5 The Application of a Simple Spatial Multi- Criteria ... - CiteSeerX

8 downloads 82622 Views 2MB Size Report
analysis to natural resource management decision making using a flexible, easy-to-use ... Multi-criteria analysis applications and software development: a se-.
5 The Application of a Simple Spatial MultiCriteria Analysis Shell to Natural Resource Management Decision Making

Robert G Lesslie1, Michael J Hill2, Patricia Hill3, Hamish P Cresswell4 and Steve Dawson1 1

Australian Government Bureau of Rural Sciences, ACT, Australia Department of Earth System Science and Policy, University of North Dakota, USA 3 CSIRO Sustainable Ecosystems, ACT, Australia 4 CSIRO Land and Water, ACT, Australia 2

Abstract: Natural resource management decision making generally requires the analysis of a variety of environmental, social and economic information, incorporating value judgement and policy and management goals. Justifiable decisions depend on the logical and transparent combination and analysis of information. This chapter describes the application of spatial multi-criteria analysis to natural resource assessment and priority setting at regional and national scales using a newly developed spatial multi-criteria analysis tool — the Multi-Criteria Analysis Shell for Spatial Decision Support (MCAS-S). MCAS-S is designed for use in participatory processes and workshop situations where a clear understanding of different approaches to spatial data management and information arrangement is necessary. The MCAS-S work environment provides for multiple map display, combination and manipulation, live update of changes, and development of spider/radar plots important in ecosystem service assessments. These and other capabilities promote clear visualisation of the relationships among the decision, the science, other constraints and the spatial data. The regional scale example illustrates the analysis of biodiversity and salinity mitigation trade-offs in revegetation in a participatory process. The national scale application illustrates reporting to policy clients on the tensions between resources use and conservation in Australian rangelands — essentially an expert analysis.

74

RG Lesslie, MJ Hill, P Hill, HP Cresswell and S Dawson

5.1 Introduction Government policy makers, local authorities and land managers with responsibility for natural resource management decision making are often required to analyse large amounts of environmental, social and economic information. The transparent and logical treatment of this information including the incorporation of community opinion, public policy and management goals can be achieved using a spatial multi-criteria analysis (MCA) approach (Kiker et al. 2005; Hill et al. 2005a; Malczewski 2006). Spatial MCA capability is available in a range of commercial geographic information systems (GIS) and this may be customised for particular purposes in spatially explicit decision support tools. However, simple, flexible spatial multi-criteria shells with easy adaptability to any problem are not readily available, particularly without a programming requirement. For participatory processes and workshop situations, a spatial multi-criteria analysis shell requires a flexible interface and transparency among spatially explicit assessments, the input data, and the classification and combination rules used to generate them. This chapter describes the application of spatially explicit multi-criteria analysis to natural resource management decision making using a flexible, easy-to-use, spatial multi-criteria analysis shell — the Multi-Criteria Analysis Shell for Spatial Decision Support (MCAS-S). The utility of the software, which provides visual cognitive links between the MCA participatory process and spatially explicit data, is illustrated using two case studies. The first is a regional level land use planning process that maps priority locations for revegetation in the West Hume region of south-eastern New South Wales (NSW). The second is a national level assessment of factors affecting the sustainability of extensive livestock grazing in the Australian rangelands.

5.2 Multi-criteria Analysis MCA is a technique that allows for the measurement and aggregation of the performance of alternatives or options, involving a variety of both qualitative and quantitative dimensions. As a means of considering the links among biophysical, economic and social data with human imperatives, it is therefore particularly useful for approaching complex interactions and effects in the context of land use and land management. There are many variants of the general MCA method that can be applied in a wide variety of contexts. Many approaches are based on the pair-

The Application of a Simple Spatial Multi-Criteria Analysis Shell

75

wise comparison method of the Analytical Hierarchy Process (Saaty 2000; Ramanathan 2001). Well-developed MCA methods usually share a number of characteristics. Generally, they are flexible, enable the capture of quantitative and qualitative data and issues, are relatively simple for clients and stakeholders to use, permit the development of many alternative scenarios, allow the exploration of trade-offs, and enable stakeholders to factor results into decision making. The MCA process is a tool to assist decision makers in reaching outcomes — it does not do the decision making, or produce a solution. Attention must be given to how information quality and uncertainty is factored in and integrated with stakeholder viewpoints and biases, political and structural realities, and achievability versus optimality. It is important that each stage of the MCA process is carried out rigorously, in parallel with stakeholder engagement. Matching the spatial and temporal scale of the input information and analysis to the issues and processes under consideration is also critical. 5.2.1 Spatial Applications There is a long history of the use of MCA in operations research, but these applications are essentially non spatial although there are recent crossover developments with traditional methods such as DEFINITE (DEcisions on a FINITE set of alternatives; Janssen and Van Herwijin 2006). The development of spatial applications of MCA has accelerated as GIS software has improved, and as computer operating systems have become more suited to implementation of GUI (Graphical User Interface) approaches. There are three major groupings: • • •

GIS-based applications utilising either established MCA modules (e.g. IDRISI) or involving the use of GIS modelling and objectbased programming linkages (e.g. ArcInfo and ArcGIS) hybrid approaches that combine GIS, MCA, models and other capability via programming stand-alone software specifically designed for an application.

Generic MCA capability within existing GIS products is tied to specific software, although in some cases standalone executables can be spawned. Table 5.1 provides a selected bibliography of the development of spatial MCA applications. Increased publication of these approaches in recent years indicates a heightened interest in MCA in combining methods and in applications that aid decision making in the coupled human–natural system.

76

RG Lesslie, MJ Hill, P Hill, HP Cresswell and S Dawson

Table 5.1. Multi-criteria analysis applications and software development: a selected list of GIS-based and standalone software-based applications Software/analysis 1. GIS-based applications IDRISI (®Clark University) GISbased MCA

Application

Model-GIS (ArcInfo) coupling with MCA ILWIS GIS

Non point source farm pollution Nature conservation Geneletti (2007) value of agricultural land Urban transport policies Arampatzis et al. (2004) Watershed manageHorst and Gimona ment; biodiversity (2005) conservation

Earthquake hazards; crop suitability; soil erosion in Ethiopia ASSESS (A System for SElecting Suit- Radioactive waste repoable Sites) written in ArcInfo AML sitory; soil conditions; catchment condition (®ESRI) Planning tool; urban ArcView (®ESRI) GIS-based MCA land use

MapInfo (®) GIS-based DSS Other GIS-based MCA

2. Hybrid applications PROMETHEE integrated with ArcGIS Land parcels ranked for housing suitability SIMLAND – cellular automata, MCA Land use change and GIS written in C and using ArcInfo GIS HERO (Heuristic multi-objective Forest planning; habitat optimisation) combined with GIS, suitability AHP and Bayesian analysis 3. Stand-alone software LMAS – Land Management Advice System MULINO-DSS (MULti-sectoral, INtegrated and Operational DSS) combines simulation models, mapping and MCA GIWIN (Geographic Information Workshop for WINdows) LRA (Land Resource Allocation) MEACROSS – Multi-criteria Analysis of Alternative Cropping Systems IWM – decision support system for Management of Industrial Wastes GSA (Global Sensitivity Analysis) in SimLab (Software for Uncertainty and Sensitivity Analysis)

Reference Ceballos-Silva and Lopez-Blanco (2003); Dragan et al. (2003) Veitch and Bowyer (1996); Bui, (1999); Walker et al. (2002) Pettit and Pullar (1999); Dai et al. (2001) Morari et al. (2004)

Brans et al. (1984); Marinoni (2005) Wu (1998). Kangas et al. (2000); Store and Kangas (2001); Store and Jokimaki (2003)

Spatial expert system

Cuddy et al. (1990)

Water resources

Giupponi et al. (2004)

Land resource Ren (1997) allocation for rice paddy fields Cropping systems Mazzetto and Bonera (2003) Industrial waste Manniezzo et al. (1998) Hazardous waste Gomez-Delgado and disposal Tarantola (2006)

The Application of a Simple Spatial Multi-Criteria Analysis Shell

77

5.2.2 The Decision-making Process Policy and program development and decision making about natural resources are often driven by questions that are deceptively simple. Questions such as: ‘Where in the landscape is agricultural production approaching the limits of sustainable resource use?’, or ‘Where can we most effectively invest in revegetation?’ raise complex issues of equity, economic performance, and biophysical impact. Revegetation, for example, may have benefits for biodiversity, water quality and amenity, and costs associated with reduced water supply and agricultural production. Usually there is no ‘right’ answer — complex trade-offs may be involved, with cost and benefit considerations resting on value judgment or opinion. The generalised spatial multi-criteria analysis and the priority-setting process are illustrated in Fig. 5.1, and its stages are described below.

Opinions and views Time series metrics Spatial measures Model outputs Measurements

Others… Productivity Sensitivity Salinity Biodiversity

Composite measures Transformation and combination

Maps for each decision criterion

indicators & weights

Analysis - coincidences - tensions

Primary data and information Synthesis Decision criteria Criteria influencing the decision (e.g. strategic plan)

- options - scenarios - trade-offs

Evaluation and decision making Decision

Problem definition e.g. issues, where & how much to invest?

e.g. funding priorities & policy interventions

Fig. 5.1. The spatial multi-criteria analysis and the priority-setting process

Problem Definition

The more specific the problem, the easier it will be to detail decision criteria. For example, an investment decision might be restricted to a particular set of outcomes such as water quality or improved farming practices, or to a particular set of mechanisms such as market-based instruments.

78

RG Lesslie, MJ Hill, P Hill, HP Cresswell and S Dawson

Decision Criteria

The criteria that are to be taken into account in making investment decisions. These criteria will depend on the nature and scope of the problem. They will usually be derived from a number of sources including legislation, agreements, policy initiatives and programs. Primary Data and Information

Primary data and information is the store of available primary information upon which decision making ultimately depends. In a natural resource decision-making context this includes land, soil, water, biodiversity, social and economic information. It takes the form of measurements, model outputs, time series information, and expert knowledge. Transformation and Combination

This is a participatory process that involves using available primary data and information (above) to obtain an assessment against each decision criterion. Outputs are usually represented as maps of composite measures (below), or indexes. For example, the creation of a map of ‘species conservation value’ may involve the combination of species distribution data, with measures of diversity, endemism, rarity and endangerment along with landscape information. The process is usually iterative. Composite Measures

Composite measures, or indices, which provide an assessment against each decision criterion. These could include measures such as species conservation value, biodiversity threat, agricultural potential, agricultural value, water availability and salinity risk. More than one composite measure may be needed to represent a decision criterion. Analysis

Analysis involves the exploration of coincidences and tensions among composite measures. Synthesis

Synthesis is the outcome of the analysis process. Consensus positions may not be reached. The synthesis process usually involves the development of options and scenarios that represent particular stakeholder perspectives,

The Application of a Simple Spatial Multi-Criteria Analysis Shell

79

and alternative pathways forward. These options and scenarios are the choices for decision makers. Decision Making

Decision making involves the resolution of inconsistencies, contradictions and competing claims.

5.3 The MCAS-S Approach As MCA is useful for understanding coupled human–environment systems, it is a process that should be routinely available to natural resource policy makers and practitioners. With this in mind, the design goal for MCAS-S was to create a portable and easy-to-use exploratory multi-criteria analysis shell with generic functionality, independence from data and project type, and freedom from GIS-based barriers to flexibility and interactive use (Hill et al. 2005b). MCAS-S functionality promotes participatory processes that require the understanding of relationships between decisionmaking requirements and the available data. Of particular value in this regard are software features that enable interactive cognitive and ‘live update’ mapping of alternative views. 5.3.1 Design Principles In designing MCAS-S, the decision was made to focus functionality on assisting the decision-making process by promoting improved visualisation and transparency of relations among spatial data, value judgments, decisions and results. The tool thus requires the use of readily available GIS for spatial data management and analysis, and does not seek to provide these functions. The first key design principle of MCAS-S is coupling the human cognitive process with the display and interrogation of spatial data (Bisdorff 1999) so that the origins of complex summary indices and queries are transparent. In MCAS-S, direction arrows connect data elements showing their contributions to combined layers. The second key principle is the automatic update of available functions according to the type of userselected data layers. These two features make the interface highly intuitive and easy to use. The third key principle is the maintenance of a live link between the primary data and all subsequent indicator, composite and analysis layers.

80

RG Lesslie, MJ Hill, P Hill, HP Cresswell and S Dawson

This allows the user to make adjustments to the treatment of primary data and immediately see the consequences as derived layers. A final major construct behind MCAS-S is enabling simple visual comparison of multiple themes in pairs (using a two-way comparison matrix) and in groups (using multi-way radar or spider plots). 5.3.2 Key Functions MCAS-S requires the user to prepare raw spatial data for a project in a raster format of consistent geographical extent, pixel resolution and projection. Primary data can then be selected from the menu and dragged into a display workspace whereupon the spatial data layer can be classified using a variety of simple classification methods and tools for tailoring input data. With the creation of individual class or rule layers, the user can apply them in weighted combinations to construct composite layers contributing to themes of interest. Key functions for combining and comparing spatial layers are illustrated below in two natural resources management applications, and represented schematically in Fig. 5.2. A composite layer interface will appear into the workspace when selected from the menu button. The weighted contribution of any selected layer to any composite can be set by using this interface. The composite map dynamically updates as the user changes the weightings on input layers. The user can see in the workspace a hierarchical, cognitive ‘map’ showing the development of individual layers culminating in a final summary layer representing a theme. Relationships among themes, specific views and particular indicators may be examined using several methods. Two-way comparison enables the user to create a two-way comparison map, explore the association among input classes, and define a colour ramp and value scale to highlight association of high or low values, or feature a particular geographical region. Multi-way comparison is used when the spatial association of two or more data layers is required. The multi-way analysis uses the radar plot as the basis for visualisation. The user can create radar plots with spokes representing layers or themes. The multi-way function provides for the creation of a binary layer that spatially delimits the area either inside or outside the limits of the radar plot, that is, the area that meets or does not meet certain user-defined criteria levels. The multi-way comparison may also be displayed as a grey-scale surface showing ‘distance’ from selected criteria values. A range of ancillary facilities includes a viewer that presents — for a selected grid cell — dynamic feedback of values for all related layers, in-

The Application of a Simple Spatial Multi-Criteria Analysis Shell

81

cluding two-way and multi-way comparison. Masking and overlay functions, a reporting utility, image and process logging, and data export are also available.

Fig. 5.2. Flow chart for MCAS-S showing stages and functionality

82

RG Lesslie, MJ Hill, P Hill, HP Cresswell and S Dawson

5.4 Applications The remainder of this chapter demonstrates the use of MCAS-S functionality in two natural resources management applications: • land use planning process that involved mapping priorities for revegetation in the West Hume region of southern NSW • national assessment of factors affecting the sustainability of extensive livestock grazing in the Australian rangelands. 5.4.1 Prioritising Revegetation Investment in West Hume NSW A practical catchment investment planning process needs to focus on indicative zones for investment in landscape change, consistent with maximising multiple environmental outcomes. In a regional context an effective prioritisation process provides for the best possible use of existing datasets and the technical expertise of participants, integrating knowledge and balancing landscape options in a transparent and objective way, and at a level that allows prioritisation of on-ground works and incremental improvement over time, bringing together new information to help decision making. In 2006, CSIRO Land and Water and the Murray Catchment Management Authority (CMA) collaborated in identifying priority areas for revegetation — to decide ‘what needs to be planted where’ — in the West Hume area of southern NSW (Hill et al. 2006a). The West Hume area is approximately 86,000 ha of predominantly mixed cropping agricultural land, in the southern part of NSW within a region managed by the Murray CMA (Fig. 5.3). This area was previously identified by the CMA as being within a salinity and biodiversity management priority area (Murray Catchment Management Board 2001). The Murray CMA co-invests with land owners in local-scale environmental activities such as revegetation with native plant species and establishment of perennial pasture. Targeting such investment in locations where environmental outcomes are likely to be maximised, and having a transparent and defensible prioritisation mechanism, is a primary CMA objective.

The Application of a Simple Spatial Multi-Criteria Analysis Shell

83

Fig. 5.3. Location of the West Hume area of southern New South Wales

Priority mapping was developed using MCAS-S from a series of rules or guidelines and spatial analysis procedures for the definition and creation of suitability ranked input data layers for biodiversity and salinity. Spatial data used for the creation of these input data layers included: • • • • • • • • • • •

historical rainfall and temperature elevation (and derivatives, e.g. slope) soil landscapes and profile classes soil water holding capacity current (or recent) land use pre-settlement vegetation threatened species (point) data groundwater pressure and quality data groundwater flow systems mapping drainage networks, stream flow and water quality location of important assets or infrastructure such as towns and roads.

The relative importance of rules or guidelines was assessed using a pair-wise comparison procedure by a CSIRO and Murray CMA focus group. The pair-wise comparison used the method of the Analytical Hier-

84

RG Lesslie, MJ Hill, P Hill, HP Cresswell and S Dawson

archy Process (AHP) (Saaty 2000; Ramanathan 2001). In this method, each guideline rule was compared with every other guideline rule and ranked as more or less important by each member of the focus group. The results were combined in a cross-tabulation and converted to a series of relative importance weightings and rankings for the salinity and biodiversity guidelines (Table 5.2). The focus group also assessed the relative importance of the biodiversity and salinity themes. Table 5.2. Weighting and ranking for biodiversity and salinity guidelines Guidelines Biodiversity guidelines (rules) 1 Revegetate for geographical dispersal 2 Revegetate biophysically heterogeneous areas 3 Revegetate rare broad vegetation types 4 Revegetate areas with rare species 5 Revegetate in areas with dense patch distribution 6 Revegetate close to large patches 7 Revegetate close to streams 8 Revegetate enclosed areas 9 Revegetate to form corridors 10 Revegetate land with low production potential Total Salinity guidelines (rules) 11 Revegetate areas: responsive groundwater flow systems 12 Protect high value water resources 13 Protect high value biodiversity assets 14 Protect high value built assets 15 Revegetate soils with high salt stores 16 Revegetate high recharge potential areas 17 Revegetate in areas with high rainfall 18 Revegetate away from saline discharge zones 19 Revegetate low value agricultural land 20 Revegetate areas with high forest production potential Total

Weighting (rank: 1=most important) 0.0974 (6) 0.0541 (9) 0.1038 (5) 0.0935 (7) 0.1502 (1) 0.1302 (4) 0.1492 (2) 0.07312 (8) 0.1305 (3) 0.0179 (10) 1.000 0.1678 0.0939 0.1039 0.0730 0.1262 0.1551 0.0683 0.0641 0.1180 0.0299 1.000

(1) (6) (5) (7) (3) (2) (8) (9) (4) (10)

Each of the 20 spatial suitability surfaces for each of the guidelines was imported into an MCAS-S project. Three new composite layers were added to the project: Salinity, Biodiversity and Combined. The 10 biodiversity guidelines were used to create the Biodiversity composite, using the weightings developed in the AHP pair-wise comparison. The same process was repeated for the Salinity composite using the 10 salinity guidelines and their associated weightings. The Salinity and Biodiversity composites were then combined to create a Combined composite, with the biodiversity layer

The Application of a Simple Spatial Multi-Criteria Analysis Shell

85

receiving a 0.4 weighting and the salinity layer receiving 0.6 weighting (Fig. 5.4). The spatial multiple-criteria analysis with MCAS-S used within a participatory process defined locations appropriate for revegetation based on the number of desirable criteria (guidelines) that were met, and the relative importance of these criteria (weighting). The analysis provided a straightforward and logical means for allocating investment in revegetation according to relative suitability in a catchment context, using simple firstprinciple criteria with available spatial data. Further decisions on sites for revegetation work are made in the context of local farm plans and farmer land management objectives.

Fig. 5.4. A MCAS-S workspace linking spatial rule layers to prioritise revegetation in West Hume

5.4.2 Assessing the Sustainability of Extensive Livestock Grazing in the Australian Rangelands The Australian rangelands occupy approximately three-quarters of the Australian continent. Land use in the rangelands is dominated by extensive sheep and cattle grazing on native pastures, with rainfall generally being too low or too variable for dryland cropping or grazing on improved pas-

86

RG Lesslie, MJ Hill, P Hill, HP Cresswell and S Dawson

tures. Understanding the implications of alternative land management options is important in the rangelands because of the large area of Australia they occupy and the reliance of rangeland communities and industries on sustainable management of natural resources. Spatial multi-criteria analysis approaches have been used to support policy analysis on aspects of sustainability in the Australia rangelands (Stafford-Smith et al. 2000; Hill et al. 2006b). From a pastoral perspective, livestock grazing in the rangelands can be characterised as sustainable where economic resilience and stability can be achieved in conjunction with the regional maintenance of native species and other ecosystem services. Informed public policy development requires an understanding of where in the landscape these ecological and economic influences are operating. Sustainability may be at risk in locations where the resource base has limited resilience to livestock grazing and pastoral land use is viable in terms of potential productivity (Fig. 5.5) (Stafford-Smith et al. 2000).

Fig. 5.5. Scope for public intervention in rangeland management (adapted from Stafford-Smith et al. 2000)

Lesslie et al. (2006b) illustrate how appropriate spatial data layers and expert opinion can be combined using the MCAS-S tool to allow spatial exploration of these relationships. First, relevant input data were combined using the MCAS-S composite function to produce composite indexes representing the attributes on each axis shown in Fig. 5.5. Primary data input layers were selected and their role in contributing to composite measures examined in a workshop process utilising expert advice. The potential pas-

The Application of a Simple Spatial Multi-Criteria Analysis Shell

87

toral productivity index was, for instance, developed by the weighted combination of input layers using the logic that potential productivity will generally be greater where: • landscapes have higher productive potential (forage potential) • rainfall is relatively consistent within and between seasons • there is better access to markets, supplies and labour. The result of composite development within MCAS-S for potential pastoral productivity is shown in Fig. 5.6.

Fig. 5.6. A representation of potential productivity for livestock grazing in the Australian rangelands

Second, the MCAS-S two-way function was used to enable exploration of the spatial relationship among derived composite layers of potential pastoral productivity and the sensitivity of the resource base to livestock grazing. The two-way layer on the right hand side of the MCAS-S desktop in Fig. 5.7 highlights locations where conditions in the upper right quadrant of the matrix shown in Fig. 5.5 apply on the basis of derived index values (i.e. where there is a coincidence of high potential productivity and high resource sensitivity index values). Locations where other relationships apply (e.g. high potential productivity and low resource sensitivity; low po-

88

RG Lesslie, MJ Hill, P Hill, HP Cresswell and S Dawson

tential productivity and high resource sensitivity; low potential productivity and low resource sensitivity) can be interactively explored in MCAS-S by simply clicking on the appropriate cell in the matrix at bottom left. In this way the two-way function facilitates the spatially explicit landscape condition assessment and policy analysis.

Fig. 5.7. A two-way analysis of potential productivity and resource sensitivity

The spatial relations among more than two layers can be interactively explored using the MCAS-S multi-way comparison facility. Figure 5.8 shows the creation of a binary layer (at bottom right) that identifies locations that meet specified conditions for each of three composite layers developed for the Australian rangelands study: potential productivity, sensitivity of the resource base to livestock grazing, and total grazing pressure. The spatial association and visualisation process involves the use of the radar plot, with spokes that represent the scale of each respective input layer. Class limits on these spokes are interactively manipulated to delimit locations that satisfy user defined criteria levels. The multi-way binary layer shown in Fig. 5.8 identifies locations where there is a coincidence of relatively low potential productivity, relatively high sensitivity of the resource base to livestock grazing, and relatively high total grazing pressure. The multi-way layer may also be displayed as a grey-scale surface showing ‘distance’ from selected criteria values.

The Application of a Simple Spatial Multi-Criteria Analysis Shell

89

Fig. 5.8. A multi-way analysis of potential productivity, resource sensitivity and total grazing pressure

5.5 Future Trends As a versatile spatial multi-criteria analysis shell, MCAS-S has many practical applications, including the assessment of ecosystem services and the visualisation of complex analyses of land use management and land management practices. The MCAS-S composite and multi-way analysis functions enable the development and comparison of multiple spatial data themes (e.g. carbon, water, aesthetics, production and ecosystem function) as described by Foley et al. (2005). Similarly, MCAS-S multi-criteria visualisation procedures can be applied with land use and land management practices information (Lesslie et al. 2006a) to assist evaluation of complex land management change and trade-off options. Additional functionality will be added to MCAS-S. In the West Hume study a pair-wise comparison of guidelines was completed in order to rank and weight them for analysis. This was conducted outside the MCAS-S software with weightings applied in the creation of the biodiversity and salinity composite layers. Further development of the MCAS-S software will

90

RG Lesslie, MJ Hill, P Hill, HP Cresswell and S Dawson

add a pair-wise comparison module to MCAS-S so that the rankings, ranking matrix table, and weightings can all be created within the software and associated with each layer. A map calculator function for composite development is also under development. Description of time and space-based complexity will enhance the functionality of MCAS-S. Time series describe many important and spatially explicit processes in the landscape. Full information extraction from time series may require application of a suite of approaches including classical time series decomposition (Roderick et al. 1999), calculation of curve metrics (Reed et al. 1996), piece-wise logistic functions to capture curve trajectories (Zhang et al. 2003), time-based classifications (Moody and Johnson 2001) and Fourier and wavelet analysis (Li and Kafatos 2000). Recent research has highlighted the importance of global sensitivity analysis (Gomez-Delgado and Tarantola 2006) and multi-criteria weight sensitivity analysis (Feick and Hall 2004) to define the response space and impact of input layers on MCA outcomes. The implementation of a Bayesian Belief Network within MCAS-S is also being explored. There are recent examples of the application of Bayesian network approaches and multi-objective model land system and environmental assessments (Dorner et al. 2007; Pollino et al. 2007). Consideration could be given to inclusion of soft systems approaches that enable users to create partial weight tables to capture soft attributes such as belief, commonality and plausibility (Beynon 2005).

5.6 Conclusion Natural resource management decision making by communities, rural industries and governments is usually undertaken in a context of uncertainty where, ideally, the decision-making process engenders trust and engagement among stakeholders. Informed debate about competing demands on natural resources and the desirability of alternative futures can benefit from the use of versatile spatially explicit decision support tools such as MCAS-S. The particular advantages of the MCAS-S tool are its flexibility, simplicity, the capability to visualise spatial data and the linkages between spatial data, and its suitability for use in group and participatory processes. The West Hume land use planning application combines simple first principle criteria with readily available spatial data to prioritise locations appropriate for revegetation based on the number of desirable criteria that were met, and the relative importance of these criteria. The process dem-

The Application of a Simple Spatial Multi-Criteria Analysis Shell

91

onstrates how a regional investment strategy can be translated into mapped priorities by focussing on identifying where multiple environmental and production outcomes can be achieved at minimal cost. The utility of the tool in a broader context of policy analysis and development is demonstrated by the Australian rangelands application. The wider adoption of this type of planning process in Australia could substantially improve the effectiveness of current investment in natural resource management.

5.7 Future Research Directions There is a growing expectation that spatial science and information can be successfully applied to decision making on complex issues affecting coupled human–environment systems. This expectation is promoted by the increasing sophistication of spatial multi-criteria analysis tools and more readily accessible economic, biophysical and social spatial data. While tools such as MCAS-S can help integrate factual information with value judgements and policy goals in a transparent and flexible way, there is still considerable scope for improvement in spatial data handling technologies and analytical methods. A particular need is for technologies that are capable of delivering information at many levels of scale and complexity with a cognitive context, through visualisation, that aligns more naturally with human thinking and decision making. Key research challenges include: • the analysis of time series data, particularly the discrimination of patterns in trend, periodicity and magnitude • the linkage of time series and signal processing methods to spatial analysis • resolving scale, particularly in relation to economic and social data and issues arising from changes in scale • managing uncertainty, especially in the linkage of scientific and social domains. Improvement in visualisation is important so that scenario outputs from spatially explicit modelling and decision processes can be made more meaningful to non-experts.

92

RG Lesslie, MJ Hill, P Hill, HP Cresswell and S Dawson

Acknowledgements The MCAS-S system was developed for the Natural Resource Management Division of the Australian Government Department of Agriculture, Fisheries and Forestry with funding from the Natural Heritage Trust. We acknowledge funding support for the West Hume study from the Murray Catchment Management Authority and the CSIRO Water for a Healthy Country Flagship program. Jean Chesson is thanked for her advice on the multi-criteria analysis and priority-setting process. The chapter also benefited from the comments of two anonymous referees.

References Arampatzis G, Kiranoudis CT, Scaloubacas P, Assimacopoulos D (2004) A GISbased decision support system for planning urban transportation policies. European Journal of Operational Research 15:465–475 Beynon MJ (2005). A method of aggregation in DS/AHP for group decisionmaking with the non-equivalent importance of individuals in the group. Computers and Operations Research 32:1881–1896 Bisdorff R (1999) Cognitive methods for multi-criteria expert decision making. European Journal of Operational Research 119:379–387 Brans JP, Mareschal B, Vincke PH (1984) PROMETHEE: A family of outranking methods in multicriteria analysis. Operational Research 22:477–490 Bui E (ed) (1999) A soil information strategy for the Murray-Darling Basin (MDBSIS). Report to Murray Darling Basin Commission, Project D5038, available at http://www.brs.gov.au/mdbsis/publications.html Ceballos-Silva A, Lopez-Blanco J (2003) Evaluating biophysical variables to identify suitable areas for oat in Central Mexico: a multi-criteria and GIS approach. Agriculture, Ecosystems and Environment 95:371–377 Cuddy SM, Laut P, Davis JR, Whigham PA, Goodspeed J, Duell T (1990) Modelling the environmental effects of training on a major Australian army base. Mathematics and Computers in Simulation 32:83–88 Dai FC, Lee CF, Zhang XH (2001) GIS-based geo-environmental evaluation for urban land-use planning: a case study. Engineering Geology 61:257–271 Dorner S, Shi J, Swayne D (2007) Multi-objective modelling and decision support using a Bayesian network approximation to a non-point source pollution model. Environmental Modelling and Software 22:211–222 Dragan M, Feoli E, Fernetti M, Zerihun W (2003) Application of a spatial decision support system (SDSS) to reduce soil erosion in northern Ethiopia. Environmental Modelling and Software 18:861–868 Feick RD, Hall GB (2004) A method for examining the spatial dimension of multi-criteria weight sensitivity. International Journal of Geographical Information Science 18:815–840

The Application of a Simple Spatial Multi-Criteria Analysis Shell

93

Foley JA, DeFries R, Asner GP, Barford C, Bonan G, Carpenter SR, Chapin FS, Coe MT, Daily GC, Gibbs HK, Helkowski JH, Holloway T, Howard EA, Kucharik CJ, Monfreda C, Patz JA, Prentice IC, Ramankutty N, Snyder PK (2005) Global consequences of land use. Science 309:570–574 Geneletti D (2007) An approach based on spatial multi-criteria analysis to map the nature conservation value of agricultural land. Journal of Environmental Management 83:228–235 Giupponi C, Mysiak J, Fassio A, Cogan V (2004) MULINO-DSS: a computer tool for sustainable use of water resources at the catchment scale. Mathematics and Computers in Simulation 64:13–24 Gomez-Delgado M, Tarantola S (2006) GLOBAL sensitivity analysis, GIS and multi-criteria evaluation for a sustainable planning of hazardous waste disposal site in Spain. International Journal of Geographical Information Science 20:449–466 Hill MJ, Braaten R, Lees B, Veitch SM, Sharma S (2005a) Multi-criteria decision analysis in spatial decision support: the ASSESS analytic hierarchy process and the role of quantitative methods and spatially explicit analysis. Environmental Modelling and Software 20:955–976 Hill MJ, Lesslie R, Barry A, Barry S (2005b) A simple, portable, spatial multicriteria analysis shell – MCAS-S. In: Zerger A, Argent RM (eds) MODSIM 2005 International Congress on Modelling and Simulation. Modelling and Simulation Society of Australia and New Zealand, 12–15 December 2005, pp 1532–1538, available at http://www.mssanz.org.au/modsim05/papers/hill.pdf Hill P, Cresswell H, Hubbard L (2006a) Spatial prioritisation of NRM investment in the West Hume area (Murray CMA region). Technical Report, CSIRO, Water for a Healthy Country National Research Flagship: Canberra, pp 1–28 Hill MJ, Lesslie RG, Donohue R, Houlder P, Holloway J, Smith J (2006b) Multicriteria assessment of tensions in resource use at continental scale: A proof of concept with Australian rangelands. Environmental Management 37:712–731 Horst D, Gimona A (2005) Where new farm woodlands support biodiversity action plans: a spatial multi-criteria analysis. Biological Conservation 123:421– 432 Janssen R, Van Herwijnen M (2006) A toolbox for multi-criteria decision-making. International Journal of Environmental Technology and Management 6:20–39 Kangas J, Store R, Leskinen P, Mehtätalo L (2000) Improving the quality of landscape ecological forest planning by utilising advanced decision-support tools. Forest Ecology and Management 132:157–171 Kiker GA, Bridges TS, Varghese A, Seager TP, Linkov I (2005) Application of multicriteria decision analysis in environmental decision making. Integrated Environmental Assessment and Management 1:95–108 Lesslie R, Barson M, Smith J (2006a) Land use information for integrated natural resources management – a coordinated national mapping program for Australia. Journal of Land Use Science 1:45–62 Lesslie R, Hill MJ, Woldendorp G, Dawson S, Smith J (2006b) Towards Sustainability for Australia’s Rangelands: Analysing the Options. Australian Government Bureau of Rural Sciences: Canberra, pp 1–12

94

RG Lesslie, MJ Hill, P Hill, HP Cresswell and S Dawson

Li Z, Kafatos M (2000) Interannual variability of vegetation in the United States and its relation to El Nino/Southern Oscillation. Remote Sensing of Environment 71:239–247 Malczewski J (2006) GIS-based multi-criteria decision analysis: a survey of the literature. International Journal of Geographical Information Science 20:703– 726 Maniezzo V, Mendes I, Paruccini M (1998) Decision support for siting problems. Decision Support Systems 23:273–284 Marinoni O (2005) A stochastic spatial decision support system based on PROMETHEE. International Journal of Geographical Information Science 19:51–68 Mazzetto F, Bonera R (2003) MEACROS: a tool for multi-criteria evaluation of alternative cropping systems. European Journal of Agronomy 18:379–387 Moody A, Johnson DM (2001) Land-surface phenologies from AVHRR using the discrete Fourier transform. Remote Sensing of Environment 75:305–323 Morari F, Lugato E, Borin M (2004) An integrated non-point source model-GIS system for selecting criteria of best management practices in the Po Valley, North Italy. Agriculture, Ecosystems and Environment 102:247–262 Murray Catchment Management Board (2001) Murray catchment management plan; a blueprint for action. Dated 17 October 2001 (draft) Pettit C, Pullar D (1999) An integrated planning tool based upon multiple criteria evaluation of spatial information. Computers, Environment and Urban Systems 23:339–357 Pollino CA, Woodberry O, Nicholson A, Korb K, Hart BT (2007) Parameterisation and evaluation of a Bayesian network for use in an ecological risk assessment. Environmental Modelling and Software 22:1140–1152 Ramanathan R (2001) A note on the use of the analytic hierarchy process for environmental impact assessment. Journal of Environmental Management 63:27– 35 Reed BC, Loveland TR, Tieszen LL (1996) An approach for using AVHRR data to monitor US Great Plains grasslands. Geocarto International 11:1–10 Ren F (1997) A training model for GIS application in land resource allocation. ISPRS Journal of Photogrammetry & Remote Sensing 52:261–265 Roderick ML, Noble IR, Cridland SW (1999) Estimating woody and herbaceous vegetation cover from time series satellite observations. Global Ecology and Biogeography 8:501–508 Saaty TL (2000) Fundamentals of Decision Making and Priority Theory with AHP Analytic Heirarchy Process. RWS Publications, Pittsburg Stafford-Smith DM, Morton SR, Ash JA (2000) Towards sustainable pastoralism in Australia’s Rangelands. Australian Journal of Environmental Management 7:190–203 Store R, Jokimaki J (2003) A GIS-based multi-scale approach to habitat suitability modelling. Ecological Modelling 169:1–15 Store R, Kangas J (2001) Integrating spatial multi-criteria evaluation and expert knowledge for GIS-based habitat suitability modelling. Landscape and Urban Planning 55:79–93

The Application of a Simple Spatial Multi-Criteria Analysis Shell

95

Veitch SM, Bowyer JK (1996) ASSESS: A System for Selecting Suitable Sites. In: Morain S, Lopez Baros S (eds) Raster Imagery in Geographic Information Systems. OnWord Press, Santa Fe, p 495 Walker J, Veitch S, Dowling T, Braaten R, Guppy L, Herron N (2002) Assessment of catchment condition. CSIRO Land and Water, Canberra Wu F (1998) SimLand: a prototype to simulate land conversion through the integrated GIS and CA with AHP-derived transition rules. International Journal of Geographical Information Science 12:63–82 Zhang X, Friedl MA, Schaaf CB, Strahler AH, Hodges JCF, Gao F, Reed BC, Huete A (2003) Monitoring vegetation phenology using MODIS. Remote Sensing of Environment 84:471–475

Suggest Documents