PROTOCOL FOR LANDSCAPE QUALITY STUDIES

METHODOLOGY AND INDEPENDENT VARIABLE

METHODOLOGY
Independent variable

Dependent variables

METHODOLOGY OF LANDSCAPE QUALITY ASSESSMENT

The common components in most studies of landscape quality are summarized by Figure 1. The landscape represents the independent variable and its characteristics may be described and measured in a variety of ways. The human landscape preferences are the ratings of the landscape scenes and are the dependent variables. The survey aims to gain ratings of the landscape from participants. Analysis of the preferences together with the characteristics of the landscape establish the relationship between the two and may generate predictive models which may be used to establish the likely landscape rating for a given set of characteristics. The results of the analysis may then be applied including mapping of landscape quality for the area.

Figure 1 Methodology of landscape quality assessment

Each of these components is examined.

INDEPENDENT VARIABLE

The independent variable comprises the landscape to be assessed in the study region. It involves photographing the landscapes and selecting, on the basis of landscape units, scenes for inclusion in the survey. It also involves measuring the characteristics or landscape factors present in the landscapes.

PHOTOGRAPHS

Photographs are surrogates for an on-site assessment of the landscape. Taking a large group to people out into the field to experience and rate the landscape may be regarded by some to be the best way to rate landscapes and that using photographs and rating scales are inferior. They distance the participant from the reality and immediacy of the landscape and through the use of a rating scale also renders it distant conceptually. Studies of these issues have found, however that the reverse is true.

Click here for a review of research into the use of photographs.

Photographs tend to provide more objective, more dispassionate responses, while site assessments can result in a more subjective response influenced by a range of site factors unrelated to landscape quality (e.g. heat, wind, odour, sound) as well as personal factors such as tiredness, hunger, thirst and boredom.

Black and white photographs reinforce likes and dislikes and produce more extreme responses than colour photographs (Shuttleworth, 1980). Photographs have the advantage over field assessments of enabling the evaluation of potential changes to the landscape, e.g. developments and seasonal changes. The availability of digital cameras has played a significant role in gaining a large numbers of photographs from which a representative sample can be drawn.

Technological developments - digital photography, Internet and email - have thus largely removed the impediments of gaining a large representative sample for rating scenes.

Kane (1976, 1981) was one of few of the Australian studies to take participants into the field to assess landscapes and generally it has been found impractical. Very few overseas studies took assessors into the field (Dearden, 1980; Briggs and France, 1980). Apart from the obvious practical advantage, photographs enable widely separate locations and seasonal changes to be assessed on a comparable basis and they also enable hypothetical changes to the landscapes, e.g. quarries, buildings, wind farms, to be assessed by rating the scene with and without the change (Trent et al, 1987; Zube et al, 1987; Lothian, 2008).

In the interests of efficiency and effectiveness, photographs provide a satisfactory alternative to field assessments providing they meet specific criteria which standardise as far as possible the scenes so that the landscape is rated in a consistent manner. Only my own studies and Preston (2001) included detailed criteria for photographs. Following are the criteria for photographs (based in part on Shuttleworth, 1980; Trent et al, 1987).

• Photograph in horizontal landscape format, not the vertical portrait format. The photographs should extend where possible to the horizon and avoid close-up confined views. Include some sky to help convey its landscape character. The issue of format is consistency and standardising the survey scenes, the two formats should not be mixed in the same survey.

Morialta Falls - horizontal format   Morialta Falls - vertical format Go to top

• Avoid photographic composition of a scene to frame a view or to lead the viewer into a scene; such composition can enhance its appearance. Aim for good lateral and foreground context to scenes, of a single landscape unit, and of typical representative scenes, not unusual (i.e. rare) scenes. However significant features such as waterfalls, cliffs and water bodies should be included. 

• Minimise extraneous features such as people, sheep or cattle, wildlife, fences, electricity poles and wires, and excavations or other eyesores, each of which can influence preferences either positively or negatively. Hull and McCarthy (1988) found wildlife enhanced preferences. Scenes should not include features of an ephemeral nature as these are not part of the permanent landscape scene.

Avoid sheep or other animals	Avoid poles and wires
Photograph over the fence, do not include it	Exclude sheds, containers, fences
Roads should not dominate the scene	A restorer's dream: junked cars in the landscape
Bay with posts and roadside curb	Bay without posts and roadside curb

• Avoid transitory effects of special atmospheric lighting such as sunsets or particularly vivid side lighting. Heavy cloud dampens the colour saturation while spectacular cloud formations can enhance the scene. I found that the rating of a sunlight scene but with extensive cloud cover averaged 1.2 lower (on 1 – 10 scale) than cloudless scenes (Lothian, 2000). Interestingly however scenes with a few scattered clouds averaged nearly 0.1 higher than cloudless scenes. Herzog and Bosley (1992) argue that mist and haze reduce the clarity of the scene and its understandability (in Kaplan’s terms) which would lower ratings. These findings support standardizing scenes with cloud-free conditions.

• Aim for sunny cloud-free conditions to standardise scenes against a blue sky. To avoid the strong side lighting of morning and evening reduces the potential time to around 6 hours, from 10 am to 4 pm. Then often the conditions are poor for photography with heavy cloud cover, mist and rain further reducing opportunities. The following scenes from the same location capture some of the range of conditions.

• Photograph at ground level. It is recognized that this can include scenes from hills and mountain tops of valleys and vistas below and these should include foreground to provide context as otherwise the scene can appear as though it was taken from an aircraft. Photographs would not normally be taken from the air as this is not the usual way the landscape would be viewed. However where aerial oblique photographs are used, ground based scenes should also be included for comparison of ratings (Ramsay, 1992).

Overall the ratings should reflect the quality of the scene, not the quality of the photograph, and standardizing photographs as far as possible through the application of these criteria will assist in ensuring this is achieved.

In some instances it may be necessary to draw from existing collections of photographs. Caution is needed to avoid selecting photographs which are well composed, have appealing lighting or clouds, or have people or other extraneous features. Experience shows that about 90% of such collections will be rejected (Lothian, 2000, 2009).

With Photoshop™ and other similar programs, photographs can be altered, for example to remove extraneous objects. While this can be used to remove electricity poles and the like, such manipulation risks the photograph ceasing to accurately represent the landscape and should therefore be used minimally to edit out unnecessary objects and not to change colours or remove intrinsic features of the landscape.

Bishop (2000) believes that advanced computer graphics is replacing photographs as a means of visual presentation and if this occurs the foregoing criteria remains relevant (see also Zube et al, 1987). Wu et al (2006) has applied sophisticated Geographical Information Systems and three-dimensional landscape modeling which enables the manipulation of large amounts of physical data. 

In summary, photographs or advanced computer graphics be used, standardized (colour, 50 mm, landscape format, cloud free conditions, avoid extraneous and transitory features, avoid composition, photograph from ground level) so as to minimize variations in the photographs other than in the landscapes they represent. Use digital manipulation sparingly to remove unnecessary features, not to enhance the scene.

Go to top

CLASSIFY THE LANDSCAPE UNITS IN THE STUDY AREA

The photographs selected for the survey aim to sample the range of landscapes present in the study region. This may be achieved by classifying the region into landscape units of broadly similar characteristics. Preston (2001) defined land types from aerial photographs while Williamson and Chalmers (1982) combined viewing distances with natural and artificial features into landscape compositions.

Previous physiographic classifications for the study region should be examined and adapted where possible for the visual landscape. As found by Williamson and Chalmers (1982), photographing the region will provide familiarity with its characteristics from which the landscape units may be classified. Landscape units need not be overly complex. The aim is to differentiate the region’s landscape sufficiently to ensure the photographs adequately sample its characteristics. Figure 2 illustrates the identification of landscape character areas for two studies.

Source: Lothian, 2000 and 2005b
Figure 2 Examples of landscape unit classification: South Australia and Barossa region

The landscape units need not be mapped. In the Coastal Viewscape project (Lothian, 2005a), the following landscape units were defined: high cliffs, low cliffs and beaches, headlands and bays, beaches and dunes, and the samphire-mangrove formation. Each were described and their extent on the coast measured. The proportions of each landscape unit per region provided the basis for the selection of photographs.

Allocate photographs to each of the landscape units and make a careful selection. The allocation need not be on the basis of area as some extensive plains are large in area but with little variation. Rather the selection should capture the characteristics of each of the units. The South Australian landscape includes an extensive arid region which covers 86% of the State, large tracts of which lack diversity; in my study (Lothian, 2000) the entire area was represented by only 29% of the scenes. In contrast, the coast, agricultural regions and particularly the Mt Lofty Ranges were far more diverse and required more photographs to capture their complexity.

Go to top

LANDSCAPE CHARACTERISTICS

The scenic quality of scenes derives from their content including land forms, trees and water which trigger responses in participants. I term such features landscape characteristics. Alternative terms include scenic quality indicators (Chenoweth et al, 1997), attributes (Preston, 2001), landscape dimensions (Williamson & Chalmers, 1982), and visual features (Wu et al, 2006). Scoring such characteristics in the scenes allows the analysis of ratings to proceed beyond mere description of the ratings to understand the contribution of landscape factors to the scene. Multiple linear regression analysis allows these landscape characteristics (the independent variables) to be compared with the ratings (the dependent variable) and to identify and quantify which landscape characteristics contribute to the ratings and their relative significance. Depending on the selection of the landscape characteristics, the models can explain a large proportion (say 75%) of the variance of the data.

An example of a model is the following for the River Murray in South Australia (Lothian, 2007):

Scenic quality rating = 2.278 + 0.71 diversity + 0.42 water + 0.32 cliffs; R2 = 0.65;

where each of the landscape characteristics were scored on a 1 – 5 scale on the basis of their significance in the scenes. The R2 indicates that this model explains 65% of the variance. A model of the same area with five landscape characteristics gained a R2 of 0.81. Such models can provide a short cut means to assess the scenic quality of an area – scoring the relevant characteristics produces the scenic quality rating.

An alternative to scoring the landscape characteristics using participants is to measure the area and perimeter of each factor in the scene. Such an approach was used in an early landmark study by Shafer et al, 1969 and has been used by several since (Shafer and Brush, 1977, Wherrett, 2000). Shafer used the term landscape zone for the predictor variables (Figure 3). The method was criticised by Carlson (1977) as being “completely formalist” while Bourassa (1991) considered the results to be spurious as there was no causal link between the predictors and the preference scores.

Figure 3 Shafer, et al, 1969 Classification of scene (simplified)

Preston (2001) was the only Australian study which coded the pixels in a grid imposed over the photographs. This study identified 110 attributes from which 14 were used to predict emotional response.

Despite Weinstein’s (1976) statement -"with enough independent variables, a regression equation can be derived that will correlate perfectly with any dependent variable, no matter how meaningless and inappropriate the predictors actually are”, likely predictors can be selected that are readily identifiable and measurable.

The review of Australian preference studies found the most common landscape characteristics to be land cover, land forms, water and built forms (Table 1).

Table 1 Landscape Characteristics in Australian preference studies

Table 2 summarises the landscape characteristics I have used in landscape studies. Naturalness and diversity have been used in most surveys as has trees and water.

Table 2 Landscape characteristics of South Australian studies

Landscape characteristics in the scene may be scored on any scale but the 1 – 5 (low - high) scale is suggested to differentiate it from the 1- 10 scale used for rating of scenic quality. This generally provides sufficient discrimination of the characteristics. Characteristics may be scored on the basis of their significance in the scene where significance was not just dependent on the proportion of the scene occupied by the feature but also reflected the prominence of the feature in the scene as judged by a group of viewers. Other characteristics (e.g. naturalness, diversity) may be scored on a continuum from low to high. Because judgement is necessary to score the factors, a small group of people, say 10 - 20 should undertake the scoring.

In summary, the likely landscape characteristics which contribute to scenic quality are identified and scored; a 1 – 5 scale is suggested. A small group of participants should score the scenes. Clear instructions of the basis of scoring need to be provided.

Go to top

Go to Dependent Variables

Go to Analysis of Results and Mapping Landscape Quality