SUSTAINABLE LAND RESOURCE ASSESSMENT IN REGIONAL AND URBAN SYSTEMS

These models have been used to help develop strategic development plans for several rural shires in Victoria, Australia. One of the major objectives of these Shire development strategies was to protect good agricultural land from urban development. There are more than 30 agricultural land suitability models that have been developed to address particular commodities of interest to these Shires. Shires where these models have been applied so far include South Gippsland.


INTRODUCTION
Strategic planning of regional and urban systems, or strategic spatial planning, has regained importance in the past decade as a fundamental approach to achieving sustainable development in the environmental and socio-economic fields (Healey et al. 1997;Layard et al. 2001).Land Resource Assessment (LRA) has become a cornerstone of the methodological approach ( CSIRO Land and Water 2000).The main purpose of LRA is the identification of environmental (i.e.biological and physical) conditions that present an opportunity, or pose a constraint, to proposed developmental actions (FAO 1976(FAO ,1993;;Davison 1992;Conacher and Conacher 2000).
Soil, landform, vegetation and climate, or combinations of these ecosystem components, have been used as indicators of a variety of biophysical characteristics in the different land evaluation, or assessment, systems developed and applied around the world (Hopkins 1977;Westman 1985).These biophysical characteristics can be used to predict the suitability of land for different uses (or activities), or as indicators of the vulnerability of the land.
Typically, each landscape attribute is separately mapped, and relevant maps overlaid to determine homogeneous land units that contain the landscape attributes of interest for particular land uses (Ahern 1999).Since the early 1960s, such approaches have rapidly evolved from handdrawn transparency maps suitable for overlay (McHarg 1969;Fabos and Ahern 1996) to sophisticated computerised systems using Geographic Information Systems (GIS).The latter are sometimes attached to automatic systems of data input from satellite imagery using remote sensing (Hossain and Morse-McNabb 2004).
The following examples serve to illustrate some of the important applications of LRA.An urban planner, architect or builder, is looking for sites to develop a large housing estate; a transport and road construction agency is studying possible alignments for a new road; a catchment management authority, or a water board, is examining best uses for available water to encourage agriculture diversification.Each has not only specific design; engineering and economic requirements that must be met, but also certain biophysical constrains that must be taken into account.Houses must not subside or be located on very steep slopes; highways will not be publicly supported if they endanger threatened species or their habitats; agricultural production should have minimum impact on the environment.
Governments and public agencies recognise the importance of the functions performed by natural systems, their fragility in the face of development, and the irreversibility of damage that may occur.Although it is not usually required by legislation, LRA provides a scientific basis for the selection of the most appropriate sustainable use for a particular area.Encompassed in this is the possibility of converting constraints into opportunities.For instance, an area of good organic soils, which is subject to flooding, is a poor site for a housing estate, though suitable and attractive for the necessary open spaces associated with the estate.Similarly, steep slopes often limit the location of buildings or certain agricultural uses, but are good places for the preservation of vegetation cover to prevent severe soil erosion and sedimentation of water features.
So far we have over 30 such commodities and commodities groups' models have been developed to satisfy the requirements of Local Councils and regional organisations in Victoria.These cover fruits, vegetables (horticulture), agro-forestry, pastures, crops and floriculture.Table 1 shows a summary list of these commodities.This paper describes the development and application of a LRA approach.Specifically, it explains an Agricultural Land Suitability (ALS) Model and an Urban Buildability Model (UBD) and their application in South Gippsland, State of Victoria, Australia.The next section introduces the basis of the approach.

MULTI-CRITERIA EVALUATION IN A GIS ENVIRONMENT
To define the suitability of an area for a specific land use (or activity), several criteria need to be considered.For this purpose, the methodology discussed in this paper uses a Multi-Criteria Evaluation (MCE) method within a Geographic Information Systems (GIS).
MCE has been developed to investigate a number of alternatives (or choice possibilities) in the light of multiple objectives (or criteria) and conflicting preferences (or priorities) (Voogd 1983;Keeney and Raiffa 1993).It is a very useful method when a set of alternatives need to be evaluated on the basis of conflicting and incommensurate criteria (Malczewski 1999).
MCE has been utilised around the world for land suitability modelling, where the primary concern is how to combine the information from several criteria to form a single index of evaluation (Carver 1991;Jankowski and Richard 1994).A criterion is some basis for a decision that can be measured and evaluated.They can be of two types: factors and constraints, and can pertain either to attributes of the individual or to an entire decision set.A factor is a criterion than enhances or detracts from the suitability of a specific alternative for the use or activity under consideration.It is generally measured on a continuous scale.A constraint serves to limit the alternatives being considered (Eastman 1999, 3).
MCE is most commonly applied by one of two procedures.The first, deployed in traditional land suitability models, uses the Boolean aggregation method of constructing suitability indexes that present a rigid binary choice of acceptance or rejection; i.e. all criteria are reduced to statements of suitability and then combined by means of one or more operators such as intersection (logical AND) and union (logical OR) of conditions.An example is the Most Limiting Factor (MLF) technique that is based on the United States Department of Agriculture land capability classification (Klingebiel and Montgomery 1961;Rowe et al. 1981;Rowan 1990).
The second procedure is known as Weighted Linear Combination (WLC) which uses a soft or 'fuzzy' concept of suitability in standardising criteria (Voogd 1983, 120).Instead of the hard Boolean decision of assigning absolute suitability or unsuitability to a location for a given criteria, it is scaled to a particular common range where suitable and unsuitable areas are continuous measures.The WLC retains the variability of continuous criteria and allows criterion to tradeoff with each other (Eastman 1999).A low suitability defined by one criterion may be compensated by a high suitability score in another criterion.Trade-off between criteria will depend on weights assigned to them, and a wide variety of techniques exist for the development of weights (Burgman, 2005).As depicted in the decision strategy space in Figure 1, the WLC model is a significant improvement over the Boolean approach by avoiding its extreme risk aversion (binary rejection) and extreme risk taking (binary acceptance) nature.

ANALYTICAL HIERARCHY PROCESS (AHP) AND EXPERTS' JUDGEMENT
In the methodology developed in this research, the MCE is implemented using an experts systems modelling approach -the Analytical Hierarchy Process (AHP) developed by Saaty (1980Saaty ( , 1994Saaty ( , 1995)).AHP provides a framework that incorporates experts' participation in the decision-making process.Compared to empirical models based purely on the correlation amongst data, the Experts Systems Modelling incorporates the knowledge of experts who can address significant issues and have an understanding of the system of concern that may be relevant to the needed decisionmaking.The model is also better suited when access to good data is limited.
As explained by Saaty (1995, 5): (AHP) enables us to make effective decisions on complex issues by simplifying and expediting our natural decision-making process.Basically the AHP is a method of breaking down a complex unstructured situation into its component parts; arranging these parts, or variables, into a hierarchical order [or decision tree; see Alexander 1964]; assigning numerical values to subjective judgements on the relative importance of each variable; and synthesising the judgements to determine which variables have the highest priority and should be acted upon to influence the outcome of the situation.The AHP also provides an effective structure for group decision making by imposing a discipline on the group thought's processes.The necessity of assigning a numerical value to each variable of the problem helps decision makers to maintain a cohesive thought pattern and to reach a conclusion.In addition, the consensual nature of group decision making improves the consistency of judgments and enhances the reliability of the AHP as a decision-making tool.
When AHP is integrated with GIS, it can deal with criteria that are interdependent, both from the effect on land and in the interaction between spatial units (Banai-Kashani 1989)."AHP incorporates both the qualitative and the quantitative aspects of human thought: the qualitative to define the problem and its hierarchy and the quantitative to express judgements and preferences concisely.The process itself is designed to integrate these dual properties."(Saaty 1995, 6); see also Bantayan and Bishop 1988.In the procedure for MCE using a WLC, it is necessary that the weights sum to one.In the AHP, these weights are calculated by taking the principal eigenvector of a square reciprocal matrix of pair-wise comparison between the criteria.All possible pairs of criteria are compared by experts on a 9-point continuous scale for their relative importance in determining the suitability of the stated objectives.The ratings are then introduced in a matrix to calculate the weights (Eastman 1999).

AN OUTLINE OF THE MODELLING PROCESS
In our methodology, the procedure to carry out the AHP is the same for all situations, although certain steps may be given special emphasis depending on the problem of interest and repetition is often necessary.The description below is specifically related to land suitability analysis.The panel of experts is chosen to provide advice for each land use (or activity) in question; for instance, in investigating agricultural suitability it may include local growers, agronomists, soil scientists, industrialists and other various experts (Sposito et al. 2000a;2000b, 2002a, 2002b, 2002c;Hossain and Tiller 2001).
• Define the issue(s) or problem, and specify the solution desired.The issue(s) are determined by the particular needs and concerns at the regional level.
• Identify the focus.The focus forms the pinnacle of the hierarchy and is the outcome being sought from the application of AHP.
• Identify the criteria.Criteria in the form of critical factors for growth for the selected agricul- tural commodities, or groups of them are based on acknowledged bibliography and agreed upon by the experts.
• Construct the hierarchy.The hierarchy is structured in the form of a decision tree with the overall objective, or focus, at the top.The hierarchy enables to assess the impact of elements of a higher level on those of a lower level, or alternatively the contribution of elements in the lower level to the importance or fulfilment of the elements in the level above."Elements (criteria) that are of less immediate interest can be represented in general terms at the higher levels of the hierarchy and elements critical to the problem at hand can be developed to greater depth and specificity" (Saaty 1994).Where necessary, primary criteria ought to be broken down into secondary and tertiary criteria.The criteria may be reviewed and modified.
• Assign intensity ratings to the range of data values for the critical factors that have been identified.The rating is made in terms of the impact on each of the lowest level criteria (factors) for each primary criterion.By assigning intensity ratings, experts can provide an assessment of the critical factors in relation to the level at which they may become limiting to the suitability (plant growth for example) or protection of the environment.
• Weight the criteria by posing a set of questions between pairs of criterion at each level of the hierarchy to establish the relative importance or priority.The pair-wise comparisons is a robust technique for capturing preferences as the user compares all factors against each other but only two factors at a time, and thus can make a more reliable judgement.In the land suitability application, environmental factors that may contribute to, or impact upon, commodity (vegetable, tree or pasture) growth and production are weighted.
The AHP weights are calculated using the WEIGHT module in IDRISI GIS software (IDRISI32 2000).A consistency ratio is also calculated to measure consistency of the pair-wise comparison.
• Model Development is then done using the Model Builder in ArcView GIS (ESRI 2000).The suitability model is defined with the AHP hierarchy, criteria weights and data value ratings formulated previously.The model is then fed with necessary data and executed to produce the resultant suitability map.
• Validate the Suitability Maps.The resultant commodity map is obtained from processing all the map overlays by reclassifying field values to AHP ratings, multiplying each by the associated weight, and afterwards summing the maps together for each level of the hierarchy.The final map ranks areas in terms of suitability for the production of the commodity under consideration that has an index range of 0 (zero) to 10 (ten), where 0 represents a site with little or no value and a 10 represents a near perfect site.For a regional or a local application this Suitability Index can be categorised into a four class rating system, such as very low, low, moderate and high suitability.The same panel of experts assesses the result and validates the final suitability map.If necessary, weighting's and intensity ratings of the 'criteria for growth' can be adjusted.

MATCHING USES/ACTIVITIES TO THE CHARACTERISTICS OF THE LAND RESOURCE
There is increasing evidence that current agricultural production systems in Australia are not sustainable in the long-term.They can cause waterlogging, rising watertables and salinisation, and leaching of nutrients, particularly nitrate, which is reflected in accelerated rates of soil acidification.Nutrient leakage, when combined with the overuse of fertilizers, subsequently leads to water eutrophication and algae blooms (National Land & Water Resources Audit 2001).
The key to improving natural resources management is ensuring that land uses (or activities) are compatible with the intrinsic characteristics of the Australian environment and accommodate the new matter and energy flows imposed upon it.The development and application of methods and practices that result in biophysically optimal spatial and temporal land use patterns is, therefore, a fundamental and urgent requirement (Lovering and Crabb 1997;Williams 1999).

APPLICATION OF THE ALS MODEL IN SOUTH GIPPSLAND, VICTORIA, AUSTRALIA
As part of the preparation of its Rural Strategy, the South Gippsland Shire Council (Map 1) sought to identify good agricultural land within its jurisdiction and protect it from unsuitable development.Eight agricultural commodities and commodities groups were selected for analysis including the main commodities grown in the region and those which offered potential for growth based on environmental and economic/market conditions.These were: brassicas (Broccoli, Brussel Sprouts, Cabbages, Cauliflowers, Chinese Cabbage, Chinese Broccoli, Kales, Jerusalem Artichokes and Mustard), carrots, cool climate grapes, peas, pome and stone fruit, potatoes, raspberries and sweet corn.Best management practices that could improve the growth potential of the commodities were also taken into account in the criteria.Experts were drawn from state government, other key public agencies (e.g.Catchment Management Authorities and water boards), industrialists, local growers and private consultancies (Sposito et al. 2000b).

Map 1 Location of South Gippsland Shire in Victoria, Australia
When different land attributes, such as soil characteristics, landscape forms and local climatic conditions, occur in certain combinations they can create a suitable environmental setting for various land uses and management.For instance, deep fertile, well drained soils in flat, temperate valleys with low winds and regular seasonal rainfall are likely to be suitable for growing a diversity of commodities.The following were the criteria sets considered in the analysis for South Gippsland.
Figure 2 shows the AHP hierarchy constructed for determining the land suitability of the brassicas commodities group, whilst Map 2 shows the resultant suitability map.According to the experts, three critical factors for Brassica are soil, slope and climate and soil and slope will have equal weights of influence (40%) whereas climate will have lesser influence (20%).Further sub-hierarchies shows factors influencing the major critical factors of soil, slope and climate.
In the case of South Gippsland, the various commodity suitability maps (a total of eight) were combined to produce the "Agricultural Suitability" map (Map 2).
For planning purposes, a range of overlays are also developed as components of the GIS for considering other significant environmental aspects of the region of concern.They include conservation values, native vegetation (Ecological Vegetation Classes -EVCs), species habitats and water bodies.Environmental risk is also incorporated by including water quality and dryland salinity layers.
In the case of South Gippsland, a set of five Policy Areas were determined by the integration of agricultural suitability (Map 3), environmental overlays and socio-economic information; i.e.; from a combination of the major environmental and socio-political factors driving natural resource management in the Shire.The 'Planning Strategy' framework (Map 4) was formed as a final product of the project.
Map 3 Agricultural Suitability Map of South Gippsland, Victoria and socio-economic factors that influence the decision for locating urban growth (Blakely 1994;Daniels 1999;DETR 2000).
The evaluation of land suitable for urban development is thus a critical planning task as it identifies opportunities and constraints for development in the region of concern.A Regional Development Plan is commonly a multi-objective planning instrument that often attempts to satisfy the needs of an expanding population without harming the environment and cultural values of the regional landscape.
In this context, the main purpose of the Urban Buildability Model (UBM), is the identification of areas suitable for urban development that do not adversely impact upon the environment.The decision-making involved in the formulation of an urban development plan for a region primarily involves two levels of analysis.Firstly, at the strategic level, the potential location(s) must be investigated within the context of the regional environment and economy.Secondly, at the local level, each available parcel of land (or site) needs to be considered in terms of its suitability for a specific land use (or activity).
The evaluation of the potential location(s) of urban development at a regional level may involve consideration of the following criteria: Spatial Framework -This is the spatial framework of the region under consideration in terms of planning and management.It relates to the evaluation of each location's ability to provide services that would strengthen the urbanrural continuum and allow for sustainable development across the region.This assessment involves analysis the location of potential service areas, and can be broadly measured in terms of distances from existing regional service centres and networks.
Environmental Sustainability -This relates to evaluation of the location in terms of its effect on the environment and attempts to ensure that the urban activities will not damage the long-term sustainability of the environment.This The evaluation of the potential location(s) for a specific land use (or activity) in an urban centre occurs within the framework defined by the regional analysis and may include following criteria.
Intra-urban Spatial Framework -This relates to the efficiency of the urban structure in providing services within the urban centre, and is achieved by establishing an appropriate network of service infrastructure and allocating the various land uses in suitable locations.This evaluation involves analysis of potential service areas in the location and can be broadly measured in terms of distances from intra-urban service centres and networks and compatibility between neighbouring land usages.Physical Characteristics -This relates to physical suitability (Land Protection Service 1985) at a site-specific level.Soil condition and landscape of a site is evaluated for a particular land use.
Model development follows a multi-criteria suitability modelling approach utilising experts to identify important factors and calculate weights for them through the AHP, as described in Section 2, above.
The UBM identifies ranges of suitability for urban development across an area.This allows for the optimal areas in terms of planning and management to be identified, as well as areas with poor suitability for urban development that would encounter construction difficulties and potential impacts to the environment such as encroaching into good agricultural land, endangered EVCs and significant landscapes if development were to proceed in these areas.
The broad criteria considered in the Urban Buildability Model are:

Physical Development Factors:
Building foundations, septic foundations, elevation and aspect of the landscape, wind, floodability

Planning and Management Factors:
Land use, planning zones, infrastructure

APPLICATION OF THE UBD IN SOUTH GIPPSLAND, VICTORIA, AUSTRALIA
Following on from the preparation of its Rural Strategy, described in the previous Section, the South Gippsland Shire required an in-depth investigation of the Coastal Policy Area (see Map 4) to determine the best possible locations for urban development.Consistent with the previous discussion, the study focused, first, at the regional level to devise a development strategic framework.Then, within the framework, it focused at the urban level in around the existing key settlements and in two greenfield locations for the possible construction of a new town (Sposito et al. 2002c).
Figure 3 shows a simplified and generic version of the UBM built for application at the regional level.Map 5 shows the regional strategic map for urban buildability (from low to high).For illustration purposes, Map 6 shows the resultant urban suitability map for the Venus Bay -Tarwin Lower twin towns which were earmarked for absorbing most of the growth in the coastal area of South Gippsland.Map identifies highly suitable land surrounding existing urban areas for further expansion and restricts areas constrained due to EVC significance and flooding conditions.

CONFIDENCE LIMITS IN LAND SUITABILITY MAPPING
Land suitability analysis using MCE methods incorporates a range of data sets that have an error component and, hence, errors are propagated throughout the modelling process.Most of the research (Goovaerts 1997;Molenaar 1998;Lowell and Jaton 1999) identifies a threefold breakdown of uncertainty types into: (1) data-derived, (2) display-derived, and (3) modellingderived.A fourth type, spatial variation, has also being identified (Hunsaker et al. 2001).In the specialised literature, there is no consensus on how to combine these different sources of uncertainty into one overall, mechanically-derived estimate of total model accuracy.Despite these problems, decisions still need to be made based on model outputs to underpin the sustainable management of natural resources.
In this context, research has been carried out to establish the confidence limits for land suitability mapping (Hossain and Wyatt 2004).Since soil information is a major input into the land suitability analysis and current soil maps lack information on accuracy, Figure 4 shows a framework developed, as part of the research, for establishing soil map accuracy by resurveying selected sample sites, creating an error matrix comparing with the existing maps and calculating confidence limits.

CONCLUSION
Land Resource Assessment provides a sound foundation for a holist and sustainable approach to strategic spatial planning.The following diagram (Figure 5) shows our land suitability modelling process and an indication of potential stakeholders and clients (agricultural extension managers, agricultural industry and regional land use planners) at each level of outputs.A Strategic Development Plan allows for the formulation of action plans and the consideration of desirable and undesirable land uses to be represented spatially for regional and local sustainable development.In the case of Local Government, this method provides vital information for the preparation of Municipal Strategic Statements and similar planning documents that plan for a social, economic and environmental sustainable future for the region.
The land suitability modelling process is under continuous development and refinement.We are now working on the incorporation of the financial prospects of commodities and more effective consideration of water availability into the model.Also, our Agricultural Land Suitability models have recently been integrated with Climate Change Impact models to predict the effect of climate change on agriculture/forestry activities.This is described in the article by Sposito (2006) in this issue of Applied GIS.
assessment involves identification of relevant environmental factors and ensuring their protection in view of environmental sustainability.Physical Characteristics -This relates to the broad physical suitability of the location in terms of urban development.It involves evaluation of soil, landscape and climatic factors considered critical to urban physical development.

Figure 4 A
Figure 4 A Framework for Determining Soil Map Accuracy Source: Hossain and Wyatt 2004

Figure 5
Figure 5 Land Suitability Modelling and Potential Stakeholders/Clients Source: Sposito et al. 2002c

Table 1
Commodity Land Suitability Models SUSTAINABLE LAND RESOURCE ASSESSMENT IN REGIONAL AND URBAN SYSTEMS ARTICLES 24.13