new banner
 

Home
Services of Scenic Solutions
Completed landscape projects
Papers and Presentations
Beautiful landscapes
Australian and International studies of landscape quality
The science of scenery
Landscape miscellany
Landscape resources
About Dr Andrew Lothian
Contact us
Site map

For other segments of the Science of Scenery, click here:

PROTOCOL FOR LANDSCAPE QUALITY STUDIES

A standard method needs to be adopted to progress the understanding of landscape quality. A standard method will enable the gradual accumulation of a substantial body of knowledge describing diverse landscapes and enable understanding to be derived from their comparable results. The model of the scientific study of landscape quality (Figure 1) identifies the components of a standard methodological approach to the study of landscape quality.

theory model

Figure 1 Components of the science of scenery

The purpose of this page is to detail the components of the method. Three framework issues are addressed and then the components of the methodology are examined.

Click on any of the following topics to go to that section.

FRAMEWORK ISSUES:

METHOD

ANALYSIS

MAPPING SCENIC QUALITY

FRAMEWORK ISSUES

TERMINOLOGY

The terminology used in the Australian studies examined varied widely, some using unique descriptors of the landscape while others used more common terms. Typically the descriptors combine a noun and an adjective: e.g. scenic quality or landscape value (Table 1).

Table 1 Descriptors of landscape

Nouns Studies Adjectives Studies
Scenic
31
Quality
37
Landscape
26
Value(s)
19
Aesthetic
13*
Preference
4
Visual
11
Amenity
3
Resources
3

* Mainly associated with Regional Forest Agreements

The combinations of the noun and adjective are illustrated by Figure 2, indicating their frequency of use. The most common phrases were those using the nouns, landscape or scenic together with the adjective quality. Scenic quality was used in 19 studies while landscape quality was used in 10 studies (five by this author). The term aesthetic value was used by 13 studies, associated almost entirely with the Commonwealth’s Regional Forest Agreement process and reflecting the terminology used under the National Estate Register.

terminology

Figure 2 Frequency of landscape terminology combinations

Other descriptors used by studies are summarized by Table 2.

Table 2 Minor descriptors used by landscape studies

Description Frequency
Visual landscape
4
Scenic amenity
3
Landscape preference
3
Scenic preserves
2
Classified & recorded landscapes
2
Visual landscape quality
2
Visual resources
1
Scenic landscape
1
Aesthetic landscape appreciation
1
Landscape setting types
1
Landscape visual quality
1
Landscape sensitivity to change
1
Scenic reaches (river)
1
Landscape assessment
1
Visual values/resources
1

This brief analysis of terminology indicates the diversity of terms that are used, and reinforces the need to settle on common terminology to reduce and avoid confusion. The science of scenic quality will be retarded by the lack of a common terminology.

I propose that the terms: scenic quality or landscape quality be adopted as alternative and equivalent terms. While purists might want to qualify this as perceived scenic quality, this is assumed and on the basis that simpler terms are best, this is unnecessary. Similarly further adjectives such as amenity or aesthetics need not be included.

It is recognised that the Australian lists, including the National Heritage Register, use aesthetic values as their basis. As this concept encompasses aspects of aesthetics beyond the visual, it remains appropriate that the visual aesthetics be covered by the above terms.

Go to top

PHYSICAL VERSUS PREFERENCE METHODOLOGY

The distinction between physical and preference studies reflects a philosophical difference in approach which has parallels with that between art and science, captured half a century ago by C.P. Snow in his prescient book, The Two Cultures (CUP, 1959). Art derives from personal experience, cognition and ability applied to the creation of new entities whereas science lets the phenomenon speak for itself and seeks to discover that which explains its characteristics.

Blaise Pascal's (1623 - 1662) pithy dictum, the heart has its reasons which the mind cannot comprehend, illustrates the dichotomy between cognition and affect.

In Australia there have been over twice as many physical studies as preference studies. The reasons for the popularity of physical studies may include the following:

The relative ease of conducting the physical study.
Although considerable effort is required in mapping the physical landscape and applying the landscape dimensions that have been defined, this is manageable; moreso if GIS is available allowing it to be undertaken quite speedily.

Control of the study
Compared with a preference study, where the results are determined by the community, a physical study is entirely under the control of the investigator.

Gaining a sufficient sample for preference studies
This has been a major deterrent in the use of the preference methodology. Early preference studies (e.g. Kane 1976, Dare, 1978, Revell, 1982) used only 10 – 15 participants and the results could be in error by up to 30%. Prior to the availability of the Internet, I took nearly six months to obtain a sample of 319 viewing slides of landscapes. However the ubiquitous availability of the Internet has changed this. Together with the expanding broadband network, the Internet makes possible the rapid ratings of scenes by a sample of hundreds, even thousands of participants with the convenience of sitting at their own computer and participating at a time that suits them. This is far more efficient than having to assemble a group of people at a fixed place where they can view slides or photographs. Table 3 illustrates this by showing the survey times for the landscape surveys I have conducted. A month is generally sufficient to obtain a large survey, however this presumes access to a large Intranet resource such as Government agencies. An alternative, so far untried, is that of open surveys on the Internet available to anyone ini the world with connection. This may be the way of the future.

Table 3 Survey Times for Author's Landscape Surveys

Survey Scenes Sample Days
South Australian Landscapes
160
319
170
Visual impact of Wind Farms
160
454*
see note
Tree Amenity Project
112
619
41
Coastal Viewscapes Project
166
3324**
24
Coastal Development Survey
82
2398**
21
Barossa and Light Region Landscape Project
150
2260**
29
River Murray Landscape Project
150
2138**
31
River Murray Development Survey
80
1427**
40
Flinders Ranges Landscape Project
147
3549**
27
Generic Landscapes Project
145
204
24
Lake District National Park Project
145
540
32

* Included 280 via Internet sampled in 7 days. ** Survey placed on SA Government Intranet

Preference studies require statistical competence
This is required for analysing the results and in carrying out more sophisticated analyses including multiple regression. However, the ready availability of statistical programs renders this difficulty far less than when calculations had to be done by programming a computer or by using a crude calculator. Nevertheless, the fear of undertaking statistical analysis is very real for many people and physical studies provide an alternative which avoids statistical analysis. In time this impediment might lessen as many of today’s graduates have a far greater familiarity with statistical techniques than in the past.

Concern over the use of photographs as surrogates for the landscape
In the early years there was concern as to whether photographs could adequately represent the physical landscape and some authors (e.g. Carlson, 1977) were opposed to their use. However research into the validity of using photographs as a surrogate of field evaluation of landscape quality (Dunn, 1976; Shuttleworth, 1980; Trent, Neumann and Kvashny, 1987; Stamps, 1990) has removed the objection that photographs are an invalid means of gaining preference ratings providing they meet certain criteria. Stamps (1990) found from a meta-analysis of studies, a correlation of 0.86 in the ratings of photographs and on-site assessments of landscape.

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

I contend that the only valid approach for the future assessment of scenic quality is by the use of preference studies. The days of the physical study are numbered; they had their use and played an important role in developing an awareness of scenic quality but a better instrument is now available and the tools are there to use it. Preference studies provide results which are replicable, which fulfil the scientific requirement for reproducibility – enabling another researcher using the same methodology to produce the same results, which cannot be said for physical studies.

While some believe that the two approaches to scenic quality can be merged (Daniel, 2001), I am less sanguine – they are literally poles apart in approach and philosophy. The best that could be expected would be for the assumptions underlying the physical studies to be tested using preference studies to ascertain whether they are correct.

Some 170 landscape dimensions have been used by physical studies in measuring scenic quality. Their individual contribution to scenic quality would be expected to vary and their efficacy in predicting scenic quality could be evaluated. Generally the results from physical studies are relatively coarse, e.g. high, moderate and low scenic quality, compared with the say, ten unit differentiation (1 – 10 scale) of preference studies which provides a three-fold improvement in outcomes.

Daniel (2001) identified precision, reliability and validity as the criteria for judging different approaches to measuring scenic quality. A comparison of the physical and preference approaches against these criteria (Table 4) reinforces the imprecision, variability and questionable validity of physical studies.

Table 4 Comparison of physical and preference studies

  Physical Study Preference Study
Precision Imprecise. Typically only three grades of scenic quality. The assessment correlates poorly with ratings derived from community preferences. Assessments are essentially nominal posturing as ordinal. Quite precise. Commonly able to discern landscape differences of half a unit, e.g. between 6 and 6.5. Precision is constant across different landscapes. Assessments may be assumed to be of interval scale. 
Reliability Often dependent on a sole rater. Variable from person to person and place to place. Statistical reliability cannot be measured. Assessments of a landscape by different experts may vary as much as assessments of different landscapes; i.e. observer variance is greater. Based on hundreds or even thousands of raters and therefore more closely reflects the community. Reliability can be measured statistically. Variations between landscapes are typically several orders of magnitude greater than variation among observer’s judgements of those landscapes; i.e. landscape variance is greater.
Validity Based on a sole expert. Studies of ratings by lay people and experts have tended to find expert ratings are similar (e.g. Kellomaki & Savolainen, 1984, Sullivan, 1994) but those of landscape architects often differed markedly from the community (Anderson & Schroeder, 1983; Brown, 1985). Relies on the expertise and experience of the rater in an appeal situation. Difficult to compare ratings of areas. Constancy over time uncertain. Being based on community preferences, it is highly valid and would better withstand scrutiny in an appeal. Comparable with ratings of other areas derived using similar methodology. Appears fairly constant over time.

Overall, a well conducted preference study can provide quite precise results, results whose reliability can be tested statistically, and it also provides a high level of validity as the preferences more closely reflect the community, not those of a single investigator.

A further reason for adopting preference studies over physical studies is based on the nature of aesthetics. Aesthetics is an affective quality (cf. affections); it does not derive from cognitive thought. In looking at an attractive landscape, one does not analyse it but rather one knows immediately and without prior thought that it is attractive. Dictionaries reinforce this distinction between the cognitive and the affective in their definition of aesthetics as “things perceptible by the senses (i.e. affective) as opposed to things thinkable or immaterial (i.e. cognitive) (Shorter Oxford, 1973).

Using an analogy from music, an individual’s liking for music does not derive from an analysis of the use of instruments, scoring for the orchestra, or a detailed analysis of the score. Rather it is immediate and without analysis.

The failure of landscape researchers in Australia in the 1970s and 1980s to develop a credible method for assessing scenic quality stemmed from approaching scenic quality from a cognitive analytical paradigm rather than from a preferences paradigm. So intent were they of being objective and avoiding the subjective, that they swung too far the other way and made analysis the sole basis for assessing scenic quality.

VISIBILITY OF THE AREA TO VIEW

Several studies, including the Visual Management Systems have assumed that the visibility of a scene to people is an essential part in measuring scenic quality. Seen areas, sensitivity levels and observer travel routes were components which measured the exposure of the area to view.

Preston (2001) argued that scenic amenity involved not only measuring people’s landscape preferences but also the extent to which places in the landscape were seen. This is illustrated by his diagram combining scenic preferences with visual exposure, scenic amenity being the product of both (Figure 3).

He stated that this approach takes a “democratic approach to assessing the importance of scenery.” However this implies that through access improvements, an area originally rated 6 may increase to 8 even though its actual scenic quality is unchanged.

The concept of visual exposure in the study assumes that scenic quality is a relative quality, relative to the visibility of the area under study. However the accessibility of an area changes over time which changes its visibility. Roads change and new routes are developed, opening up to view new areas.

Preston diagram
Source: Preston, 2001
Figure 3 Criteria for assessing the significance of scenery

Additionally, there is a wide range of means of viewing landscapes apart from the car: mountain bikes, four wheel touring, cycling, orienteering, bushwalking, wilderness travel, hang gliding. The coast, rivers and lakes may be viewed from motorboats, jetskis, yachts, cruise ships and canoes as well as from the air. There is a tendency to consider only the obvious means, which confines the discussion to cars and access roads. There are many other forms of travel which do not require roads.

By considering the area’s visibility, the study’s results are determined by when they were derived and the accessibility then available which may differ in the future.  The results do not stand the test of time but rather need to consider the accessibility then available. To take another environmental resource, that of biodiversity, this exists regardless of whether or not it is viewed by the public. Its worth is not contingent on its accessibility or view. I contend that, like biodiversity, scenic quality exists regardless of its visibility, in other words it is an absolute, not relative quality. Its value is not a function of its visibility but rather is independent of this.

This is contrary to the opinion, often expressed that because an area is not visible it does not matter if its visual amenity is marred by a wind farm, powerlines or quarry. Arguments before appeal courts sometimes take the position that the visibility of the site and the number of people viewing it are important considerations, it being implied that the less visible and the less the number of viewers then its scenic quality is not significant.  However, taking this to its logical conclusion could result in the widespread loss of scenic quality in the vast tracts of Australia not regularly visited.

In this context, Brown and Itami (1982) commented: “…while areas of the landscape that are seen by the most people may be more valuable to society than visually inaccessible landscapes, this is a measure of visual accessibility and not scenic resource value.”

Therefore in the standard methodology, considerations of visibility are irrelevant. The scenic quality ratings derived will provide an absolute, not a relative, measure of the significance of the area.

Go to top

Go to Methodology of Landscape Assessment

 
 

© 2009-14 Scenic Solutions   Privacy   Terms of Service   Site Map   Contact us