Landscape visualisation and climate change: the potential

Landscape visualisation and climate change: the potential for influencing perceptions and behaviour Stephen R.J. Sheppard Collaborative for Advanced La...

53 downloads 631 Views 1MB Size
Environmental Science & Policy 8 (2005) 637–654 www.elsevier.com/locate/envsci

Landscape visualisation and climate change: the potential for influencing perceptions and behaviour Stephen R.J. Sheppard Collaborative for Advanced Landscape Planning (CALP), Department of Forest Resources and Landscape Architecture, University of British Columbia, 2045-2424 Main Mall, Vancouver, BC, Canada V6T 1Z4

Abstract The urgent need to mitigate and adapt to climate change is becoming more widely understood in scientific and policy circles, but public awareness lags behind. The potential of visual communication to accelerate social learning and motivate implementation of the substantial policy, technological, and life-style changes needed, has begun to be recognised. In particular, realistic landscape visualisations may offer special advantages in rapidly advancing peoples’ awareness of climate change and possibly affecting behaviour and policy, by bringing certain possible consequences of climate change home to people in a compelling manner. However, few such applications are yet in use, the theoretical basis for the effectiveness of visualisations in this role has not been clearly established, and there are ethical concerns elicited by adopting a persuasive approach which deliberately engages the emotions with visual imagery. These questions and policy implications are discussed in the context of a theoretical framework on the effects of landscape visualisation on a spectrum of responses to climate change information, drawing in part on evidence from other applications of landscape visualisation. The author concludes that the persuasive use of visualisations, together with other approaches, may be effective, is justified, and could be vital in helping communicate climate change effectively, given ethical standards based on disclosure, drama, and defensibility. # 2005 Elsevier Ltd. All rights reserved. Keywords: Visualisation; Climate change; Visual communications; Carbon consciousness; Behavioural response

1. Introduction This paper addresses the potential role of realistic visualisation tools in rapidly increasing peoples’ awareness of climate change and possibly affecting behaviour and policy, together with associated ethical dilemmas. 1.1. The need for increased public awareness and policy response on climate change Evidence of climate change and its impacts on society and the environment have become generally accepted by scientists (Pearce et al., 1996; IPCC, 2001), along with the importance of mitigating and adapting to climate change (Fawcett et al., 2002; Adger, 2003). However, public awareness and policy changes are lagging behind. Boardman and Palmer (2003) found low levels of ‘carbon consciousness’ among European consumers and businesses, and Lorenzoni and Langford E-mail address: [email protected]. 1462-9011/$ – see front matter # 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.envsci.2005.08.002

(2001) reported that the majority of their study participants were not concerned about climate change. Many authors have addressed the substantial gap between the possession of environmental knowledge or behavioural intent and actual pro-environmental behaviour (e.g. Kollmuss and Agyeman, 2002). In some countries, organisations such as the UK Climate Impacts Programme (UKCIP, 2000a) have been set up to address this gap through education and communication. Tickell (2002) has summarised the difficulties in communicating climate change (discussed further below), and warned that it may require a calamity to induce people and governments to take the necessary radical actions. The principal question therefore becomes: what are the most effective ways to stimulate climate change mitigation and adaptation behaviours in society before crises occur, and reduce harm to the environment and society? The main hypothesis examined in this paper is that certain kinds of visual communication (i.e. realistic landscape visualisations) which attempt to look into the future and which engage the emotions, may substantially enhance awareness-building on

638

S.R.J. Sheppard / Environmental Science & Policy 8 (2005) 637–654

various complexities and implications of climate change, and may help motivate behavioural change at the individual to societal levels (Sheppard, 2004). Since few such visualisations addressing climate change have yet been produced or subjected to research evaluation, this paper attempts a proactive, systematic analysis of their potential, informed by early precedents. It is recognised that many more abstract, statistical, or iconic aspects of climate and related landscape change may not be readily communicated with realistic landscape visualisation, and other more conventional tools may be needed in addition. As background to the discussion of visualisation applications to climate change issues, this section provides a brief rationale for using visual imagery, and landscape visualisation in particular, in promoting environmental awareness and behaviour. 1.2. The rationale for using visual imagery and visualisation to raise environmental awareness and influence behaviour Human responses to environments and visual displays can be broadly categorised as cognitive (related to knowledge and understanding), affective (related to feelings, attitudes, and emotions), behavioural (related to changes in behaviour of the viewer), and physiological (biological or physical effects on the observer’s body) (Zube et al., 1982). The latter type of response is not considered further in this paper. Perception is defined here as the process of seeing or otherwise perceiving phenomena, leading to particular responses or states which include both cognitive and affective outcomes. The focus of the paper is primarily on individual perceptions, individual behavioural responses, and organisational behaviour (i.e. policy). The ability of visual imagery to communicate messages quickly and powerfully has long been recognised in fields of human activity ranging from commercial advertising, to the media, to political campaigns: seeing is believing. The cognitive advantages of visual information over written or verbal information have been widely documented (e.g. Tufte, 1990), for example, when mentally visualising conditions that cannot be seen directly in the real world (e.g. famines in remote countries, or the design of a proposed building). Visual stimuli can also trigger innate and instant reflexes and feelings which can be persistent even in the face of new information (Slovic et al., 2002); Zajonc (1984) has termed this the ‘‘primacy of affect’’ over cognition. Visual imagery can therefore be a powerful tool to reach people’s emotional side, as witnessed in the more dramatic imagery from the news media. Imagery (visual or otherwise) which triggers affective responses has been shown to improve cognition in some cases (Winn, 1997), and to influence people’s decision-making (Slovic et al., 2002). Clearly, the intent of some visual imagery is not just to inform or engage emotionally, but also to influence people’s behaviour, as in advertising and public service health announcements.

Community-based social marketing using such techniques has proved effective in bringing about behaviour change (McKenzie-Mohr and Smith, 1999; Kollmuss and Agyeman, 2002). Among the various forms of visual imagery available, 3D visualisations present special characteristics which help demonstrate consequences of environmental change in a compelling manner. Scientific information is commonly communicated by forms of data visualisation such as charts, diagrams, maps, graphics, and 3D computer models (Cox, 1990; MacEachren and Ganter, 1990). These 2D and 3D representations are often somewhat abstract or simplified (Fig. 1). Landscape visualisation, sometimes referred to as visual simulation or landscape modelling (Ervin and Hasbrouck, 2001), represents actual places and on-theground conditions in 3D perspective views (Fig. 2), often with fairly high realism (Sheppard and Salter, 2004). These specific forms of virtual reality are now typically computergenerated in three or four dimensions, and can convey detailed information on the assumed future appearance of environments (sky, terrain, buildings, and vegetation). This amounts to a unique form of visual communication, conveying information in the dominant form to which the human species is genetically adapted (i.e. visual landscapes), but capable of showing future worlds as they would be seen if the viewer were actually there. Historically, landscape visualisations have been used primarily in urban design and impact assessment (Porter, 1979; Sheppard, 1989; Bosselmann, 1998). GIS-based methods for landscape visualisation (Appleton et al., 2002) are now being applied more broadly to depict alternative future scenarios for visioning, public input, and decision-making (Tress and Tress, 2002; Steinitz et al., 2003; Sheppard and Meitner, 2005). Nicholson-Cole (2005) describes the promise of landscape visualisation in conveying strong messages quickly, condensing complex information, engaging people in issues of environmental change, and motivating personal action. However, we lack comprehensive knowledge and integrative theories to validate, explain, and predict a broad range of responses to landscape visualisations (Appleyard, 1977; Sheppard, 2001). These techniques also harbour risks and key limitations, and there are a number of ethical and

Fig. 1. Example of a conceptual or abstract 3D visualisation (showing terrain and elevation). Credit: John Lewis, CALP, UBC. Reproduced from Sheppard, Lewis, and Akai, 2004, Landscape Visualisation: An Extension Guide for First Nations and Rural Communities, courtesy of Sustainable Forest Management Network, Canada.

S.R.J. Sheppard / Environmental Science & Policy 8 (2005) 637–654

639

Fig. 2. Visualisations of landscapes can now be modelled in detail and realistically rendered, as in this depiction of two forest management alternatives after 38 years of plan implementation. Credit: Ken Fairhurst, Mike Meitner, Ryan Gandy, and Duncan Cavens, CALP, UBC. Courtesy of CANFOR.

professional issues raised by attempts to influence behaviour using visual imagery. It is therefore important to consider carefully available evidence and theoretical arguments for the effectiveness and ethics of using visualisation on climate change issues, in order to avoid false expectations, critique the inevitable use of visualisations of climate change, and guide further research.

influences will be required to effect societal change on climate change; the attempt made here is to articulate the additional or value-added contributions of visualisation within the toolset. In the rest of this paper, the term ‘visualisation’ refers to landscape visualisation, unless otherwise noted.

1.3. Scope and structure of paper

2. A theoretical framework for the impact of visualisation on awareness, attitudes, and behaviour in relation to climate change

These issues are discussed in the context of an initial theoretical framework on the possible effects of landscape visualisations on human responses to climate change; thus, Section 2 addresses the question: ‘‘Can landscape visualisation deliver on improving awareness and stimulating action on climate change?’’ Section 3 examines dilemmas relating to such use, essentially addressing the question: ‘‘Should we use landscape visualisations to engage the emotions and influence behaviour, and if so, how?’’ Possible approaches are described involving ethical standards for use of visualisations in a persuasive role on climate change, resulting policy implications, and research priorities. Visualisation can have multiple roles in society: in entertainment, planning and decision-support, education, and research on public perceptions. In this paper, the focus is on visualisations as awareness-building tools and potential change agents in fostering action by both the general public and policy-makers, although with more emphasis on individual behaviour. In the context of climate change, the most urgent applications address the need for behavioural change to mitigate climate change, though the ideas discussed in this paper apply equally to adaptation priorities. While the focus is on landscape visualisation, it is understood that in reality a combination of techniques and

This section addresses the issue of whether landscape visualisation can be expected to impact awareness, individual behaviour, and policy. We briefly consider theoretical concepts, available evidence, and experience from professional practice, relating to people’s responses to general environmental issues presented via landscape visualisation. Then, we consider how this knowledge on visualisation can be applied to the question of people’s responses to climate change. 2.1. Human responses to landscape visualisations Some of the potential benefits promised by landscape visualisation techniques in the arena of awareness-building, behaviour change, and environmental decision-making can be summarised as in Table 1. These benefits include both self-evident capabilities (e.g. the ability to depict alternative future scenarios side by side) and theorised influences on people’s responses (e.g. increased engagement and perceived salience). However, despite the widespread use of landscape visualisations in planning and design, findings on responses to them are generally not scientifically documented or comprehensive (Sheppard, 2001; Lange, 2001;

640

S.R.J. Sheppard / Environmental Science & Policy 8 (2005) 637–654

Table 1 Potentially beneficial attributes of landscape visualisation for promoting environmental awareness and actiona Integration of science and intuition Engagement of lay-people Personal salience Presentation of choices for the future Flexibility of tool a

The combination of the predictive capabilities of modelling and GIS with the intuitive and experientially rich media of photography and realistic representation, with meaningful socio-cultural associations for communities that may help strengthen more informed perspectives in decision-making The attractiveness of virtual reality and its novel applications to conventional planning problems may be helpful in getting multiple stakeholders to engage in public processes (Sheppard and Meitner, 2005; Nicholson-Cole, 2005) The ability to localize and ground the information by detailed depiction of recognisable and well-known sites as they would be seen by local residents or users, as opposed to a detached plan or aerial view or an expert’s conceptualisation The ability to present alternative futures side-by-side and over time, posing ‘what-if’ questions (Steinitz et al., 2003) in the search for preferred or acceptable environmental solutions over the long term Digital visualisation techniques can be modified or customized to emphasize important information or condense complex details, to fit the presentation to the needs and capabilities of the user (Sheppard, 2005)

Potential disadvantages of landscape visualisation are discussed in the text in Section 3.1.

Sheppard and Salter, 2004), and there is clear potential for offsetting disadvantages (discussed below). Most recent visualisation evaluation studies focus on direct self-report evaluation of the tools, and here there is considerable evidence of their usability or perceived communications effectiveness (Bengtsson et al., 1997; AlKodmany, 2000; Appleton and Lovett, 2003; Sheppard and Meitner, 2005), including the ability to engage lay-people (Lewis and Sheppard, in press) (Fig. 3). Some research has evaluated effects on cognition, but mostly with somewhat abstract forms of visualisation (MacEachren and Ganter, 1990; Al-Kodmany, 2000), which can be helpful in explaining concepts, ecological processes, and overall environmental conditions not easily expressed in text or data tables. Winn (1997) has argued that more complex, interactive 3D virtual reality displays may provide qualitatively and quantitatively superior forms of learning on environmental science and global change issues. Researchers have found that visualisation can demonstrate cognitive advantages over other media (e.g. Furness et al., 1998; Danahy, 2001), though MacEachren (2001) notes that learning may be hampered by the virtual devices used to attract and sustain attention. We should also be careful to discriminate between learning quickly (widely reported in practice) and learning correctly, which is seldom measured

(MacEachren and Ganter, 1990; Winn, 1997; Lewis, 2000; Salter, 2005). Daniel and Meitner (2001) have described several studies which show that visualisations can arouse positive or negative emotional reactions in observers. Few studies have attempted to measure the intensity of emotional reaction, though practical experience suggests public reactions to visualisations can sometimes be vehement. A few studies have evaluated the ability of visualisations to match the affective and related experiential and evaluative responses obtained from real world environments (e.g. Bosselmann and Craik, 1987; Sheppard, 1989, 2001; Meitner and Daniel, 1997; Bishop and Rohrmann, 2003), with mixed results. Zube et al. (1982) and Bishop and Rohrmann (2003) nonetheless believe that the greater the realism, the more similar the responses will be to real life. Very few studies have been carried out on the behavioural impacts of landscape visualisation, either during the exposure to the visualisation material or afterwards. Some experiments have evaluated behavioural actions such as pathway choices in a virtual environment (Bishop et al., 2001), or tracked behaviour in collaborative immersive settings during planning workshops (Campbell and Salter, 2004). Orland (1992, p. 259) has speculated that visualisations may be used ‘‘in a more persuasive mode to motivate

Fig. 3. The impact of imagery: substantially increased dialogue was obtained with members of a First Nation community on resource management issues when using realistic landscape visualisations, compared with using simple GIS maps. Credit: John Lewis, CALP, UBC. Courtesy of Cheam Band, BC. Reproduced by permission of Sustainable Forest Management Network.

S.R.J. Sheppard / Environmental Science & Policy 8 (2005) 637–654

people to do something about the impact being represented,’’ as happens with photography, film, and television media. McKenzie-Mohr and Smith (1999) provide numerous examples of visual information which, with other community-based social marketing techniques, can influence sustainable behaviour or adaptation if they are vivid, personal, and concrete. Interactions between cognitive, affective, and behavioural responses may help explain visual triggers to behavioural responses. Slovic et al. (2002, p. 398) state that ‘‘many theorists have given affect a direct and primary role in motivating behaviour.’’ They review evidence that more vivid and sensational narratives associated with feelings rather than cognition have more influence on risk behaviours. This suggests that introducing compelling visual evidence that is readily translatable into personal risks or implications can be more successful in strengthening motivations to change behaviour than cognitive information alone. Maiteny (2002) argues that behavioural change to support sustainability can only take place when the individual has gone through a deep-rooted personal transformation, as in experiential learning processes. Sheppard (2004) and Nicholson-Cole (2005) speculate that the strength and symbolic emotive content of visualisation imagery may motivate people to act in a more sustainable manner. However, it must be recognised that both the forms of and causal influences on environmentally significant behaviour are very complex, varying with pre-determined attitudes, individual capability, context, etc. (Stern, 2000), and therefore the effectiveness of any form of external intervention such as visualisations, on their own, may be limited by other factors. There is also the important issue of how the use of visualisation could influence relevant policy, either directly through presentations to key decision-makers and policymakers or indirectly through public opinion and collective individual actions. Again, however, there is little scientific information on such policy responses in the visualisation literature. There is some evidence from perception studies using visualisation which indicate differing levels of acceptability of various resource management scenarios (e.g. Sheppard and Meitner, 2005; Ribe, 2005), although the practical linkages between judgements of unacceptability in research and actual policy impacts in the real world have not been widely demonstrated. Professional practice does provide anecdotal examples suggesting sometimes profound effects of visualisation on government behaviour: for example, computer visualisations of potential housing development under the Official Community Plan in Maple Ridge, BC, when shown to members of the city council, caused such surprise and negative reaction that the council initiated a process to develop sustainable community plans with a very different footprint. Overall, many dimensions of visualisation use and subsequent human response (especially behavioural) are not well understood, and we should be wary of generalizing

641

too broadly (Bosselmann and Craik, 1987). There would however seem to be advantages in engagement and cognition, with some evidence of emotional arousal and the potential to affect behaviour on environmental issues. 2.2. Applications of landscape visualisation to the problem of climate change So, can visualisation, in combination with other information sources/media, affect human responses to climate change? To address this question systematically, we need to consider the specific context of society’s perceptions of climate change, before developing a conceptual framework of how visualisation may relate to these perceptions. 2.2.1. Perceptions of climate change The potential for visualisation in relation to climate change can be considered in terms of:  perceptual characteristics and difficulties relating to climate change itself;  potential and observed responses of people and agencies to climate change, and observed gaps between awareness and behaviour;  early precedents for responses to imagery and visualisation of climate change-related phenomena. The characteristics of climate change make it difficult to see directly. Carbon dioxide and even the carbon source itself is effectively invisible. There is potential to recognise some quite tangible visual or landscape-related effects of climate change, such as sea-level rise, flooding, or drought (a typology of such effects is provided below). However, Tickell (2002) has described the problems of communicating change which occurs over long time periods and which carries uncertain and uncomfortable future consequences. Winn (1997) and Nicholson-Cole (2005) have described the scale, complexity, and abstraction of climate change as challenges to communication. This is exacerbated by problems of media coverage, ranging from down-playing the issue (Monbiot, 2004) to ‘‘inappropriate use of model predictions, inaccurate representations of physical processes, false associations of unrelated phenomena, and use of emotion-laden language’’ (Furness et al., 1998, p. 19). It is also difficult to relate global or national figures on climate change to local scales (Furness et al., 1998), and there are as yet few reliable predictions for local climate effects or spatially explicit consequences at high levels of resolution, other than for sea-level rise. It is therefore difficult for individuals to relate their actions (good or bad) to the larger context of climate change (Sheppard, 2004; Nicholson-Cole, 2005). Information has begun to flow on the nature of conceptualisations of and human responses to climate change. Several researchers report evidence of persistent

642

S.R.J. Sheppard / Environmental Science & Policy 8 (2005) 637–654

misconceptions of climate change, e.g. confusing clean air or ozone-layer issues with climate change (e.g. Furness et al., 1998; Boardman and Palmer, 2003). Kollmuss and Agyeman (2002) and Lorenzoni and Langford (2001) describe various types of barrier or gaps between perceptions/intentions and behaviour conducive to climate change mitigation, including gaps in cognition and awareness (ignorance), gaps between knowledge and action, and gaps between intent and action. Lorenzoni and Langford (2001) identify different perspectives that people adopt when confronted by climate change issues, including denial, doubt, disinterest, and engagement. Other responses include confusion, pessimism, guilt, frustration, and, more rarely, motivation to change behaviour (Kollmuss and Agyeman, 2002; Maiteny, 2002). Nicholson-Cole (2005) found that a sample of Norfolk residents expressed their feelings about climate change in a somewhat negative but abstract, distant sense, though personal experiences and impacts on local environments were a common feature of people’s visual conceptions of climate change. Her results highlight the influence of popular visual media on the mental imagery of climate change that respondents held (Nicholson-Cole, 2005), and suggest, in particular, that respondents were most emotionally affected by national and local imagery, rather than global. Gaps in behavioural motivation seemed to be related to difficulties in personalising climate change (Nicholson-Cole, 2005); respondents felt that they needed to understand the personal implications of climate change and be regularly reminded if they were to act. However, it should be noted that these findings represent self-report data on behavioural intent, and not actual measured behaviour. In the context of climate change, visual communications are beginning to be used to accelerate social learning, and the possibility of their motivating the substantial policy, technological, and life-style changes needed has begun to be recognised. Cohen (1997) used GIS and remote sensing imagery with other information to communicate climate change scenarios to Canadian stakeholders, who reported that the scenarios made a difference in their visions of the future and potential interventions in policy debates. Scientific agencies provide animated 3D visualisations of modelled climate change phenomena seen from space, such as changes in sea ice over time (Fig. 4), on the web; similar visualisations of the ‘ozone hole’ are commonly credited with helping to spur legislation to ban fluorocarbons. New books on climate change highlight the impacts of climate change as illustrated with vivid landscape photography of changing iconic landscapes such as snow-capped volcanoes and glaciers (Lynas, 2004). There has been much debate over the recent film entitled ‘‘The Day After Tomorrow’’ which dramatically depicts fictionalized climate change events (http://www.nydailynews.com/entertainment/story/ 189422p-163828c.html). Lowe et al. (2005) found that the film, which included spectacular ‘realistic’ visualisations of weather extremes and sea-level rise threatening the characters, did result in changes in attitude towards climate

Fig. 4. Computer visualisations of simulated summer-averaged sea-ice thickness at the South Pole over a 150 year period, where white represents the thickest ice and the red line depicts the maximum extent of 1cm thick sea-ice. Credit: Copyright, # 2002, University Corporation for Atmospheric Research, http://www.vets.ucar.edu/vg/seaice/index.shtml. Scientific credits: Mark Holland, CGD/NCAR; Cecilia Bitz, U. Washington; Bruce Briegleb, CGD/NCAR; Elizabeth Hunke, LANL; Bill Lipscomb, LANL; Richard Moritz, U. Washington; Julie Schramm, CGD/NCAR. Visualization credits: Fred Clare, NCAR/SCD; John Clyne, NCAR/SCD; Tim Scheitlin, NCAR/SCD. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of the article.)

change, with significantly more concern, anxiety, and motivation to change behaviour, though the latter effect was short-lived. It appeared, however, that the disaster film genre, overly dramatic special effects, and the failure to distinguish fact from fiction amid the high realism, led to low credibility of the film itself, making inferences for more scientifically based visualisation difficult to draw.

S.R.J. Sheppard / Environmental Science & Policy 8 (2005) 637–654

Few visualisations of climate change effects on the landscape have yet been produced to scientific standards, and even fewer have yielded results on responses. The feasibility of using realistic landscape visualisations to depict climate change is not in question: it can be done quite simply using limited 2D perspective imagery with photoimaging software to depict specific possible consequences of climate change, or by more sophisticated 3D/4D modelling of future scenarios with multiple attributes and visualisation inputs. The ease of manipulating images with convincing photo-realism is demonstrated in Fig. 5, which depicts a

643

hypothetical mitigation strategy to reduce fossil fuel usage, seen from within a real community. Fig. 6 provides examples of 3D modelling with semi-realistic visualisation software developed for the Royal Commission on Environmental Pollution, to depict a hypothetical British landscape in existing conditions and with alternate mitigation strategies, here seen in birds-eye view. Dockerty et al. (2005) describe one of the first studies to visualise scientifically modelled effects of climate change in an actual rural landscape in Norfolk, at the detailed site level. Here, local scenarios visualised were developed from the

Fig. 5. Photo-realistic landscape visualisations comparing existing conditions and a hypothetical low-carbon future in southern England. Credit: Cecilia Achiam, CALP, UBC.

644

S.R.J. Sheppard / Environmental Science & Policy 8 (2005) 637–654

four basic socio-economic scenarios described by Nakicenovic and Swart (2000), by applying an agricultural land use model and locally specific decision-rules and assumptions, to develop high-resolution 3D visualisations showing changes in the farming landscape (Fig. 7). However, no studies of formally collected responses to these visualisations have yet been published. In other visualisation-based research on responses to long-range scenarios for forest management planning (Sheppard and Meitner, 2005), rural community members questioned the failure to build in climate change factors to such models, raising concerns over modelling credibility. Schroth et al. (2005) used interactive 3D aerial views of semi-abstract 3D visualisations of the Entlebuch Biosphere Reserve in Switzerland, to assess qualitatively responses to retreating snow cover on winter recreation under climate change modelling assumptions based on topographic elevation (Fig. 8). When shown to community stakeholders, the visualisations caused surprise and disquiet, which led quickly to a new and different discussion of summer recreation opportunities (Schroth et al., 2005), suggesting that the medium can stimulate a rapid adaptation response.

Fig. 6. Hypothetical British landscape visualised in existing conditions and with alternate future scenarios including mitigation strategies featuring wind-turbines and biofuels plantations, and varying levels of development. The two future scenarios represent two major energy choices (S1 – energy demands stabilized at 1998 levels; S4 – energy demands at half the 1998 levels*), affecting the need for large-scale renewable or nuclear energy plants that adversely impact on landscape and seascape character, and the capacity of small-scale generation to cope with local energy demands. Credit: Images from ‘‘Visualizing renewable Energy in the Landscape of 2050’’: Copyright of The Countryside Agency. Images by ethos-uk.com. *S4 scenario uses energy projections from the Royal Commission for Environmental Pollution.

2.2.2. Towards a theoretical framework for visualisation of climate change Based on a synthesis and simplification of the available evidence and theoretical arguments discussed above, Fig. 9 presents a conceptual diagram in the form of a spectrum of possible individual responses to visual information on climate change, representing different levels of carbon consciousness and commitment to action on climate change issues. We can envisage the range of perceptual and behavioural responses to visual information in the form of an ordered categorisation moving from a low state of awareness through to action, which corresponds generally with a progression from cognitive processing of information, through emotional responses, to behaviour change. It is not assumed that each stage on the spectrum is a necessary precursor for response types further along the spectrum, or that over time any individual will necessarily move through the range of responses in this specific linear sequence. For example, emotional effects may actually precede cognitive effects from a given presentation, or occur at the same time. The diagram is presented as a simple way of structuring our thinking and emerging research programmes on the range of response types and effects we may look for as the result of applying visualisations to issues of climate change. It is recognised that there are many other models of the motivations and controlling factors on pro-environmental behaviour, which may not be reflected in this simple framework. The responses illustrated here can be related to other systems of categorising types of respondents, e.g. Lorenzoni and Langford’s (2001) classification of the deniers, uninterested, doubters, and engaged; or to Kollmuss and Agyeman’s (2002) analysis of barriers to pro-environmental

S.R.J. Sheppard / Environmental Science & Policy 8 (2005) 637–654

645

Fig. 8. Semi-realistic 3D visualisations of forecasted snow conditions and existing ski runs in the Sorenberg area of Switzerland, under current conditions and after 50 years with climate change, with safe snow elevations for skiing depicted in white. Credit: Olaf Schroth/VisuLands 2005, #Geodata: RAWI Lucerne.

Fig. 7. Model-derived 3D visualisations of existing conditions in a Norfolk agricultural area, with two scenarios under climate change in 2020: a low mitigation, high carbon emissions scenario (A2) and a local stewardship scenario with lower carbon emissions (B2). Credit: Katy Appleton, University of East Anglia. Reproduced from Dockerty et al., 2005, courtesy of Computers, Environment and Urban Systems.

behaviour. The gaps between the stages along the spectrum represent some of the main gaps between knowledge and action. However, for visualisation to be judged effective in addressing climate change, viewers do not need to move along the entire length of the spectrum; if visualisation succeeds at any one of these levels, e.g. in reducing ignorance or denial, then this should be beneficial, particularly if it is more

effective than other forms of communication. The response spectrum suggests that the early stages along the spectrum may be more straightforward to achieve with visualisation than the upper stages: the ability of pictures to convey information is well accepted, whereas the ability to impact behaviour is much less certain. In order to bridge the various kinds of perceptual or behavioural gap shown in Fig. 9, it would seem that different forms of visualisation may be required, as suggested in Fig. 10. In improving understanding of climate change, we may expect that more conceptual 3D visualisations may be appropriate, through simplification and focus on the most cogent information such as overall CO2 cycles, shifts in biogeoclimatic zones, or cumulative effects. Cognitively effective landscape visualisation might focus on augmenting reality to make the invisible visible, making the abstract tangible, collapsing long time scales into short periods, and easily switching between different scales (Winn, 1997; Furness et al., 1998). Clarity of message and of depicted conditions appears important in effective cognition (Sheppard, 1989; Nicholson-Cole, 2005). While realism may not be necessary for (and may in some cases detract from) cognition, Furness et al. (1998, p. 13) believe that, in most

646

S.R.J. Sheppard / Environmental Science & Policy 8 (2005) 637–654

environmental psychology literature that familiar landscapes tend to be associated with stronger and more positive affective responses (e.g. Kaplan and Herbert, 1988), and as noted above, people seem most affected by personal implications of climate change (McKenzieMohr and Smith, 1999; Nicholson-Cole, 2005).  Immediacy: near-term conditions (Lorenzoni and Langford, 2001) or possibly longer term conditions made to seem nearer term through speeding up time, combined with meaningful future considerations such as their neighbourhood as seen by the viewers’ grand-children.  Containing images of people, animals, or other symbols with strong affective content (Nicholson-Cole, 2005).  Demonstrating the future consequences of people’s actions or inactions (Furness et al., 1998). In theory, behaviour change may require emotional and cognitive attributes described above, but intensified to be vivid, memorable, and transformative (McKenzie-Mohr and Smith, 1999; Maiteny, 2002): ‘‘a compelling virtual environment will likely heighten (people’s) motivation to act responsibly’’ (Furness et al., 1998, p. 28). Both Winn (1997) and Furness et al. (1998) argue that vivid and compelling visualisations can be obtained through a range of techniques, notably:

Fig. 9. A theoretical spectrum of human responses to visual information on climate change.

cases, ‘‘. . . realistic representation will lead to more direct and more robust knowledge construction.’’ Experience has already shown that the novelty or interest in computer visualisations can attract people and engage them in collaborative learning processes (Winn, 1997; Campbell and Salter, 2004; Sheppard and Meitner, 2005). In order to reach the emotional side of viewers, the following additional attributes of visualisation would appear to be important:  Realism, in the sense of photo-realistic or ‘lifelike’ imagery in re-creating experiential qualities (Appleyard, 1977) and making abstract concepts ‘concrete’’ (McKenzie-Mohr and Smith, 1999).  Depicting personally relevant environments, such as local and recognisable neighbourhoods (Nicholson-Cole, 2005), or iconic, well-known landscape symbols to which people can relate (Sheppard, 2004). It is well known in the

 immersion in a virtual environment: large images and panoramic ‘wrap-around’ displays can increase the sense of presence (Furness et al., 1998), engagement (Appleyard, 1977), and intensity of experience (Sheppard et al., 2001);  dynamic or animated imagery that increases enthusiasm and engagement (e.g. Dykes, 2000) and/or provides freedom of virtual movement for the viewer (Orland and Uusitalo, 2001);  interactivity with the displayed data in real-time, to increase engagement (Orland and Uusitalo, 2001; Campbell and Salter, 2004; Schroth et al., 2005). Beyond media attributes, the content of the climate change message also appears critical if behaviour is to be influenced. McKenzie-Mohr and Smith (1999) argue that messages which emphasize environmental losses due to inaction are consistently more persuasive than those which simply emphasize benefits of action; such threats should be combined with positive implications of action to overcome the barrier of helplessness which might otherwise afflict viewers (Kollmuss and Agyeman, 2002). Nicholson-Cole (2005) stresses the importance of tailoring the visualisation material to the target audience in light of the variability in human responses: this can be interpreted as identifying the most applicable stage on the response spectrum for the target audience in selecting the appropriate visualisation stimulus. A more systematic discussion of approaches to the content of visualisations addressing aspects of climate change is provided in Section 3.2.

S.R.J. Sheppard / Environmental Science & Policy 8 (2005) 637–654

647

Fig. 10. Theoretical effects of different types of landscape visualisations in stimulating perceptions and behaviour in response to climate change.

In summary, therefore, there is strong evidence of the cognitive effectiveness of visualising global change, and a strong likelihood of the ability to engender emotional responses. It also appears theoretically possible that the use of landscape visualisation could affect behaviour with regard to climate change, if the imagery provides: (1) disclosure: a window into the future which is personally meaningful and tangible, making the global both local and personal, showing possible negative and positive outcomes; (2) drama: a vivid and compelling presentation with emotional content. However, the behavioural part of the theoretical response spectrum (Fig. 10) is mostly in doubt and in need of testing. Developing theories on the potential of visualisations to influence attitudes and behaviour must also recognise that their effectiveness would be dependent on many factors including the intended purpose of the exercise (i.e. nature of response sought), sociocultural and environmental context, the type of audience, and the contribution of other forms of information.

3. Ethical and policy issues of using visualisation to influence perceptions and behaviour on climate change In this section we turn our attention to the ethical questions: ‘Should we use visualisations for the purpose of persuasion, and if so, how?’ Is there a ‘right’ way to do this? 3.1. Risks and dilemmas in using visualisation of climate change Attempts deliberately to harness the power of visualisations to influence behaviour, what Luymes (2001) describes as the rhetoric of visual simulation, have radical policy implications in terms of visualisation use and the role of the presenter. The conventional role of visualisation as an informative tool in decision support is associated with the supposed neutral role of science in not imposing value judgements on the public.

648

S.R.J. Sheppard / Environmental Science & Policy 8 (2005) 637–654

Arguments for a more deliberate attempt to use visualisation to influence the public or impact government policy emphasize the need to forestall an actual crisis in the environment as an over-riding imperative. A persuasive approach should not necessarily be equated with inaccurate or distorted visualisations: visualisations properly prepared can simply disclose the truth, which may itself be persuasive. In fact, it could be argued that visualisations which disclose possible futures are less misleading than not using them, and there may be an obligation to disclose the truth to those who cannot or will not see. Luymes (2001) has advocated the use of powerful visualisation tools to shape public values on sustainability. History is replete with examples of the inability of cultures to foresee the impact of their actions or inaction: e.g. the use of DDT, fluorocarbons, and indeed carbon emissions. Precedents do exist for other kinds of long-term predictions and disclosure of consequences, including environmental impact assessment, multiple-rotation forest modelling, and biodiversity modelling in response to climate change (Berry et al., 2002). There is also an argument that scientifically produced visualisations are needed to counter the misinformation propagated by entertainment media such as the recent film ‘‘The Day After Tomorrow,’’ while utilizing a similar medium. More generally, Michaelis (2003) has argued for a government strategy to change public behaviour on greenhouse gas emissions through a process of leadership, dialogue, and facilitation: there would appear to be strong role for visualisation in helping project ‘‘visions for a sustainable way of living’’ (Michaelis, 2003, p. S143). This author concludes that the persuasive use of visualisations, together with other tools and approaches, is justified if they can be effective, and may even be vital in communicating climate change urgently. We should take the perceptual leap and seek to engage the public’s emotions. This seems worth doing even if behaviour does not change immediately, by engaging people to think more seriously about the future and showing them they have a choice among alternative futures. This at least would remove the first barrier of ignorance on the response spectrum of carbon consciousness and action (Fig. 9). Once the decision is taken that a policy of persuasive use of powerful visual imagery can be justified in the fight against climate change, a second tier of issues becomes evident. What risks do we take in such usage, how can they be minimized, and are the resulting risks worth taking? There are many process decisions and constraints in the production of any realistic simulated landscape imagery, and many different problems can result (Sheppard, 1989). There is space here only for a brief discussion of some of these, as follows: 1. The risk of biased responses: How can unintentional bias or deliberate attempts to mislead be prevented? (McQuillan, 1998; Orland et al., 2001). Winn (1997) points out that virtual reality media may create their own

misconceptions. There is the risk of aesthetic responses to visual conditions over-riding more important but nonvisible implications of climate change. Both Luymes (2001) and Orland and Uusitalo (2001) express concerns about the high credibility and low transparency of realistic immersive forms of virtual reality, which foster sometimes false assumptions of authority, and which encourage suspension of disbelief in the virtual worlds depicted. These problems can be increased when deliberate efforts are made to engage the emotions (Slovic et al., 2002). 2. The risk of disbelief: Lack of credibility (Sheppard, 1989) of the visualisation imagery or the underlying modelling/ assumptions would seriously damage the effectiveness of the visualisation exercise. Sheppard and Meitner (2005) and Schroth et al. (2005) record participants’ concerns about apparent or potential manipulation of visualisation imagery to support a particular message. Realistic visualisations of possible climate change effects (Fig. 11) which are presented without corresponding scientific explanations may not be taken seriously. Even with such information, the necessary disclosure of massive uncertainty (Webster, 2003) in climate change and land use projections could itself trigger lack of confidence in the scientists and their models. 3. The risk of confusion: The enormous complexity of climate change and associated policy options (Keeney and McDaniels, 2001) could simply confuse people over the number of visualised scenarios, contingencies, associated risks, and consequent choices. 4. The risk of overkill: Even visual information can become too much, either through information overload (Orford et al., 1999) or sheer habituation to what becomes mundane (McKenzie-Mohr and Smith, 1999). There may

Fig. 11. Examples of photo-simulations of hypothetical sea-level rise in Holland in 2020, used in public communications. Credit: Stone/Getty Images; courtesy of The Guardian ‘‘The Drowned World’’ September 11, 2004; and Glen Taylor.

S.R.J. Sheppard / Environmental Science & Policy 8 (2005) 637–654

be different desired levels of stimulus and response intensity for different purposes, e.g. multiple iterative exposures to promote better learning versus single sharp exposures for shock value. Winn (1997) has raised the fear of the virtual reality medium itself being ‘overhyped’ and losing effectiveness. 5. The risk of upsetting people: There are differing views on the desirability and effectiveness of emotional content (such as fear or unease), leading to negative or defensive reactions to the issue (Nicholson-Cole, 2005), versus judicious use to trigger behavioural change (McKenzieMohr and Smith, 1999). Furness et al. (1998) note that students who become personally engaged in serious consideration of global change may become very distressed, raising conflicts between the goal of ‘truthfulness’ and the desire not to upset people. 6. The risk of perpetuating the problem by encouraging acceptance of climate change: What if the visualisation imagery is valid but simply too subtle to influence attitudes, or depicts positive imagery that may result from some aspect of a given climate change scenario? Dockerty et al. (2005) have generated landscape visualisations for Norfolk (Fig. 7) showing what some may consider to be attractive fields of sunflowers which would replace the existing agricultural crops. 3.2. Ethical standards and policy implications Given the issues just raised, how should we embark upon a policy of deliberate use of visualisations to influence attitudes and behaviour on climate change? The following sections briefly explore three topics key to the ethical implementation of visualisations of climate change. 3.2.1. Defensible methodologies for preparing visualisations An obvious starting point would be to develop guidelines for uses of visualisation related to climate change, appropriate to the needs of public motivation and decision-making. This would address the principal missing ingredient beyond disclosure and drama for effective visualisation of climate change: defensibility. If there is no overall scientific or logical underpinning of the visualisations, they are unlikely to change peoples’ minds or convince policy-makers suspicious of new media. However, there is little guidance on how much manipulation or selectivity should be allowed to create the appropriate kinds of drama while avoiding the pitfalls described in Section 3.1. What accuracy or response validity standards can be applied when visualisations represent predictions of unknowable futures? Should accuracy be measured against likelihood or feasibility of occurrence, or as veracity to the modelled outcomes? To what degree would existing conventions developed for typical uses of landscape visualisation in design and planning apply to the special circumstances of climate change mitigation and adaptation?

649

Sheppard (2001, 2005) has proposed a code of ethics for visualisation which identifies the following principles or criteria that may be relevant to climate change applications: accuracy of visualisation relative to expected conditions; representativeness of views in space and time; visual clarity of presentations; interest and engagement of the audience/ users; legitimacy or accountability of the visualisation, including transparency of data and of the production process, and delivery of visualisations by a trusted source (Sheppard and Meitner, 2005; Nicholson-Cole, 2005); and accessibility of the visualisations to the public and potential users. Other ethical and policy issues of particular concern to climate change are likely to include:  Representing uncertainty, e.g. through presentation devices (Dockerty et al., 2005) or through multiple alternative visualisations per given scenario as a form of sensitivity testing (Appleyard, 1977; Sheppard, 2001, 2004).  Down-scaling from global, regional, or even local climate change scenarios: we need robust ways of developing decision rules for visualising the detailed landscape conditions necessary for personal engagement and local recognition, but based on very broad land use or environmental scenarios with limited resolution data-sets (e.g. UKCIP, 2000b; Dockerty et al., 2005).  Presentation of non-visible conditions (in the visualisation and with other media) to mediate and augment the visible aspects of climate change, through a blend (Fig. 12) of realism (in this case, individual trees on the hillside) and abstraction (e.g. false colour).  Permissible types of drama, including dramatic content such as forest die-back or the effects of extreme storms, dramatic viewing conditions such as animated ‘fly-bys’ and unusual lighting, or dramatic display formats such as stereo vision, big screens, and panoramic angles of view. One defensible policy on permissible drama would perhaps be that a defined ‘reality’ remains the guide: for example, animated large-screen panoramic imagery of a future landscape may be both dramatic and legitimate if it conveys the expected appearance at the actual image size and angle of view with which it would be seen in the neighbourhood, but not if it exaggerated the effects of climate change, gave a magnified view of the landscape features affected, or selectively left out key elements. However, the type of decision or response required does need to be taken into account, and cognitive awareness, as suggested in Fig. 10, may require the very types of omission or abstraction that could be misleading with the experiential response. Ensuring that multiple media are used, retaining the cognitive information along with the realistic experiential media (e.g. 3D landscape models augmented with draped polygons and labels) may reduce the risk, together with disclosing any manipulation of suppression of data. The

650

S.R.J. Sheppard / Environmental Science & Policy 8 (2005) 637–654

Fig. 12. Non-visible conditions (in this case tree species changing over time on a BC mountainside) can be displayed with colour-coding in detailed landscape visualisations to express important environmental changes due to policy implementation or climate change, even though the actual landscape appearance might not change significantly. Credit: Jon Salter, CALP, UBC.

aims of the exercise (e.g. fostering sustainable behaviour) should also be made explicit at the outset. Allowing the sceptical user to navigate and interrogate the visualisation imagery and underlying databases, and to choose their own viewing conditions or view sequence, may help in assuring defensibility (Furness et al., 1998; Sheppard and Salter, 2004). Another strategy for building defensibility would be to ensure effective stakeholder participation in the development of socio-economic scenarios, the application of decision-rules for visualising the scenarios, and even the process of generating the visualisations (Sheppard and Salter, 2004). 3.2.2. The content of visualisations of climate change scenarios The selection of appropriate global, regional, and local scenarios to visualise would be a key strategic decision. While the IPCC (Nakicenovic and Swart, 2000) and various levels of national and regional governments have developed generic alternative scenarios for future climate change and adaptation (e.g. UKCIP, 2000b), systematic mechanisms for downscaling these to local landscapes are needed. This usually means accepting the limited range and assumptions inherent in the internationally or nationally selected generic scenarios, although none of the seminal IPCC scenarios currently contemplates substantial short-term reductions in carbon emissions to reduce future climate change risks (UKCIP, 2002). Visualisation of these scenarios alone might be mistaken by the public as expressing the full range of possible solutions, rather than what some may see as a limited and politically influenced subset. Such scenarios do not articulate estimated long-term benefits of a radical short-term approach to climate change mitigation, relative to a slower response which effectively endorses additional discretionary carbon

emissions. Mechanisms for developing and analyzing additional visualisation scenarios which can be derived independently from the global scenarios might therefore be strategically advisable. There may be other reasons to develop independent scenarios for visualisation purposes. These could address specific local environmental or cultural issues which would be more meaningful to local planners and communities, and require less work to link to much larger multi-dimensional global or regional scenarios and modelling systems. The challenge would lie in still having these visual scenarios be credible. One approach would involve local stakeholders in the scenario generation process, though this may become constrained by the participants’ inability or unwillingness to consider the more radical alternatives posed by climate change (Berry, 2005). Another possibility is to assume certain likely conditions, based on prior scientific analysis, without detailed linkages of visualisation to complex modelling systems. For example, there is a strong likelihood in the dry Okanagan basin of BC that increasing drought conditions will continue or get worse under regional climate predictions (Cohen et al., 2004), leading to more widespread forest die-back; even without specific modelling, visualisation that shows the possibility of 80% die-off of forest trees at some point in the lifetime of current middle-aged residents of the area, generated with the advice of recognised experts and with appropriate caveats on the general uncertainties, may not be too speculative. In such circumstances, there is a limited set of visual possibilities: more of the trees turn brown or all the trees turn brown. It may be sufficient to say: ‘we are not sure which of these futures will occur, but you should be prepared for all of them,’ in order to stimulate adaptation behaviour. Much more thought needs to be given to this issue, and where to

S.R.J. Sheppard / Environmental Science & Policy 8 (2005) 637–654

651

Table 2 An initial typology of selected climate change impacts in terms of their potential to be readily and realistically visualised Climate change impact type

Capability to be visualised realisticallya Often readily apparent or imageable to lay-viewer at landscape level

Sea-level rise (coastline) Permanent flooding Seasonal flooding Changes in seasonal patterns/timing Increased storm severity Ice retreat Snow depth reduction Soil erosion from concentrated precipitation Stream character change (drying up, loss of riparian vegetation, etc.) Lake level drop Drought induced vegetation stress Drought induced vegetation die-back Vegetation succession/invasion from biome shift Desiccation of forest floor Increased forest disease epidemics Increased vegetation fire risk Fire events Loss of rare plant species Loss of wildlife species (through extreme climatic events or long-term shifts) Crop failures Farm abandonment Cultural event disruption (e.g. seasonally disrupted traditional ceremonies) Urban landscaping stress Urban brown-outs

Visually subtle or invisible (requiring augmented realism) at landscape level

H H H H H H H

H H H

H H H H

H H

H H H H H H H H H H

a Some climate change impacts vary widely in their visibility, and are noted in both columns. Some impacts may or may not be realistically imageable depending on whether time-lapse visualisations are used.

draw the line between defensible (though not necessarily quantifiable) depictions of possible futures, and less scientific, alarmist scenarios. A range of climate change impacts would lend themselves to disclosure in visualisations of a future local landscape, depending on the geographic conditions of the area. Table 2 provides an initial typology of climate change effects, with an initial categorisation of how readily they could be expressed in realistic landscape visualisations: an initial indication of their potential for visual drama or meaningful communication in that medium. One use of such a classification would be to identify those phenomena which are not well suited to portrayal by landscape visualisations and where other media need to be used. Issues in attempting to represent future climate-influenced conditions in visualisations include conditions or events which have occurred before but which now occur with different frequency or timing, and the importance of threshold conditions (e.g. tree death and dam-failure) that are more visibly demonstrated than more subtle and gradual changes. In the context of motivating behavioural change at the policy level, the potential content of visualisation imagery need not be limited to the direct impacts of climate change on the local environment and community (as shown in Table 2 and Fig. 8). It might also be effective cognitively and affectively to visualise:

 the root causes of climate change and sources of carbon entering the ecosystem cycle, e.g. oil wells and coal mines;  mechanisms of GHG emission into the atmosphere, e.g. powerplants, vehicles, and houses;  indirect impact through adaptation, e.g. raised dykes for flood abatement, changed agricultural land-use, or logging to reduce fire-risk;  mitigation measures to reduce climate change and CO2 emissions, e.g. proposed low energy transportation systems, dispersed renewable energy sources, houses with energy-generating technology, etc. (Fig. 5). 3.3. Research priorities Research programmes to support such usage of visualisation would need to address: 1. Empirical testing of visualisation to support or adapt the theoretical framework of effects on the response spectrum, and especially, effects on behaviour. This should combine more psychological research on climate change perception and behaviour (e.g. Lorenzoni and Langford, 2001) with classical environmental perception research methods using controlled visual landscape imagery as stimuli (e.g. Daniel and Meitner, 2001).

652

S.R.J. Sheppard / Environmental Science & Policy 8 (2005) 637–654

2. Exploration of ethical issues through policy debate and initial case studies developed especially to illuminate these difficult questions. 3. Scientific monitoring of real-world projects or processes of long-term planning, resource modelling, and climate change adaptation, where these address alternative futures as mediated by landscape visualisation-based tools or related visual imagery. This would track actual consequences of the use of visualisations on learning and behaviour. Such a combined programme would help to identify some of the underlying triggers for attitudinal or behavioural change on climate change policy and life-styles, as well as risks and benefits of particular visualisation approaches. It would also be instructive to evaluate the effect of visualisations relative to other policy and communication tools. Visualisation research could be built into ongoing studies focused on people’s reactions to possible climate change strategies and mitigation approaches, e.g. alternative technologies for house construction and energy sources (Fawcett et al., 2002). With an appropriate research design, landscape visualisations (e.g. Fig. 5) being used as surrogates for real world conditions in determining preferences or acceptability of possible climate change strategies could provide additional research results on the performance/effects of alternative visualisation methods. Research programmes which are using landscape visualisation stimuli for perception research on planning issues could also be retuned to address climate change issues.

4. Conclusions There is an alarming gap between awareness and action on climate change. Fawcett et al. (2002) admit that ‘‘Appropriate policy on energy use and carbon emissions would only have a small part to play in the social revolution which would be needed to achieve voluntary acceptance of reduced consumption levels’’ (p. 80). Emerging techniques of landscape visualisation promise considerable improvements over other media in communicating some climate change issues, fostering social learning, and possibly in influencing popular perceptions and behaviour on climate change. There is considerable evidence of the effectiveness of visualisation as a planning tool and its ability to enhance cognition, but less research has taken place on responses to realistic landscape visualisations, especially their behavioural impacts. However, experience in practice suggests that landscape visualisations can sometimes have substantial effects on policy, and there is evidence with other visual media of behavioural effects through engaging emotional responses. New immersive and interactive systems offer attractive possibilities for engagement, awareness building, and reaching the emotions of the viewer; however, there are

ethical concerns including unrecognised bias in using such tools. Despite these issues, the rhetoric of visualisation (Luymes, 2001), applied deliberately and defensibly as a persuasive force, may be vital in helping to accelerate social learning, engage the emotions, influence individual behaviour, and project the rationale for policy change. The hypothetical mechanism is realistically and dramatically to bring home the impacts of future climate change scenarios to people in their back yard or in other meaningful and iconic landscapes: ‘making climate change personal.’ A theoretical framework is needed for examining responses to visualisations of climate change, addressing their ability to influence people’s hearts, minds, and behaviour. One possible framework has been outlined in this paper. However, even if the theorised effects of visualisation on behaviour are not established and the framework presented here is not borne out, there may be advantages in using visualisation to stimulate awareness and deliberative discourse. We should test carefully every potentially powerful weapon in the fight against climate change, especially those which promise rapid results. Visualisation tools are potentially too powerful either to be ignored or used without careful consideration. The hope is that the dilemmas of whether and how to use visualisations proactively as agents of learning and behavioural change, can be resolved through development of ethical standards incorporating the ‘3 Ds’ of visualising climate change: disclosure, drama, and defensibility.

Acknowledgments I would like to acknowledge the initial inspiration for this paper and the subsequent review from Dr. David Little of Arthur D. Little Limited. Dr. Russ Parsons also provided helpful comments on early drafts. I would like to thank Green College and the Environmental Change Institute at Oxford University, and members of The Zuckerman Institute for Connective Environmental Research (ZICER) at University of East Anglia, for providing venues and/or vital information contributing to the development of this paper. Thanks also to Cecilia Achiam for her graphics and guidance, and to the many CALP members who have supported me. The preparation of the manuscript was supported in part by sabbatical funding from the Sustainable Forest Management Network, Canada.

References Adger, W.N., 2003. Social capital, collective action, and adaptation to climate change. Econ. Geogr. 79 (4), 387–404. Al-Kodmany, K., 2000. GIS in the urban landscape: reconfiguring neighbourhood planning and design processes. Landsc. Res. 25 (1), 5–28. Appleton, K., Lovett, A., Sunnenberg, G., Dockerty, T., 2002. Rural landscape visualisation from GIS databases: a comparison of approaches, options, and problems. Comput. Environ. Urban Syst. 26, 201–211.

S.R.J. Sheppard / Environmental Science & Policy 8 (2005) 637–654 Appleton, K., Lovett, A., 2003. GIS-based visualisation of rural landscapes: defining ‘sufficient’ realism for environmental decision-making. Landsc. Urban Plann. 65, 117–131. Appleyard, D., 1977. Understanding professional media: issues, theory, and a research agenda. In: Altman, I., Wohlwill, J.F. (Eds.), Human Behavior and Environment, vol. 1. Plenum Press, New York, NY, pp. 43–88. Bengtsson, P., Johansson, C.R., Akselsson, K.R., 1997. Planning working environment and production by using paper drawings and computer animation. Ergonomics 40, 334–347. Berry, P.M., 2005. Personal communication. Berry, P.M., Dawson, T.P., Harrison, P.A., Pearson, R.G., 2002. Modelling potential impacts of climate change on the bioclimatic envelope of species in Britain and Ireland. Global Ecol. Biogeogr. 11 (6), 453–462. Bishop, I.D., Rohrmann, B., 2003. Subjective responses to simulated and real environments: a comparison. Landsc. Urban Plann. 65, 261–277. Bishop, I.D., Ye, W.S., Karadaglis, C., 2001. Experiential approaches to perception response in virtual worlds. Landsc. Urban Plann. 54, 115– 123. Boardman, B., Palmer, J., 2003. 4CE—Consumer Choice and Carbon Consciousness: Electricity Disclosure in Europe. 4CE Final Report. Environmental Change Institute, Oxford, UK, 81 pp. Bosselmann, P., 1998. Representation of Places: Reality and Realism in City Design. University of California Press, Berkeley, CA. Bosselmann, P., Craik, K.H., 1987. Perceptual simulations of environments. In: Bechtel, R.B., Marcus, R.W., Michelson, W. (Eds.), Methods in Environmental and Behaviour Research. Van Nostrand Reinhold, New York, NY (Chapter 5). Campbell, D.C., Salter, J., 2004. The Digital Workshop: Exploring the Effectiveness of Interactive Visualisations and Real-Time Data Analysis in Enhancing Participation in Planning Processes. Report Submitted to Forestry Innovation Investment Forestry Research Programme, Deliverable #5. Collaborative for Advanced Landscape Planning, UBC, Vancouver, BC, 32 pp. Cohen, S.J., 1997. Scientist–stakeholder collaboration in integrated assessment of climate change: lessons from a case study of Northwest Canada. Environ. Model. Assess. 2, 281–293. Cohen, S.J., Nielsen, D., Welbourn, R. (Eds.), 2004. Expanding the Dialogue on Climate Change and Water Management in the Okanagan Basin, BC. Final Report, Environment Canada, Agriculture and AgriFood Canada and University of British Columbia. Project A463/433. Submitted to the Adaptation Liaison Office, Climate Change Impacts and Adaptation Directorate, Natural Resources Canada, Ottawa, 300 pp. Cox, D.J., 1990. The art of scientific visualization. Acad. Comput. 46, 20– 56. Danahy, J.W., 2001. Technology for dynamic viewing and peripheral vision in landscape visualization. Landsc. Urban Plann. 54, 125–137. Daniel, T.C., Meitner, M.J., 2001. Representational validity of landscape visualizations: the effect of graphical realism on perceived scenic beauty of forest vistas. J. Environ. Psychol. 21 (1), 61–72. Dockerty, T., Lovett, A., Sunnenberg, G., Appleton, K., Parry, M., 2005. Visualising the potential impacts of climate change on rural landscapes. Comput. Environ. Urban Syst. 29 (3), 297–320. Dykes, J., 2000. An approach to virtual environments for visualization using linked geo-referenced panoramic imagery. Comput. Environ. Urban Syst. 24, 127–152. Ervin, S.M., Hasbrouck, H.H., 2001. Landscape Modeling: Digital Techniques for Landscape Visualization. McGraw-Hill, New York, NY. Fawcett, T., Hurst, A., Boardman, B., 2002. Carbon UK. Industrial Sustainable Development Group, Environmental Change Institute, University of Oxford, Oxford, UK, 65 pp. Furness III, T.A., Winn, W., Yu, R., 1998. The impact of three dimensional immersive virtual environments on modern pedagogy: global change, VR and learning. In: Proceedings of Workshops in Seattle, Washington, and Loughborough, England in May and June 1997. Human Interface Technology Lab, University of Washington, Seattle, WA available online at http://www.hitl.washington.edu/publications/r-97-32/, rtf file accessed June 1, 2004.

653

IPCC (Intergovernmental Panel on Climate Change), 2001. Climate Change 2001: Impacts, Adaptation and Vulnerability. Summary for Policymakers. World Meteorological Organisation, Geneva. Kaplan, R., Herbert, E.J., 1988. Familiarity and preference: a cross-cultural analysis. In: Nasar, J.L. (Ed.), Environmental Aesthetics: Theory, Research, and Applications. Cambridge University Press, New York, NY, pp. 379–389. Keeney, R.L., McDaniels, T.L., 2001. A framework to guide thinking about climate change policies. Risk Anal. 21 (6), 989–1000. Kollmuss, A., Agyeman, J., 2002. Mind the gap: why do people act environmentally and what are the barriers to pro-environment behaviour? Environ. Educ. Res. 8 (3), 239–260. Lange, E., 2001. The limits of realism: perceptions of virtual landscapes. Landsc. Urban Plann. 54, 163–182. Lewis, J.L., 2000. Ancient values, new technology: emerging methods for integrating cultural values in forest management. Unpublished M.Sc. Thesis. Faculty of Forestry, UBC, Vancouver, BC. Lewis, J.L., Sheppard, S.R.J. Culture and communication: can landscape visualisation improve forest management consultation with indigenous communities? Landsc. Urban Plann., in press. Lorenzoni, I., Langford, I.H., 2001. Climate change now and in the future: a mixed methodological study of public perceptions in Norwich (UK). CSERGE Working Paper ECM 01-05. CSERGE, Norwich, UK. Lowe, T., Brown, K., Dessai, S., de Franca Doria, M., Haynes, K., Vincent, K., March 2005. Does tomorrow ever come? Disaster narrative and public perceptions of climate change. Tyndall Centre for Climate Change Research Working Paper 72. University of East Anglia, Norwich. Luymes, D., 2001. The Rhetoric of Visual Simulation in Forest Design: Some Research Directions. International Union of Forest Research Organizations (IUFRO) Research Series. CABI International, Wallingford, UK, pp. 191–204 (Chapter 13). Lynas, M., 2004. High Tide: News from a Warming World. Flamingo, Oxford, UK. MacEachren, A.M., 2001. Cartography and GIS: extending collaborative tools to support virtual teams. Prog. Hum. Geogr. 25 (3), 431–444. MacEachren, A.M., Ganter, J.H., 1990. A pattern identification approach to cartographic visualization. Cartographica 27 (2), 64–81. Maiteny, P.T., 2002. Mind in the gap: summary of research exploring ‘inner’ influences on pro-sustainability learning and behaviour. Environ. Educ. Res. 8 (3), 299–306. McKenzie-Mohr, D., Smith, W., 1999. Fostering Sustainable Behavior: An Introduction to Community-Based Social Marketing. New Society Publishers, Gabriola Island, Canada. McQuillan, A.G., 1998. Honesty and foresight in computer visualizations. J. Forestry 96 (6), 15–16. Meitner, M., Daniel, T., 1997. The effects of animation and interactivity on the validity of human responses to forest data visualizations. In: Orland, B. (Ed.), Proceedings of Data Visualization’97, St. Louis, MO. Michaelis, L., 2003. Sustainable consumption and greenhouse gas mitigation. Climate Policy 3S1, S135–S146. Monbiot, G., 2004. Beware the fossil fools. In: The Guardian, Tuesday, April 27, accessed June 8, http://www.guardian.co.uk/comment/story/ 0,3604,1204194,00.html. Nakicenovic, N., Swart, R. (Eds.), 2000. Emissions Scenarios 2000: Special Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK. Nicholson-Cole, S.A., 2005. Representing climate change futures: a critique on the use of images for visual communication. Comput. Environ. Urban Syst. 29 (3), 255–273. Orford, S., Harris, R., Dorling, D., 1999. Geography: information visualisation in the social sciences. Soc. Sci. Comput. Rev. 17 (3), 289–304. Orland, B., 1992. Evaluating regional changes on the basis of local expectations: a visualisation dilemma. Landsc. Urban Plann. 21, 257–259. Orland, B., Budthimedhee, K., Uusitalo, J., 2001. Considering virtual worlds as representations of landscape realities and as tools for landscape planning. Landsc. Urban Plann. 54 (1–4), 139–148.

654

S.R.J. Sheppard / Environmental Science & Policy 8 (2005) 637–654

Orland, B., Uusitalo, J., 2001. Immersion in a virtual forest: some implications. In: Sheppard, S.R.J., Harshaw, H.W. (Eds.), Forests and Landscapes: Linking Sustainability, Ecology, and Aesthetics. International Union of Forest Research Organizations (IUFRO) Research Series.CABI International, Wallingford, UK, (Chapter 14), pp. 205–224. Pearce, D.W., Cline, W.R., Achanta, A.N., Frankenhauser, S., Pachauri, R.K., Tol, R.S., Vellinga, P., 1996. The social costs of climate change: greenhouse damage and the benefits of control. In: Bruce, J.P., Lee, H., Haites, E.F. (Eds.), Climate Change 1995: Economic and Social Dimensions of Climate Change. Cambridge University Press, Cambridge, UK. Porter, T., 1979. How Architects Visualize. Van Nostrand Reinhold, New York, NY. Ribe, R., 2005. Aesthetic perceptions of green-tree retention harvests in vista views: The interaction of cut level, retention pattern and harvest shape. Landsc. Urban Plann. 73 (4), 277–293. Salter, J.D., 2005. Designing and testing a prototypical landscape information interface for lay-people. Unpublished Masters Thesis. Faculty of Forestry, University of British Columbia, Vancouver. Schroth, O., Lange, E., Schmid, W.A., 2005. From information to participation: applying interactive features in landscape visualizations. Paper presented at Our Shared Landscape Conference, Ascona, Switzerland, May 2–6. Sheppard, S.R.J., 1989. Visual Simulation: A User’s Guide for Architects, Engineers, and Planners. Van Nostrand Reinhold, New York. Sheppard, S.R.J., 2001. Guidance for crystal ball gazers: developing a code of ethics for landscape visualization. Landsc. Urban Plann. 54 (1–4), 183–199. Sheppard, S.R.J., 2004. Making Unsustainability Obvious to the Public. Green College Environmental Seminar Presentation, Oxford, 4 February. Sheppard, S.R.J., 2005. Validity, reliability, and ethics in visualization. In: Bishop, I., Lange, E. (Eds.), Visualization in Landscape and Environmental Planning. Taylor and Francis, London, (Chapter 5), pp. 79–97. Sheppard, S.R.J., Cavens, D., Salter, J., Meitner, M., 2001. Seeing the big picture: the benefits of digital landscape immersion. In: Proceedings of the LANDTECH ASLA/CSLA Annual Meeting, 24 September 2001, Montreal, Canada. (CD-ROM). Sheppard, S.R.J., Meitner, M.J., 2005. Using multi-criteria analysis and visualisation for sustainable forest management planning with stakeholder groups. Forest Ecol. Manage. 207 (1–2), 171–187. Sheppard, S.R.J., Salter, J., 2004. The role of visualization in forest planning. In: Encyclopedia of Forest Sciences, Academic Press/Elsevier, Oxford, UK, pp. 486–498. Slovic, P., Finucane, M., Peters, E., MacGregor, D.G., 2002. The affect heuristic. In: Gilovich, T., Griffin, D., Kahneman, D. (Eds.), Heuristics and Biases: The Psychology of Intuitive Judgement. Cambridge University Press, New York, NY, (Chapter 23), pp. 397–420.

Steinitz, C., Rojo, H.M.A., Bassett, S., Flaxman, M., Goode, T., Maddock III, T., Mouat, D., Peiser, R., Shearer, A., 2003. Alternative Futures for Changing Landscapes: The Upper San Pedro River Basin in Arizona and Sonora. Island Press, Washington, DC. Stern, P.C., 2000. Toward a coherent theory of environmentally significant behaviour. J. Soc. Issues 56 (3), 407–424. Tickell, C., 2002. Communicating climate change. Science 297, 737 (Editorial). Tress, B., Tress, G., 2002. Scenario visualisation for participatory landscape planning: a study from Denmark. Landsc. Urban Plann. 982, 1–18. Tufte, E.R., 1990. Envisioning Information. Graphics Press, Cheshire, CT. UKCIP (Climate Impacts Programme), Mackenzie Hedger, M., Gawith, M., Brown, I., Connell, R., Downing, T.E. (Eds.), 2000a. Climate Change: Assessing the Impacts—Identifying Responses: The First 3 Years of the UK Climate Impacts Programme. UKCIP, Oxford, UK, 123 pp. UKCIP (Climate Impacts Programme), 2000b. Socio-Economic Scenarios for Climate Change Impact Assessment: A Guide to their Use in the UK Climate Impacts Programme. UKCIP, Oxford, UK, 123 pp. UKCIP (Climate Impacts Programme), 2002. Climate Change Scenarios for the United Kingdom: The UKCIP02 Scientific Report. UKCIP, Oxford, UK, 112 pp. Webster, M., 2003. Communicating climate change uncertainty to policymakers and the public: and editorial comment. Climatic Change 61, 1–8. Winn, W., 1997. The Impact of Three-Dimensional Immersive Virtual Environments on Modern Pedagogy. HITL Technical Report R-9715. Human Interface Technology Laboratory, University of Washington, Seattle, WA. Zajonc, R.B., 1984. On the primacy of affect. Am. Psychol. 39 (2), 117–123. Zube, E.H., Sell, J.L., Taylor, J.G., 1982. Landscape perception: research, application, and theory. Landsc. Plann. 9, 1–33. Stephen Sheppard teaches in sustainable landscape planning, aesthetics, and visualisation in the Faculty of Forestry and Landscape Architecture programme at the University of British Columbia. He received a BA/MA in agricultural and forest sciences at Oxford, an MSc in forestry at UBC, and a PhD in environmental planning at UC, Berkeley. He directs the Collaborative for Advanced Landscape Planning (CALP), an interdisciplinary research group using perception-testing and immersive/interactive visualisation to support public awareness and collaborative planning on sustainability issues. He has over 25 years’ experience in environmental assessment and public participation internationally. He has written or cowritten two books on visual simulation, and co-edited ‘‘Forests and Landscapes: Linking Ecology, Sustainability, and Aesthetics,’’ vol. 6, in the IUFRO Research Series. He is currently contributing to the BC chapter of Canada’s National Assessment of climate change impacts and adaptation. Current research interests lie in perceptions of climate change, the aesthetics of sustainability, and visualisation theory and ethics.