset (Ministry for the Environment, 2000), the 1:50,000 Topographical data, and also uses computationally intensive visibility analysis. The spatial resolution of the second version is 100m while the first version was 500m.

1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12