3 Initial Landscape Classification Parameters

3.1           INTRODUCTION

3.1.1         This section of the Report demonstrates how a set of parameters can be derived for use in understanding the landscape, and classifying landscape into different types. It then sets out the proposed landscape classification parameters for this Study.


 3.2            TERMINOLOGY

3.2.1         The Study Brief requires that what it terms 'landscape classification criteria' be mapped in order to provide a baseline for the Study as a whole. The word 'criteria' has for the purposes of this and future reports, been replaced by the word 'parameters' as the latter is the term more usually used in the field of landscape assessment and properly suggests the identification of defining features or terms.

3.2.2         The term 'landscape classification parameters' is therefore taken to mean those natural or human features which can be found in different landscapes but which vary in terms of their presence or contribution, thereby defining the character of any particular landscape type. Examples of landscape parameter are 'geology', 'soils', 'vegetation', 'settlements', etc. These are all features which vary in terms of presence and composition from landscape to landscape but which help to define the characteristics of any particular landscape.

3.2.3         Landscape classification parameters are therefore the physical building blocks of the landscape which are to be mapped as part of the Study. The process of landscape classification involves analysing landscape parameters to produce areas of homogenous character or qualities, (i.e. landscape character types and landscape character areas such as 'upland peaks and valleys', 'rolling arable lowland', etc). The process of landscape classification is the next step in the Study and is set out in Section 4 of this Report.

3.2.4         Classification parameters should not be confused with evaluative statements, such as 'area of high landscape value', 'SSSI' or 'scenic viewpoint', etc. Such evaluative statements require the application of evaluative criteria and should be distinguished from the process of mapping the basic conditions of the landscape. The process of evaluation is the next Task in the Study (Task 3) and will be set out in Technical Report No.2.

3.2.5         Each 'landscape classification parameter' has a number of 'attributes'. Attributes are the different types of any given landscape classification parameter. Therefore, the landscape classification parameter Vegetation, contains numerous attributes, which might include 'Woodland', 'Tall Scrub', 'Low Scrub', 'Grassland', etc, etc. Similarly the landscape classification parameter Solid Geology might include the attributes 'Igneous Rocks', 'Sedimentary Rocks', 'Volcanic Rocks', 'Superficial Deposits', etc.


3.3           DERIVATION OF INITIAL LIST OF LANDSCAPE CLASSIFICATION PARAMETERS

3.3.1         In order to identify which landscape parameters need to be mapped as part of the Study, a number of further questions need to be addressed. These are:

    3.3.2         These questions are addressed in the following sections of the report.

                 WHAT DOES THE WORD 'LANDSCAPE' MEAN ?

3.3.3         In order to establish a list of landscape features or 'parameters' to be mapped as part of the Study, it is essential to interpret what the term 'landscape' means. Only through a comprehensive understanding of the term and it implications, can a meaningful list of parameters be derived that fully reflects the meaning and also values of landscape.

3.3.4         'landscape' is in itself not a map-able resource, as it has no single defining feature or dimension. Rather 'landscape' concerns certain aspects of the natural and built world, and our relationship with and attitudes towards those aspects. Therefore to map a 'landscape', one has to map a number of different dimensions which constitute it.

3.3.5         However, these different dimensions of 'landscape' are not self-evident and require adequate definition before they can be mapped. To identify these different aspects or parameters of the world which together comprise 'landscape', it is necessary to first define "landscape" so as to understand what makes it different from associated terms and ideas such as "land" or "terrain", etc. Only then is it possible to identify all the different features that need to be identified and mapped in order to classify landscape into its different types.

3.3.6         This section of the report identifies the various semantic associations or dimensions of the terms "landscape" in order to identify firstly what issues are covered by the terms and thereby, which issues need to be mapped as part of this Study.

                THE VISUAL/AESTHETIC DIMENSION

3.3.7         In English, the term landscape originates from the Middle Dutch landscap. This term in turn, was associated with a particular style of painting in from the 17th and 18th Centuries. The Concise Oxford Dictionary defines landscape as "Natural or imaginary scenery, as seen in a broad view".

3.3.8         The Pocket Oxford Dictionary defines landscape as "features of land-area as seem in broad view: picture of this...".

3.3.9         Both definitions stress the pictorial or visual elements of landscape. Clearly then in both its English origin and its common usage, there is a significant visual dimension to the concept of 'landscape'.

3.3.10        In order to capture the visual and aesthetic dimensions of landscape, recording of visual attributes is most appropriately carried out on sketches and photographs and in written descriptions. Features such as viewsheds, vistas and panoramas can be recorded appropriately on plan.

                THE NATURAL/ECOLOGICAL DIMENSION

3.3.11        Professor Brian Hackett, looking at the issue of landscape planning from a professional perspective, stresses the ecological and biotic associations of the term 'landscape'.

3.3.12        "The word 'landscape' has various meanings such as the view seen by an onlooker or the environment inhabited by man and other living beings. A person who takes a professional approach to his understanding of landscape, or is directly involved in it, is likely to accept the explanation that based upon land, modified by the climate and that the result influences the distribution of the types of flora and fauna" (Hackett, 1971, p1).

3.3.13       Although there are now urban landscapes which appear almost devoid of influences of flora and fauna, most landscape is still largely influenced by these factors (which in any case maintain a significant hold on the term from the origins of 'landscape painting').

3.3.14       Clearly, then natural land-based and biotic factors are a key element in our understanding of what 'landscape' is. To capture the natural and ecological dimensions of landscape, mapping should include solid and drift geology, soil class, topography, (relief / slope / aspect), flora / fauna, habitats and hydrology.

                THE POLITICAL/ECONOMIC DIMENSION

3.3.15       Marxist critiques suggest that what we call 'landscape' or 'townscape' is in fact the physical expression of factors of production, the manifestation of the spatial dynamic between labour and capital, and the locational characteristics of capital (Harvey, 1988; Pugh, 1989).

3.3.16       From any political angle, it is clear that the influence of systems of economics and politics (seen as the science of wealth distribution) have a particular bearing on both the visual conditions of the landscape as well as upon the functioning of its natural systems. What we term landscape is in large part a product of our attempt to live on land and to exploit it to our own advantage. This results in a specific political and economic dimension to the landscape.

3.3.17       Any attempt to map landscape must take account of the political and economic dimensions of landscape. The political / economic dimension is most appropriately captured by mapping land use, transportation infrastructure and built form (proportion of built / unbuilt area, building height and building age).

                THE CULTURAL/SOCIAL DIMENSION

3.3.18       Simon Schama suggests natural / ecological and visual / aesthetic factors are not enough in themselves adequate to provide a definition of 'landscape'. For Schama, what defines landscape is its contribution to social memory and myth. He states that "it was always the inherited tradition" that made landscape out of mere geology and vegetation¡Kand it is culture, convention and cognition... that invests a retinal impression with the quality we experience as beauty... If our entire landscape tradition is the product of a shared culture, it is by the same token a tradition built from a rich deposit of myths, memories and obsessions..." (Schama, 1995, p11-14).

3.3.19       The suggestion here is that land becomes landscape only when its cultural values are recognised and when its role in the development of folk-memories is appreciated.

3.3.20       In order to capture the cultural and mythic qualities of landscape, it is necessary for the landscape mapping study to identify those aspects of it, which make manifest or contribute to its cultural and psychological significance.

3.3.21       Such features might include fung shui woodlands, temples, buildings or districts with specific cultural connotations (often those designated as scheduled ancient monuments). However, it also includes associations with particular landscape features, which contribute to its importance. This includes paintings, literature (prose and poetry), music and photography (film and photographs) which are associated with particular places and particular landscapes.

                THE CHARACTER/SENSE OF PLACE DIMENSION

3.3.22       In addition to the above, there is also a dimension of landscape which whilst an attribute of the landscape is a product of the above influences. This is what Alexander Pope in 18th Century England termed "genius loci", or what may be termed "sense of place" or "landscape character". Certainly in the UK, the concept of "character" is inseparable from the term "landscape", in professional circles at least, and is in this sense an important contributor to its meaning.

3.3.23       Quite what causes specific places to possess a specific character is not entirely clear and is that which all designers aim for. However, it is certainly to do with distinctiveness, consistency, coherence and legibility. In the area of landscape the UK Countryside Commission provides the following definition of landscape (Countryside Commission, 1991, p3) - the term landscape "is not just concerned with the visual appearance of the landscape, it's qualities of shape form and colour but also more particularly, with the way in which the various components come together to create different landscapes. When these components occur in a distinct and consistent pattern, they give character to the landscape".

3.3.21       Such features might include fung shui woodlands, temples, buildings or districts with specific cultural connotations (often those designated as scheduled ancient monuments). However, it also includes associations with particular landscape features, which contribute to its importance. This includes paintings, literature (prose and poetry), music and photography (film and photographs) which are associated with particular places and particular landscapes.

                THE CHARACTER/SENSE OF PLACE DIMENSION

3.3.22       In addition to the above, there is also a dimension of landscape which whilst an attribute of the landscape is a product of the above influences. This is what Alexander Pope in 18th Century England termed "genius loci", or what may be termed "sense of place" or "landscape character". Certainly in the UK, the concept of "character" is inseparable from the term "landscape", in professional circles at least, and is in this sense an important contributor to its meaning.

3.3.23       Quite what causes specific places to possess a specific character is not entirely clear and is that which all designers aim for. However, it is certainly to do with distinctiveness, consistency, coherence and legibility. In the area of landscape the UK Countryside Commission provides the following definition of landscape (Countryside Commission, 1991, p3) - the term landscape "is not just concerned with the visual appearance of the landscape, it's qualities of shape form and colour but also more particularly, with the way in which the various components come together to create different landscapes. When these components occur in a distinct and consistent pattern, they give character to the landscape".

3.3.24       Certainly, if a picture of the landscape is to be formed which gives full expression to its nuances, it will have to include the concept of character. Character is most appropriately captured by breaking down a plan of the landscape into 'landscape character areas' (as the UK Countryside Agency recommends for non-urban landscapes ) or coherent 'districts' (as suggested by Kevin Lynch (Lynch, 1960) for urban areas).

                OTHER DIMENSIONS OF 'LANDSCAPE'

3.3.25       Those dimensions and attributes of landscape discussed above are those most commonly dealt with in the literature. There remain however, two further aspects of the wider term 'landscape', less frequently addressed, but certainly germane to this Study. These are the issues of 'urban landscape' and of 'coastal landscape'.

                URBAN LANDSCAPE

3.3.26       As noted above, landscape is amongst other things, an aggregation of natural and built elements of land. The extent to which land is altered by human beings, the extent to which buildings are built on to it is largely one of degree and there is a continuum between rural landscapes (largely natural backdrop punctuated by some built elements) and urban landscapes (largely built backdrop punctuated by some natural elements). The distinction between natural landscape and urban landscape is therefore a largely artificial one. This was recognised by the eminent landscape historian Professor W.G.Hoskins as early as 1955 in his seminal text 'The Making of the English Landscape', where he states:

                URBAN LANDSCAPE

3.3.27       "We may study with our guide books all the historic, individual features of a town and get to know them. But then¡Kother questions begin to arise in the mind, which even the best guide books are unable to answer. Why is the town just like this, this shape, this plan, this size? Why do its streets run in this particular way...Why are there sometimes two market places, why are the ancient churches just where they are? - and so on. In short, what gives the town its particular landscape".

3.3.28       What Hoskins stresses in his work, is that there can be no distinction between urban and rural, because the same forces influence both areas. As he notes in the passage just quoted, in order to understand urban areas properly, it is essential to look at them at the same level as a specific landscape and not as collections of buildings. Seen from this perspective, villages and towns and urban areas are not different from 'landscape', but are simply landscape at a certain point in transition, given their character by their relative number of buildings in any particular view.

3.3.29       The inter-penetrability of landscape and townscape has been recognised by Grady Clay (Clay, 1994) who has proposed a typology of landscapes which deals with urban and rural places equally. As suggested by Clay and as has been practised in the UK (Tidal Thames Landscape Strategy and Fareham Borough Landscape Assessment) it is possible to produce a typology for urban areas which deals with urban and rural landscapes at the same level.

                COASTAL LANDSCAPE

3.3.30       Coastal landscape is defined by its relationship with the sea and cannot exist without the sea. Therefore, to adequately define such landscapes, they must include both land and sea. Landscapes such as 'beaches' or 'rocky island' are defined as much by the sea as by the land.

3.3.31       There also exist landscapes which exist somewhere between land and sea, such as 'mud flat' or 'salt marsh'. These landscapes form an almost perfectly intermediate state between land and sea.

                COMMENTARY

3.3.32       Clearly, if all the features that constitute the term 'landscape' (be it natural, terrestrial, urban or coastal) are to be mapped, clearly all the different dimensions of the term noted above must be fully considered and their meanings and implications taken into account in the selection of landscape classification parameters.

                 WHAT COMPONENTS OF LANDSCAPE CONTRIBUTE TO ITS VALUE ?

3.3.33       In order to select a list of landscape parameters to be mapped, it may be necessary to understand what physical features of landscape are valuable. This is because one of the later aspects of the Study is to develop a measurable indicator for input into the SUSDEV model. Understanding at an early stage of the Study, what physical characteristics we value in landscape will allow us to define parameters representing those features so that their value may be meaningfully represented in the SUSDEV model.

3.3.34       Although it is somewhat pre-emptory to produce a definitive list of evaluation criteria at this stage of the Study, (as this is the objective of Task 3 of the Study), it is useful to review literature to identify any physical landscape features that previous studies have considered to contribute significantly to landscape value. In this way, these features can be included in the list of landscape features to be mapped and whatever evaluation technique is used in the future, one can be certain that these features have at least been mapped.

3.3.35       There are numerous approaches to landscape evaluation, but what may be termed 'psychophysical' and 'surrogate component models' of evaluation try to identify those features in any given landscape which contribute to scenic quality and which may therefore be used as indicators of public preference. These models have attracted some criticism, particularly for ignoring the relationships between features, but they are sufficiently prevalent in the literature that they should be addressed at this stage, even if this technique is not ultimately employed in this Study. The following are typical of the findings of such approaches to landscape evaluation.

3.3.36       Zube, Pitt and Anderson (1975) define a number of landscape features which they attempted to prove can be correlated with landscape value. These include:

3.3.37       Kaplan et al (1989) suggest that suggest that landform and land-cover have the most profound influence on landscape value. Components of landform that contribute to value comprise relief, topographic contrast and spatial diversity. Components of landform that contribute to value comprise naturalism, compatibility of land use, height contrast and variety of land-cover.

3.3.38       Fabos et al. during the METLAND study (1971-73) suggested that land-use (both diversity and compatibility) and tree-cover are key determinants of landscape value.

3.3.39       Brush and Shafer (1975) identify water and vegetation as key determinants in landscape quality, when viewed as a picture plane. Whilst using a similar technique, Polakowski (1975) identifies sky, vegetation, open space, water, built structures and topography as key determinants.

3.3.40       In her discussion of surrogate components models developed by Crofts and by Arthur et al., Wherrett (1996) concludes that the following components are most commonly linked to landscape value:

                COMMENTARY

3.3.41       Given that a number of current approaches to landscape assessment involve evaluation of the individual components of landscape, it would seem to be prudent at this early stage in the Study to map all those features which might individually contribute to that value, noted above. At least in this way, this potential method of evaluation remains open during the remainder of the Study.

                WHAT IS THE ORIENTATION/PURPOSE OF THE PARTICULAR STUDY ?

3.3.42       Landscape assessments are carried out for a wide variety of different purposes. The purpose and orientation of any given assessment will affect the landscape classification parameters that are selected for mapping and analysis.

3.3.43       In the United States and in Asia, landscape assessments are often orientated towards one or more specific objectives. These can include issues such as watershed management, recreation or historic asset management. In each case, the orientation of the Study will require an adjustment of the parameters mapped.

3.3.44       Certain studies are more orientated towards historic aspects of landscape. These are termed historic landscape assessments and are carried out in addition to a general landscape assessment, as in the case of the Hampshire Historic Landscape Assessment in the UK (1999). Such studies identify historic components in far greater detail than general landscape assessments.

3.3.45       Similarly ecological landscape assessments focus on those landscape components which are most closely related to the ecology of landscape and which define its natural ecological areas, as in the Study on the Ujung Kulon National Park in Indonesia. Such studies identify more specifically components that are not relevant to a general landscape assessment.

                COMMENTARY

3.3.46       The Brief for the current Study does not identify a specific orientation or area of interest for the Study but treats it as a general landscape assessment Study. The only particular focus is upon the development of a landscape indicator for use as part of the SUSDEV21 model. It is therefore not appropriate to collect very specific or specialised data on say hydrology, recreation, landscape history or ecology, but rather to cover all these areas at a general level, as is the case with most landscape assessments.

3.3.47       The following section of the report looks at a number of general landscape assessments and identifies the range of landscape parameters that is generally collected for such studies.

                WHAT PARAMETERS ARE TYPICALLY MAPPED IN OTHER STUDIES ?

3.3.48       Looking at what types of parameter are usually mapped in other Studies will be a useful guide to which features it will be appropriate to map in the current Study. In Section 2.0 of this report, information is provided regarding what has been mapped in five Overseas Case Studies. The parameters that have been mapped in a selection of other studies are noted below.

                US STUDIES BY IAN MCHARG, JULIUS FABOS ET AL.

3.3.49       Approaches such as those pioneered by McHarg in the United States map individual landscape resources on separate plans. In such studies, such as his Staten Island Study , Potomac River Basin Study and Quadrant Study, McHarg (1969) would typically map:

                COMMENTARY

3.3.50       Landscape assessment in the United States was developed principally from the fields of ecology, biology and natural sciences. This accounts for the mapping of aspects of landscape such as 'wildlife habitats' and 'ecological associations', features that are likely to be relevant to landscapes in Hong Kong. US studies are also typically driven by specific objectives (such as watershed management), and therefore features such as 'water table' and 'tidal inundation' may not need to be mapped in all cases.Countryside Character Map, UK (Countryside Commission, ongoing).

3.3.51       This Study mapped the following landscape classification parameters:

                COMMENTARY

3.3.52       This is a national study and therefore deals with data at a relatively broad scale (e.g. 'settlement pattern'), whilst also concentrating on using ecological and soils data to define natural areas. In a more detailed study, dealing with urban landscape and significant landscape diversity, it is likely to be necessary to go into more detail than this. The parameters also reflect prevailing land use conditions in the UK (such as 'farm type').

                TIDAL THAMES LANDSCAPE STRATEGY AND DESIGN GUIDELINES (NATIONAL RIVERS AUTHORITY, 1996)

3.3.53       This Study mapped the following landscape classification parameters:

                COMMENTARY

3.3.54       This is an urban study, centred on the river landscape of the Thames through London. The specific requirements of the Brief mean that there are landscape parameters that will not be relevant to the current Study (e.g. 'River Channel Types'). The 'Urban Morphology' attributes will however be of relevance to the current Study.

                'INTERIM LANDSCAPE CHARATER ASSESSMENT GUIDANCE', COUNTRYSIDE AGENCY/SCOTTISH NATURAL HERITAGE, 1999

3.3.55       This Guidance suggests that the following landscape classification parameters be used:

                COMMENTARY

3.3.56       These parameters are generally applicable in the Hong Kong context (except perhaps 'Patterns of Field Enclosure'). Note the specific reference to temporal aspects.

                'LANDSCAPE AND LANDSCAPE ASSESSMENT', DEPARTMENT OF THE ENVIRONMENT AND LOCAL GOVERNMENT, IRELAND, 2000

3.3.57       The Guidance suggest that the following factors should be recorded;

                COMMENTARY

3.3.58       This is a typical list of parameters, applicable to all scales of assessment. Interest in 'Historical Field Pattern' reflects specific geographical conditions.

                'OUR NATIONAL LANDSCAPES', TONY JACKMAN/BIOLOGICAL RESOURCES CENTRE, 1983

3.3.59       This was guidance published in New Zealand which recommends the following parameters be studied as part of a resource model:

                COMMENTARY

3.3.60       This is very much a resource based model and contains certain data which may not be relevant to a Study adopting a more character-based approach, such as 'climate' or 'animal and bird life'.

                PLANNING AND DEVELOPMENT STUDY ON HONG KONG ISLAND SOUTH AND LAMMS, HONG KONG SAR GOVERNMENT (ONGOING)

3.3.61       This is one of the more comprehensive landscape classification exercises carried out in Hong Kong in recent years, and established a database comprising:

                COMMENTARY

3.3.62       This study was primarily a characterisation exercise rather than a classification. A specific remit to address visual and recreational issues led to the inclusion of 'Recreational facilities' in the mapping exercise.

                STUDY ON SOUTH-EAST NEW TERRITORIES DEVELOPMENT STRATEGY REVIEW HONG KONG ISLAND SOUTH AND LAMMS, HONG KONG SAR GOVERNMENT (ONGOING)

3.3.63       This is another comprehensive recent landscape study, which mapped the following landscape features:

                COMMENTARY

3.3.64       This is a reasonably comprehensive list. Note the mapping of 'fung shui' features, important in the Hong Kong context.


3.4           DERIVATION OF LANDSCAPE CLASSIFICATION ATTRIBUTES

3.4.1         The selection of attributes for each parameter should be the fewest number that will give full potential to its particular landscape classification parameter to express the variety of possible landscape types. It will also be determined by the following:


3.5           INITIAL LIST OF LANDSCAPE CLASSIFICATION PARAMETERS AND ATTRIBUTES

3.5.1         From the above examples of 'best practice', a reasonably consistent group of parameters emerge. However in addition to those items mentioned above, the urban Hong Kong context, it will be pertinent to obtain detail data on building height, age and density. Temporal issues are rarely addressed explicitly, but will need to be one of the factors to be mapped in this case.

3.5.2         Therefore, based on experience gained hitherto, an initial list of classification parameters is set out below:

                 NATURAL FEATURES

HUMAN FEATURES

TEMPORAL FACTORS

3.5.3         The data available to the Study on these parameters is set out in Appendix 5, together with attributes on which information is available for each parameter.

 

Go to previousGo to main pageGo to next