Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Decision support systems for land evaluation : theoretical and practical development Miller, David B. 1985

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1985_A6_7 M54_2.pdf [ 9.14MB ]
JSON: 831-1.0096193.json
JSON-LD: 831-1.0096193-ld.json
RDF/XML (Pretty): 831-1.0096193-rdf.xml
RDF/JSON: 831-1.0096193-rdf.json
Turtle: 831-1.0096193-turtle.txt
N-Triples: 831-1.0096193-rdf-ntriples.txt
Original Record: 831-1.0096193-source.json
Full Text

Full Text

DECISION SUPPORT SYSTEMS FOR L A N D EVALUATION: THEORETICAL A N D PRACTICAL DEVELOPMENT by DAVID B. MILLER B.ES. (Plan.), University of Waterloo, 1981 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF T H E REQUIREMENTS FOR T H E D E G R E E OF MASTER OF SCIENCE in T H E F A C U L T Y O F G R A D U A T E STUDIES ' INTERDISCIPLINARY STUDIES (Resource Management Science) We accept this thesis as conforming to the required standard T H E UNIVERSITY OF BRITISH COLUMBIA OCTOBER, 1985 I ® DAVID B. MILLER, 1985 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the The University of British Columbia, I agree that the Library shall make it ^  freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. INTERDISCIPLINARY STUDIES The University of British Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Date: OCTOBER. 1985 ABSTRACT The challenge of resolving land use allocation and policy questions depends to a large degree on the conversion of data into information, and the effective integration of information into the decision process. Land evaluation is one of the fundamental means of generating information for land planning. Information products have however, been inconsistently and ineffectively used in the decision process. This thesis develops a decision centered approach to land evaluation as a response to this concern. Included in this development is a description of important theoretical concepts, as well as a practical demonstration of the use of decision support systems as a design approach. Initially, a conceptual model is introduced illustrating the technical and use components of information generation, as well as the adaptive design cycle. Various terms and techniques involved in the technical aspects of land evaluation are reviewed. Decision making concepts including decision structure, environment, analysis, and criteria are outlined. Three existing methods of land evaluation are then compared from a use or decision making perspective. Having completed a review of current approaches, Decision Support Systems are introduced as a logical progression towards a decision centered approach. Decision Support System design is demonstrated using a portion of the Central Fraser Valley Regional District as a case study area combined with an interactive microcomputer land planning tool (LANDPLAN). The demonstration emphasizes the advantages of the flexible, interactive capabilities of Decision Support Systems in aiding the decision process. Iterative design is also promoted with several needs identified if a more complete system is to be developed. In particular, data on strategic long term supply and demand factors is required, as well as continuous rating functions for assessing land performance. ii Table of Contents ABSTRACT ii LIST O F TABLES vii LIST O F FIGURES viii A C K N O W L E D G E M E N T S . ix 1. INTRODUCTION 1 1.1 C O N C E P T U A L M O D E L 1 1.2 PROBLEM S T A T E M E N T A N D EXPLORATION 4 1.3 THESIS OBJECTIVES A N D OUTLINE 9 2. OVERVIEW O F L A N D CLASSIFICATION A N D E V A L U A T I O N 12 2.1 L A N D CLASSIFICATION 13 2.1.1 Identifying Land Units 15 2.1.2 Parameter Assessment 16 2.2 L A N D E V A L U A T I O N 19 2.2.1 Approaches to Land Evaluation 19 Land Suitability Analysis 21 F A O Framework 23 Plan Evaluation 26 Programming Techniques 28 2.3 S U M M A R Y 29 3. DECISION M A K I N G CONCEPTS 31 3.1 DECISION DIAGNOSIS 32 3.1.1 Decision Environment 32 3.1.2 DESCRIPITIVE MODELS 35 3.2 DECISION ANALYSIS 40 3.2.1 Techniques 41 3.3 DECISION CRITERIA 45 3.4 S U M M A R Y 50 iii A N A L Y T I C A L COMPARISON A N D ASSESSMENT O F C U R R E N T L A N D E V A L U A T I O N SYSTEMS 51 4.1 T E C H N I C A L COMPARISON 51 4.1.1 Ecological Land Evaluation 51 4.1.2 S I R O - P L A N 54 4.1.3 Land Evaluation Model 56 4.1.4 Summary 61 4.2 A N A L Y T I C A L COMPARISON 61 4.2.1 Ecological Land Evaluation 61 Decision Environment ; 61 Descriptive Models 63 Decision Criteria 64 4.2.2 S I R O - P L A N 66 Decision Environment 66 Descriptive Models 67 Decision Criteria 67 4.2.3 Land Evaluation Model 70 Decision Environment 70 Descriptive Models 71 Decision Criteria 71 4.3 S U M M A R Y 73 DECISION SUPPORT SYSTEMS 75 5.1 E V O L U T I O N 76 5.2 DEFINITION 79 5.3 D E S I G N 80 5.3.1 Decision Framework 81 5.3.2 Technology levels 81 5.3.3 Tactical Options 83 iv 5.3.4 Design Guidelines 83 5.4 Roles in DSS 85 5.5 S U M M A R Y 87 6. A DECISION SUPPORT SYSTEM FOR L A N D E V A L U A T I O N 88 6.1 S T U D Y AREA 88 6.2 E V A L U A T I V E TASK 91 6.3 INTELLIGENCE 92 6.3.1 Decision Diagnosis 92 Physical Environment 92 Decision Environment 98 6.3.2 Available Information 101 6.3.3 System Requirements ...102 6.4 DESIGN 103 6.4.1 Development Options 103 6.4.2 R O M C Framework 104 6.4.3 Technique Selection 104 6.4.4 L A N D P L A N 105 Policy Set 106 Data 108 Rating Codes 112 6.4.5 Interpretive Matrix 115 6.4.6 System Summary 115 6.5 C H O I C E 118 6.5.1 System Analysis 118 6.5.2 Applications 119 6.5.3 Iterative Design 125 7. CONCLUSIONS 134 v REFERENCES 136 APPENDIX O N E 146 APPENDIX T W O 147 APPENDIX T H R E E 155 APPENDIX F O U R 163 vi U S T O F TABLES Table Page 1. Decision Analysis Techniques 42-43 2. Technical Summary of Current Approaches 62 3. Comparative Summary of Current Approaches 74 4. R O M C Framework 86 5. Land Use Conflicts - C F V R D 99 6. Decision Participants - C F V R D 100 7. R O M C Framework - C F V R D 105 8. L A N D P L A N Data Items 112 9. L A N D P L A N Data Assessment 113 10. Sensitivity Analysis 123 11. Impact of Value Scenarios 124 12. Data Assessment - Value Scenarios 129 vii LIST O F F I G U R E S Figure Page 1. Conceptual Model of Information Generation and Use 2 2. Information Degeneration Cycle 6 3. Conceptual Model Including Decision Support Systems 10 4. Information Classification Levels in Planning 14 5. Decision Environments 33 6. Decision Levels and Information 36 7. The Structure of Land Evaluation Decision Problems 37 8. Decision Analysis Technique Selection Process 46 9. Ecological Land Evaluation 53 10. Land Evaluation Model Structure 59 11. Land Evaluation Model Evaluative Output 60 12. Canada Land Data System 78 13. Intelligence - Design - Choice Paradigm 82 14. Technology Levels in Decision Support Systems 84 15. Design Process for the Decision Support System Demonstration 89 16. Location Map of Study Area 90 17. Study Area Landscape Map 93 18. Map of Study Area Planning Units 110 19. Map of Equal Policy Weight Scenario 116 20. Land Use Compatibility Matrix 117 21. Use of the Land Use Compatibility Matrix 121 22. Map of the Agricultural Value Bias Scenario 126 23. Map of the Development Value Bias Scenario 127 24. Map of the Conservation Value Bias Scenario 128 viii A C K N 0 W 1 IFTXIFIMFINTS Completion of a thesis and graduate program requires considerable support and direction. In this regard, 1 would especially like to thank Dr. Hans Schreier, thesis supervisor, for his critical comments, general direction, and continuous support throughout my studies. I am grateful to Mr. Anthony Dorcey for his invaluable assistance with theoretical direction and useful critiques of the thesis. Dr. Pat Miller and Dr. Les Lavkulich also deserve thanks for their advice while serving on my committee. In addition, I would like to acknowledge the financial support of the Natural Sciences and Engineering Research Council during my two years at graduate school. Finally, my thanks and appreciation go to my father for his advice and encouragement, and to the rest of my family for their moral support ix Chapter 1 INTRODUCTION With increasing demands being placed on land, public and private decision makers are having to regularly resolve land use conflicts by determining use allocation and policies. This has become a very complicated, difficult and often contentious task (Flaherty and Smit, 1982). Furthermore, the information analysis involved in this task is becoming increasingly complex with the many variables and disciplines involved (Beek, 1978). The task of resolving land use conflicts depends to a large degree on the conversion of data into information, and the effective integration of information into the decision process. Data can be defined as a symbol representing a state of affairs (Murdick, 1980). A soil pH of 7.5 is, for example, one data item encountered in soil analysis. If allowed to stand on its own, such data is not particularly revealing to anyone no intimately familiar with the data item. The volume of data involved in most complex analytical tasks will also quickly exceed human processing and cognitive capabilities (Simon, 1969). Data must therefore be converted into useful, meaningful information. Information is the knowledge gained when a meaning is assigned to data which has been retrieved, processed and interpreted (Dumanksi and Kloosterman, 1978). Information may predispose a person to action while data are not a stimulus in most situations (Murdick, 1980). 1.1 CONCEPTUAL MODEL Figure 1 is a simple illustration of the process from data collection to decision making. It also acts as the conceptual model for much of the thesis and will be explored initially through this introduction. 1 DATA - Soils - Land use - Resources CONVERSION - Generation of i nformati on - Land c l a s s i f i c a t i o n and evaluation INFORMATION - Land units - Value measures DECISIONS - Use of information to make choices between a l t e m a t i ves ADAPTIVE DESIGN AND USE FIGURE 1 - Conceptual Model of Information Generation and Use. 3 The importance of effective and efficient conversion of data into information with reduced complexity is beyond question. Thie, Wiken and Ruben (1984) support this view: The effective use of our land resources thus depends to a great extent on delivering the right information, in the right format and at the right time to the decisionmakers. If one of these criteria is not met, otherwise valuable information may have little or no impact on the decisions. By understanding the overall information needs and information technology such situations can be avoided (p.l). Information for land resource planning and management is generated in a variety of ways. This thesis deals with the generation of information through land evaluation techniques. Land evaluation is a broad concept which means different things, to different people, in different contexts. At this juncture, a definition by Girt et al. (1976) is offered as a beginning, with a more detailed discussion provided in Chapter Two. They explain that land evaluation involves: ... a system which indicates the relative worth or utility of allocating a particular use, as opposed to all others considered, to an area. (p.3). The value measure which results can be an absolute quantitative measure of land performance, or a qualitative ranking of land quality. Whatever the product or output, these value measures are the key to land evaluation. They become the indicators of relative importance which are used as guides in making choices between alternatives (Sinden and Worrell, 1979). This leads to the decision making component of the conceptual model. Decision making involves choosing between alternatives. These alternatives can represent goal setting or goal achievement (Inbar, 1979). Goal setting involves the translation of aspirations into articulated goals while goal achievement deals with alternative actions to implement the goals. A decision may be made to supply more recreational opportunities, followed by a choice of alternative locations. Decision making also involves searching for a good argument on which to base choice (Montgomery, 1983). 4 Decision making hinges on both factual and value premises (Inbar, 1979). Information therefore has a factual and value content The factual content, as represented by measures such as land productivity, is widely recognized and usually becomes the focus of attention during the conversion component. Value content, reflecting such influences as societal wants or political bias, is rarely well defined or explicit It is nevertheless, an equally integral part of the decision making component 1.2 PROBLEM STATEMENT AND EXPLORATION Ideally, the conceptual model functions as a continuous cycle improving decision making as iterations occur. Data is converted into information which is used in decision making. If the information itself proves inadequate, then corrective steps are taken to improve the conversion component Conversely, if the use of the information is not satisfactory or appropriate, then improvements are made through increased understanding or procedural changes in the decision process. Unfortunately, this model has not been functioning in a consistent or effective manner. Environmental information is inconsistently, and some would argue rarely, employed effectively in land use decision making. A problem exists when there is a difference or gap between a current state of affairs and a desired state of affairs (VanGundy, 1981). A problem gap can certainly be detected in the use of environmental information in land resource decision making. As noted by Smit et al. (1981): It is apparent from the controversy which continues to cloud land use decision making that the ability to evaluate land use alternatives in a comprehensive manner remains rather limited. Despite the availability of an impressive inventory of land resources in Canada, few mechanisms are available to use this information to determine whether there is a need to preserve land for certain use (p.4). Petak (1980) agrees: Although technical methods of dealing with environmental problems are being perfected, application of those methods lacks clear direction. However, even if the perception of aims were clear, the wide choice of means 5 provided by modern technology makes accomplishment of these aims a very complex undertaking (p.287). Complaints regarding the collection and. use of information permeate a wide variety of studies (see McCurdy and Meyers, 1978; Coastal Zone Resource Subcommittee, 1978; Holdgate, 1981; Dearden, 1978). While it may be easy to detect a problem gap, it is more difficult to identify causes and, ultimately, provide and implement solutions. Use of information in land use planning provides a graphic example. While the problem gap is recognized, solutions have been slow in coming. This becomes easier to understand when the plethora of causes, and potential obstacles to solutions, are considered. The persistent generation of information of marginal usefulness can be partly attributed to the information degeneration syndrome. This occurs when information is produced and analyzed without an examination of how and why the information is to be used (Lang, 1978). In other words, the link between the conversion and decision making components is dysfunctional. Figure 2 provides a schematic illustration. When degeneration occurs, Lang (1978) explains that: Information gathering becomes an end in itself, perhaps even a subsitute for action, in the hope that the data will inspire flashes of insight and reveal their significance to us (p.228). The technical problems involved in land evaluation can certainly be overwhelming. Between accumulating the necessary data, choosing classification approaches, comparing data bases, and attempts at quantification, the 'how and why' considerations of use are often lost In addition, technical improvements are not necessarily the answer to decision problems in every case. Improvements in technology are not always matched by concurrent understanding of, or ability to assist, the decision making process (Cook and Hammond, 1978). The problem is most acute when technical experts who complete evaluations, work in isolation from decision makers, who have to use the information. 6 know how and why information i s to be used i n the decision process analyze and co l l e c t available information generate new information degeneration occurs when working r e p e t i t i v e l y around t h i s loop without ascending to a higher l e v e l FIGURE 2 - Information Degeneration Cycle (Adapted from Lang, 1978). The blame for information degeneration should not rest entirely with those responsible for information generation. Information users, especially within organizations, can actively encourage degeneration. When dealing with information, organizations have been known to operate in a surveillance mode more than a problem solving mode. In this circumstance, information becomes an end in itself as a signal and symbol of competence in decision making. March and Shapira (1982) summarize the ineffectiveness of organizations when managing information: They gather information and don't use it Ask for more and ignore it Make decisions first and look for the relevant information afterwards. In fact, organizations seem to gather a great deal of information that has little or no decision relevance, information that, from a decision theory point of view, is simply gossip (p.777). Aggravating the problem is the lack of ability to organize and take advantage of the information which does exist When a decision problem is confronted, important 7 information is often poorly understood or poorly organized. Many have also warned against scattered and unorganized data and emphasized integration of information sources as a prerequisite for effective use (Kessel, 1984). Even if useful information exists, the proper evaluation technique may not be selected, or the results may be inappropriately applied. Poulton (1983) lists five reasons for poor choice or application of evaluation methods. 1. Failure to appreciate limitations inherent in the evaluation task. 2. Failure to relate evaluation methodologies to the information needs of decision makers. 3. Failure to consider evaluation as an integral part of the planning process. 4. Failure to categorise the available methods so that an efficient search amongst them can be conducted. 5. Failure to identify flaws in the logic or interpretation of the procedures adopted. These failures are particularly important and are emphasized throughout this thesis. One final obstacle created by the nature of environmental problems deserves mention. A dichotomy exists between environmental systems operating in an integrated holistic manner, and the decision making or institutional forums operating in a disjointed, separated manner (Lang, 1978; Dearden, 1978). This creates the dilemma of having to produce information which reflects the nature of environmental systems for use in a decision making forum ill-prepared for the necessary ecological perspective. Attempts to express ecological concerns through institutional systems have often been difficult (Holling, 1980; Lang, 1978; Dearden, 1978). Hickling (1975) also emphasizes the conflict created by trade-offs in the decision process. He identifies four key tradeoffs: 1. The need for simplification to facilitate understanding and analysis vs. the need for complexity to reflect reality. 2. The urgency of providing timely information vs. the lack of available information. 3. The need for defensible commitment vs. the need to be flexible and experimental. 8 4. The need for incremental analysis to reflect the realities of policy development vs. the need for comprehensiveness. These conflicts are difficult to resolve, partly as a result of uncertainty over (1) the operating environment, (2) policy values and , (3) choices in related decisions. This is not to say that the situation is beyond hope. Effective data collection and information analysis is making a positive contribution to land resource decisions. Information contributing to land use decisions is certainly vastiy improved - from earlier efforts. Still, even where obstacles to the effective use of information seem insurmountable, technical specialists and decision makers must continue to strive for improvement To this end, a number of general areas requiring improvement and research have been identified. These include: 1. More emphasis on socio-economic - land use relationships like the calibration of capability classes with crop yields or economic returns (Beek, 1978; Burroughs, 1976; McRae and Burnham, 1981) 2. More emphasis on producing functional information based on "an understanding of svstem relationships, rather than just descriptive inventories (Beek, 1978; Dorcey and Hall, 1981). 3. Less emphasis on research for its own sake and more on incorporating the results of information generation into the decision making process ( Burroughs, 1976; Lang, 1978) This thesis examines the last of these concerns, namely the interaction of land evaluation with decision making. The contribution of this thesis therefore, lies not in the advancement of land evaluation technology per se., but in the combination of decision making perspectives and requirements and land evaluation technology to demonstrate a decision centered approach to land evaluation. The result should provide insight into the design of more effective land evaluation systems as well as directions for future information research, generation and evaluation. A logical evolution of the concepts involved in a decision centered approach to land evaluation is the development of Decision Support Systems (DSS). These are 9 highly interactive and adaptive tools for the conversion of data into information. They are designed to improve and complement, not replace, the judgement of the decision maker. Figure 3 completes the conceptual model introduced earlier by inclusion of Decision Support Systems as an approach to operating the basic model. 1.3 THESIS OB.TECTTVES AND OUTLINE Having introduced the conceptual model and the functional problems currently preventing its effective operation, a goal and several objectives for this thesis can be identified. The goal of this thesis is to: Develop and demonstrate the design of an approach to land evaluation appropriate to decision making requirements. This goal can be divided into several objectives: Outline important decision making concepts as they relate to land evaluation. Demonstrate the linkage between land evaluation and decision making by comparison of existing land evaluation approaches. Outline and demonstrate the use of a decision support system framework for design and use of land evaluation systems. These objectives are pursued by first examining the current state of the art using the conceptual model illustrated in Figure 1 as a guide. Decision Support Systems are then introduced as a logical path towards improving the generation and use of information in land resource planning. Chapter Two deals with the technical aspects of converting data into information using land evaluation. The following chapter examines the decision making component of the conceptual model by introducing important elements of the decision process. The link between these two is established in Chapter Four with a comparative analysis of three existing land evaluation systems, namely ecological land evaluation, SIRO-PLAN, and a Land Evaluation Model. Having examined current knowledge and 10 DATA - Soils - Land use - Resources CONVERS ION - Generation of Information - Land classification and eval nation INFORMATION - Land units - Value measures DECIS IONS - Use of information to make choices between alternatives ADAPTIVE DESIGN A N D U S E DSS Evaluative ou tDu t (i nformation) Representations Operations Memory Aids Control Aids O O O O O DECISION SUPPORT SYSTEMS DSS Tools - Data, Comouter language FIGURE 3 - Conceptual Model Including Decision Support Systems. 11 approaches, a design framework is introduced based on Decision Support System theory. Finally, Decision Support System design is demonstrated in an area of the Central Fraser Valley Regional District in the Lower Mainland of British Columbia using the interactive micro-computer tool LANDPLAN. Chapter 2 OVERVIEW OF I A N D CI OSSIFICATION AND EVALUATION An important tool in the conversion of many types of data into information is the idea of classification and subsequent evaluation. Scientific classification involves the establishment of categories with a high degree of internal similarity or association combined with clear distinctions between groups. Classifications are used to combine and organize data into more manageable or revealing information sets. Land classification involves the assignment of categories to the earth's surface (Flaherty and Smit, 1982). These categories are usually created by identifying a combination or complex of land attributes which, by varying individually and in relation with other attributes, create a local character (Mabbut, 1968). The task of converting land data into information through classification involves a confusing array of terms and concepts. Adding to the confusion is the often ambiguous, inconsistent, and even contradictory meanings attached to these terms. Leaving aside the various terms used to describe particular concepts and approaches temporarily, a general evaluative classification procedure can be identified. This general procedure was concisely summarized by B. Smit et al. (1984) as follows: It invariably begins with basic surveys of soil, water, climate and other characteristics of the biophysical resource. These data are used in the compilation of an inventory of land resources. Land units are identifed which are relatively homogeneous with respect to biophysical attributes. For land evaluation purposes, the biophysical attributes employed are those which relate most closely to the requirements of the uses being considered. Once land units are identified, it is then necessary to predict or rate their performance for particular uses or groups of uses (p.468). Traditionally, this procedure results in three types of products or information classifications (Girt et al.,1976 ; Smit et al.,1981). These can be identified as: 1. Classifications showing inherent characteristics with no explicit interpretation or reference to land use. For example, a soil survey will group similar soils into soil orders, groups or types depending on the level of detail involved. 2. Classifications showing characteristics or information sets which have implications for land use. A land capability classification for agriculture provides a recognizable illustration. 12 13 3. Classifications showing land use recommendations and policies. In this case, the surface categories represent land use zones or policy areas corresponding to the desired land use pattern, however determined. Land evaluation provides a link between the second and third levels of information. (Smit et a l , 1981). This typology is illustrated in Figure 4. 2.1 LAND CLASSIFICATION Before considering the place of land evaluation in more detail, it is important to identify and define some fundamental concepts involved in the first two types of classification. The traditional land classification process involves two basic decisions -how to distinguish between or group land units, and what index of performance to use. Land classification starts with land characteristics which are simply attributes that can be measured or estimated (Brinkman and Smyth, 1973). A set of interacting land characteristics can express a land quality which is a distinct factor influencing land performance for a particular land use (Vink,1975; Beek and Benema, 1974). For instance, availability of water and nutrients are major land qualities related to plant growth (Beek and Benema, 1974). Land qualities are the simplest direct functional link between biophysical data and land use. As such, they may be used as diagnostic criteria to identify properties worthy of consideration for higher level (type two) land classification projects (Beek, 1978). Although land qualities are usually used as tools in the transition between data and some measure of performance, they can be left as type one classifications on their own. Given basic land data in the form of characteristics or qualities, the means of combining the qualities to determine land units and performance ratings is the next consideration. Two types of procedures are important The first procedure explicitly identifies homogeneous land units. The second combines land qualities to generate performance ratings. These ratings can be applied to predefined units created by the 14 Inherent Characteristics - Surveys of attributes and qual i ties ( s o i l survey, vegetation map) LAND CLASSIFICATION Interpretive Characteristics - Indication of single use performance rating (capability) \ LAND EVALUATION Allocative Characteristics - Indication of recommended use or policy (zoni ng map) \ F I G U R E 4 - Information Classification Levels in Planning (Adapted from Smit et a l , 1981) 15 first procedure, or used to create units by grouping land areas based on the ratings themselves. 2.1.1 IDENTIFYING LAND UNITS Mabbutt (1968) described three general approaches to identifying land units: Genetic, Landscape, and Parametric. Genetic approaches involve subdivision of land units on the basis of causal or genetic environmental factors. Examples include the delineation of natural regions based on climate or morphology and classification of soils by soil genesis. Although genetic classifications may provide some useful general information, they are not an integral part of most land evaluation exercises because the regions produced are "not of such scale and precision as to be manageable land units" (Mabbut, 1968, p.13). In an effort to create a more practical approach, the search for genetically defined natural regions was largely abandoned in favor of identifying land units based on distinctive observed or visible land character. This is known as the landscape approach. It takes an integrated view of the land complex, foregoing the identification of specific land attributes for a sense of a complete landscape created by a general pattern of distinctive components. For example, a landscape unit consisting of a floodplain in association with scrub vegetation can be identified from examining an air photo and thus included as a distinctive unit in a landscape classification. There are advantages and disadvantages to the landscape approach. It can yield a reasonable level of information fairly quickly at a low cost General recognition of land patterns can result from a limited sample base (Mabbutt, 1968). The approach can however, be criticized on several counts. Unless the procedure is very sophisticated and mechanized, with sensors replacing human judgement, it can suffer from a lack of reproducibility. In other words, two people can examine the same area and yet define markedly different land units. A landscape classification does 16 not supply much specific information amenable to quantitative analysis. In some circumstances, this can be significantly limiting. The parametric approach can be defined as the "division and classification of land on the basis of selected attribute values" (Mabbutt, 1968, p.21).- It involves the identification of critical attributes, the formation of specific classes based on attribute values, and the grouping of land areas into units' according to those value classes. The parametric approach has several significant advantages. Being based on specific attribute values, it can be standardized thereby removing subjectivity and increasing reproducibility (Riquier, 1974; McRae and Burnham, 1981). The approach is also quantitative and specific which facilitates further analyses. In fact, many of the techniques used to determine performance ratings in conjuction with a type two classification are parametric in nature. Parametric approaches are not however without their disadvantages. The quantitative and specific nature of the classification can give a misleading impression of accuracy with the resultant mistrust when imperfections are identified (McRae and Burnham, 1981). It can also be difficult to choose and isolate individual land attributes. The interdependance between attributes, and our incomplete knowledge of the functional interactions in many land systems, may lead to classification on the basis of an attribute which does not in fact make a significant contribution to land character and performance (Riquier, 1974). 2.1.2 PARAMETER ASSESSMENT The classifications resulting from the previously described procedures do identify homogeneous land units. They do not however, provide any link to land use. To provide this link, it is necessary to select and evaluate land characteristics and qualities to act as diagnostic parameters. For example, soil fertility and land drainage are frequently used as diagnostic parameters in the assessment of agricultural land 17 capability. This involves selecting those qualities which are critical in determining land use performance or impact Selection can however, be a difficult and uncertain task unless the functional relationships are very well understood (Nix, 1968; Riquier, 1974). Once selected, parameters can be used to provide a basis for a single use performance or capability rating. This corresponds to the second level of information in Figure 4 as well as being an initial stage in many land evaluation exercises. The selection of qualities to serve as parameters depends on the rating or index used. A performance index or rating is an indication of how well a particular land unit will meet an objective dictated by the the rating. This could involve production of a crop, environmental conservation measures, or any other element important for a particular land use. Four basic indices should be defined: productivity, vulnerability, capability, and feasibility (Smit et al, 1984). A productivity classification assesses land units based on the physical yield that would result if the unit was allocated for a given use with specified management and input levels (Smit et al., 1984; Nix, 1968). Vulnerability refers to the potential environmental impact of an activity or the degree to which a use would affect the physical quality of an area (Smit et al., 1985). The concept of carrying capacity can also be linked to vulnerability. As a land performance rating, carrying capacity analysis is often completed in tandem with other ratings by placing values on limiting factors in an effort to determine the maximum level of growth and development (Ortolano, 1984). A feasibility rating attempts to measure the consequences of a proposed use together with its economic viability (Dent and Young, 1981). Land capability and suitability are often used interchangeably but there is a distinct difference. To be capable is to be qualified for or have the ability to support a use while to be suitable for a use implies an additional aspect of appropriateness requiring a higher level of evaluation (Steiner, 1983). Land capability therefore refers 18 to the ability of an area to support a particular use with varying results without permanent damage (Lang and Armour, 1980). Capability assessment provides a common example of an interpretive, single use classification using land qualities for direct assessment This method assesses a land unit on the basis of limiting factors imposed by permanent land characteristics (Davidson, 1980). The method depends on several assumptions including good management practices and permanent characteristics. A common example is the seven class capability classification used in the Canada Land Inventory (see Coombs and Thie, 1979). Capability assessments are rapid and simple and can be useful at the reconnaisance or regional level for general analysis. More detailed assessments are however difficult and direct economic interpretation is constrained by the scale of most capability assessments and the assumptions relating to management practices. To this point basic enviromental data has been converted into information through land classification and land qualities have been assessed for single uses. Land use policy and allocation, as represented by the third stage in Figure 4, is often based on the information supplied by these first two levels of information generation. Agricultural land preservation is a case in point Policies and regulations controlling land use in rural areas to preserve agricultural land have been based primarily on capability mapping. The Canada Land Inventory mapping form the foundation of both the Foodland Guidelines in Ontario (Government of Ontario, 1978) and the Agricultural Land Reserves in British Columbia. This has however been questioned as an adequate information base from which to make land use decisions (Girt et al., 1976; Smit, 1981; Rees, 1977). As noted by Smit (1981): While such schemes provide useful information on the land resource, they do not consider societal needs or requirments from the use of land, nor do they consider competing uses for land (p.l). Other deficiencies include the lack of quantitative measurement and the static nature 19 of the assessment To provide a higher level of information more appopriate for land policy planning, attention has turned to land evaluation systems. 2.2 LAND EVALUATION As noted previously and illustrated in Figure 4, land evaluation bridges the gap between more traditional classifications and land use policy. It involves more than the assessment of qualities for a particular use, but also a measure of value. Land evaluation has been previously defined as a procedure which indicates the relative utility of a particular use in a particular area. This definition could obviously embrace a wide variety of systems. Some generally agreed upon principles can serve to narrow the focus: 1. Land evaluation should include consideration of the inputs or other resources, required to make a land use viable (Girt et al, 1976; FAO, 1977). 2. Evaluation for one land use cannot be completed without considering its competitors for land (FAO, 1977; Smit, 1981). 3. Evaluation is a multidimensional procedure examining several uses simultaneously, with each use having a number of dimensions. A multidisciplinary approach is therefore required (Girt et al, 1976; FAO, 1977). 4. Evaluation must be relevant to the physical, economic, and social context of the subject area including the economic system and level of development, the available land base, and the level of land use conflict (FAO, 1977) 2.2.1 APPROACHES TO LAND EVALUATION The fundamental difference between land evaluation systems is in the method of assigning value and the related allocation rules (Gold, 1974). It usually dictates the technical approach from data organization to comparative mechanics. Measures of value can be approached in several general ways. Measures of value can be intrinsic . based strictly on the land unit itself, or extrinsic which includes the locational setting of the area in the evaluation. Land evaluation can also be approached from a general or specific use perspective. A general approach attempts to derive a measure of the relative value of a variety of uses if located in a particular area. Conversly, a use specific evaluation begins with a use in mind, and evaluates several areas for that use. A distinction can also be drawn between physical land evaluation - assessing the physical ability to support a use - and integral land evaluation - considering basic economic and other societal goals in relation to physical capability (Smit et al., 1984). When considering socio-economic factors, two general' approaches can be taken. In the two-stage approach . a preliminary physical evaluation is completed with socio-economic analysis undertaken as a second stage on physically attractive alternatives. If a parallel approach is taken, the socio-economic criteria is included throughout the evaluation (FAO, 1983). Value can also be current or potential . A current evaluation considers only the present situation, without major land improvements or socio-economic changes. Potential value refers to a measure based on assumed improvements or changes. Finally, a value measure can be qualitative or quantitative . A qualitative value is descriptive and rarely provides anything beyond a relative measure. On the other hand, a quantitative evaluation attempts to produce a meaningful number which indicates value on its own. Land evaluation may also include different forms of measurement Four scales can be identified (Hill, 1968): 1. The nominal scale which classifies and numbers entities. 2. The ordinal scale which ranks entities. 3. The interval scale which provides equal intervals between entities and indicates the differences or distances of entities from some arbitrary origin. 4. The ratio scale which provides equal intervals between entities and indicates the differences or distances of entities from some nonarbitrary origin. As will be demonstrated shortly, the scale of measurement is an important consideration and should be understood when using a product As the general value measure varies, so to do the means for deriving the measure. This variety can be illustrated by documenting four types of value measures 21 and the techniques related to them. Land Suitability Analysis In many cases, there is insufficient data or expertise to adhere to the principles of quantitative land evaluation. As a result, the evaluation may be completed in relative terms with largely qualitative measures to produce a land suitability analysis (Vink, 1975; McDonald and Brown, 1984). The fitness of a given tract of land for a defined use is a commonly accepted definition of land suitability (Steiner, 1983; Vink, 1975). Fitness is meant to include a consideration of outputs from the land as compared to the necessary inputs and environmental conditions (Smit et al., 1984; Steiner, 1983). Suitability can also include market, nonmarket, and nonmonetary costs and benefits (Hopkins, 1977). Land Suitability Analysis involves a high degree of expert judgement in the selection and measuring of land characteristics, assigning ratings or priorities for composite suitability, and estimating land conversion impacts. The approach is meant to emphasize the physical impact of land use, but these impacts are implicitly measured by the inclusion of data on physical land characteristics (McAllister, 1980). The output of a land suitability analysis is usually a set of maps, showing the relative level of suitability for each land use as well as summary which may indicate composite suitability or preferred use. The resulting product can be visual, showing degrees of suitability with colour shades, or quantitative, with numerical ratings assigned to the degrees of suitability (McAllister, 1980; Hopkins, 1977). Popular examples of land suitability analysis include the American Soil Conservation Service Agricultural Land Evaluation and Site Assessment (LESA) system (Steiner, 1983) and the McHarg approach which is discussed in detail in Chapter Four. The LESA system generates a value measure by combining the soil potential index with an agricultural site assessment (ASA) which considers factors other than soil quality (Steiner, 1983). Soil potential index (SPI) is a physically based assessment formula with the form: 22 SPI = P-CM-CL wherein P is the performance measure in dollars, CM is the relative cost of corrective measures and CL is the relative cost as a result of continuing limitations (Steiner, 1983). During a land suitability analysis, information pertaining to a variety of land qualities and a number of land uses must be combined. There are a variety of techniques available to compare and combine information. Hopkins (1977) provides a review of these techniques. The least rigorous approach is to rely on gestalt judgement The gestalt method views the landscape as a whole and is related to a landscape classification. Land units are described verbally and expert or gestalt judgements are made regarding constraints and opportunities presented by the area in question (Hopkins, 1977). Obviously, gestalt assessment is difficult to defend and few people have the expertise to make credible gestalt judgements. While it is recognized that some gestalt judgements are necessary for land rating, a 'feel of the land' is not enough for an evaluative assessment (Brinkman and Smyth, 1973). Another basic approach involves the use of map overlays and ordinal combination. It involves three steps as follows (Hopkins, 1977): 1. A map is prepared for each relevant factor or land quality 2. A table or other device is used to rate the relative performance of each factor when combined with each land use. 3. Overlay suitability maps Of individual factors to produce a composite map for the suitability. The suitability maps can then be examined on their own or overlayed with one another to indicate the most 'fit' use for a land area. Two major technical criticisms are made of ordinal combination. Firstly, to be valid, the numbers or shades in the index should be on an interval scale. If the differences between shades or numbers are not equal for each factor, direct comparison 23 is not necessarily valid (Hopkins, 1977). Secondly, the operation is equivalent to addition, which requires the assumption that each factor is independent (Hopkins, 1977). Once again, this is rarely the case. A common response to these criticisms is to weight the various factors and use a linear combination method. In this case, the factors are rated on separate scales and then combined using an importance weight for each factor. This changes the measurement units by the ratio of the importance weight to create a common interval scale. The importance weight can be manipulated in a number of ways including the use of a standard formula for a weighted average, transference to a common value range, or a simple scoring system where each factor is assigned a certain percentage of the rating (Hopkins, 1977). While this method does deal with the measurement problem, it does not address the possible interdependance between factors. Interdependance is however considered in the nonlinear combination method. A nonlinear combination involves combining factors on the basis of known relationships that can be converted info mathematical functions. Unfortunately, few relationships are well enough understood in terms of suitability to be converted into nonlinear combinations. A second group of techniques are based on logical combination (Hopkins, 1977). These techniques assign ratings to sets of combinations or factor types on the basis of verbal rules. For example, a specified rating could be assigned if a land use had a particular soil type in combination with a steep slope. Carefully constructed rules which are sensitive to the natural system can handle interdependence between factors. Logical combination methods are now becoming popular through the use of expert systems which try to mimic the question and answer sequence of expert advice (Davis, 1985). FAO Framework A second general approach to land evaluation is provided by the Food and Agriculture Organization of the United Nations (FAO). The FAO framework for land 24 evaluation has been promoted in a number of publications (FAO, 1977; FAO, 1983; Beek and Benema, 1974). It is designed specifically for agricultural land use and uses an agricultural systems approach, examining both the inputs and the outputs of the land use system. The value measure is derived from the combination of land qualities, physical inputs and outputs, and land requirements, all of which are included in the search for an optimal combination (Beek, 1978). The evaluation begins by first describing land utilization types (LUT's). A LUT. represents a production or land use system described in some detail on the basis of such attributes as capital and labour intensity, market orientation, technical refinement, and tenure characteristics. For example, a dairy and pasture agricultural system is a land utilization type frequently found in North America. Land types are also identified along with the accompanying land qualities. The LUT is then compared or matched with the land type. This requires an assessment of the productive requirements and response of the land type. Three approaches are most frequently encountered when relating productivity to land type: Analogue, Site Factor, and Systems Analysis. Analogue methods involve extrapolating results from specific sites to areas considered similar or analogous (Nix, 1968). This approach does not require extensive a priori knowledge of functional realtionships but does depend on accurate selection of attributes used in identifying analogous areas. Nix (1968) has noted problems in obtaining reliable data and concluded that the resulting predictions are really location specific, and should not be considered widely applicable or generally valid. Site Factor or parametric methods are numerous and varied. 1 The bulk of practical applied research in the past decade has been directed at developing parametric techniques. This approach arrives at an assessment by relating critical land i See Riquier (1974) for a review of parametric methods in an international and third world context 25 characteristics to performance. The characteristics are then combined, usually with some form of linear regression equation, to produce a rating (Nix, 1968). The Storie Index for measuring crop production is an often cited example (Mabbut, 1968). Advantages of the method include its quantitative nature, and its amenability to further analysis. It can however be difficult to identify the critical relationships on which the method depends. Land characteristics also react differently in different settings and so, like the analogue method, it is often location specific. Recently, attempts have been made to generate performance ratings through systems analysis. These approaches divide a complex land system into component processes, and then synthesize the components into a symbolic representation, or a mathematical model, of the complete system (Nix, 1968). The resulting productivity models can then be used to predict response by manipulating input parameters to reflect the conditions of the land type being considered. Although considerable effort is being directed towards construction of performance models, the information or knowledge required for systems analysis is lacking in most areas. The above techniques incorporate biophysical information into the evaluation. The Framework also indicates the importance of socio-economic concerns but simply indicates the need for a general economic analysis with no explicit approaches suggested. Following a general economic and social analysis, land suitability classes are delineated to cummunicate the analytical and comparative results. Three class levels are identified: 1. Land Suitability Orders reflecting kinds of suitability. 2. Land Suitability Classes reflecting degrees of suitability within orders. 3. Land Suitability Subclasses reflecting kinds of limitation, or main kinds of improvement measures required within classes. These may be further divided into land suitability units. 26 The FAO framework is very similar to general land suitability analysis, but from a use specific perspective. The difference therefore, lies in the degree to which the requirements of specific LUT's are incorporated into the value measure. As a general observation, the approach seems most applicable in developing areas where specific choices must be made between cropping systems. Documented applications include projects in Brazil (Beek and Benema, 1974), Surinam (FAO, 1977) and Kenya (FAO, 1977). Plan Evaluation A different value measure from the previous approaches based on physical land characteristics can be derived by using goal achievement or plan evaluation. This perspective relates land value to the degree to which the land use attains identified goals or objectives (Fabos, 1978). Plan evaluation is not strictly a land evaluation tool and can be found in both public policy analysis and impact assessment studies (McAllister, 1980). Several of these techniques have however become significant in fulfilling the role of land evaluation in bridging the gap between single use performance rating and land use allocation policy. Planning Balance Sheets (PBS), as developed by Nathaniel Lichfield in Great Britian, is one such technique. PBS, essentially an extension of cost benefit analysis, attempts to determine the plan or option most advantageous to the public interest by examining transactions in relevant impact categories. These transactions are the costs and benefits involved in an option, identified in monetary terms wherever possible and related specifically to the sector of individuals affected (Poulton, 1983). Net benefits or costs delivered to each sector are tallied and alternatives ordered according to the score. The evaluator must eventually make subjective judgements as to the importance of sector categories to societal welfare. There are some advantages to PBS in that nonmonetary measures can be incorporated and the focus on transactions between sectors can supply useful 27 information on the equity affects of land use options. While this may be true, the nonmonetary impacts must still be compared directly to monetary values. The concept of transactions is also purely artificial and can serve to confuse evaluative information (Poulton, 1983). In the end, major value judgements are still required to complete the comparative evaluation (McAllister, 1980). A second approach, the Goals Achievement Matrix (GAM) also originated in Great Britian and was developed by Morris Hill (Hill, 1968) as a response to the limitations of cost benefit analysis (McAllister, 1980). With a GAM, the impacts or alternatives are assessed according to explicitly stated community goals. The approach is quantitative, but does not rely on conversion to monetary measures. As explained by McAllister (1980): Each impact is estimated by scientific methods when possible and measured in whatever unit most closely describes the goal to which it corresponds. A grand index of 'goals-achievement' can be calculated for each alternative action by multiplying the impacts by a set of value weights (p. 159). This grand index can then be used as a value measure for land use options. The ability of GAM to account for a wide variety of evaluative measures, and the organization of these measures around goal statements, have been considered important strengths of the approach (McAllister, 1980). Weaknesses include challenges to the simple arithmetic involved in the aggregation of the final index (Poulton, 1983) and problems in establishing value weights (McAllister, 1980; Poulton, 1983). These techniques and others like them (see McAllister, 1980 and Poulton, 1983) have been incorporated into tasks involving land evaluation. By way of an example, Fabos (1978; 1979) has developed a complicated procedure based on landscape assessment and plan evaluation. In the context of this system, called METLAND, landscape evaluation is viewed as a process used to evaluate alternative plans (Naveh and Lieberman, 1984). As explained by Fabos (1978), the model has three phases: The first phase - composite landscape assessment - produces landscape, ecological and public service value profiles. In the second phase, alternative plans are formulated. In the third phase, the effects of the alternative 28 plans on each value profile are evaluated (pp.4-5). The assessment procedure uses a parametric approach and calculates the profiles in dollar values. Ultimately, the system provides a balance sheet for land areas in dollars estimating the losses and gains to each value profile of land use changes. Although METLAND does offer some interesting insights into land evaluation techniques, it has remained essentially a research project and is too complex for widespread practical application. A slightly different approach to value measures through plan evaluation is provided by SIRO-PLAN, developed in Australia by the Commonwealth Scientific Industrial Research Organization (CSIRO). This technque is examined in detail in Chapter Four so will not be discussed in depth at this time. It calculates suitability ratings linked directly to policy statements. A plan of preferred use is produced by weighting the policies and maximizing a linear additive function. Subsequent plans are1 produced and compared on the basis of policy achievement, thereby moving incrementally towards the most satisfactory plan. The land value measure is thus related directly to policy achievement through a system similar to the Goals Achievement Matrix. Programming Techniques The Land Evaluation Group at the University of Guelph have promoted the concept of integral land evaluation through mathematical programming. The Land Evaluation Model (LEM) developed is also examined in detail in Chapter Four. In this case, it is argued that the relative value of different land use allocations cannot be determined without a measure of how the allocations affect the ability to meet societal goals expressed by acceptable production levels and requirements for goods, services, and amenities (Flaherty and Smit, 1982; Smit et al., 1984). The LEM accounts for societal goals by developing a model using mathematical programming techniques. Various goal scenarious can be tested with the model which produces four value 29 measures: 1. Feasibility of meeting socio-economic goals using the available bio-physical resource. 2. Overall flexibility in land use options. 3. Relative or critical importance of a land unit for a particular use. 4. Sensitivity of the first three measures to changing physical, social or economic conditions. Finally, land evaluation is not viewed as an attempt to define the best use but as an indicator to decision makers of the importance of particular land units in achieving societal goals (Smit et al., 1981). Other goal programming approaches to land evaluation have developed in recent years. Examples include a dynamic programming method designed by Hopkins (Hopkins, 1974; Goulter, Wenzel and Hopkins, 1983) and applied to watershed analysis as well as a goal programming model developed by Cocklin et al. (Cocklin et al., 1984) to assess poplar development in Eastern Ontario. Programming and modelling approaches can make very sophisticated use of functional information and can incorporate economic concerns in a parallel assessment. However, the data and knowledge requirements are currently prohibitive for practical applications on a wide scale. Hopefully as research continues and acceptance of these models builds among decision makers, they will be able to make a significant contribution to the analysis of complex policy questions. 2.3 SUMMARY This chapter has provided a very brief overview of both traditional types of land classification and rating as well as land evaluation. Land classification has been described as a means of converting land data into information. Land Evaluation is seen as providing a higher level of information by bridging the gap between traditional descriptive or single use classifications, and allocation policies. The amount of research effort being expended, and the variety of techniques available, is considerable and would seem to provide a solid basis for converting data into information useful in 30 land use decision making. However, regardless of the sophistication or elegance of the specific land evaluation system, the information must still be effectively integrated into the decision process if the conceptual model is to function properly. This requires an understanding of the decision making component, as outlined in the following chapter. Chapter 3 DECISION MAKING CONCEPTS For the conceptual model to function effectively, the information product must be integrated into the decision process. It is at this stage that the realities of decision making intrude on what was previously a largely technical or scientific exercise. It is also at this stage that the ultimate purpose of land evaluation - to aid and improve policy, management and allocative decisions - can no longer be ignored. Land evaluation should therefore be seen as a process involving strategic and routine, as well as public and individual decision making. As such, any land evaluation system contains some implicit perspectives on the decision-making process. Indeed, these perspectives are reflected in the differences between systems. Roome (1984) illustrates the combination of factors involved in these differences : ... the individual criteria that are combined in evaluation indices differ with respect to how they are measured, whether the measurement is relative or absolute, the degree to which value judgements are used in their specification and, indeed, with respect to the individuals or groups who define the value judgements and assumptions which underlie the evaluation (p.248). With this in mind, one of the basic contentions of this thesis is that explicit understanding of the value content and decision making perspectives and requirements as well as the value content is essential for effective design and use of a land evaluation system. This understanding begins with an awareness of pertinent decision making theories and concepts. Three areas are of particular concern: 1. The decision environment, including decision structure and descriptive models which provide an understanding of observed decision processes. 2. The appropriate techniques for decision analysis. These techniques provide an analytical structure, decision rules for choice, and measures of aggregate value. 3. The decision making component of land evaluation criteria. Decision making concerns have a contribution to make to the list of characteristics to be found in an effective land evaluation system. Each of these characteristics and their implications for land evaluation will now be 31 32 explained. 3.1 DECISION DIAGNOSIS Decision diagnosis is the act of deciding the nature of the decision process and environment by examination and analysis (Keen and Morton, 1978). Two analytical questions have important implications for land evaluation: 1. What type of decision environment is involved ? 2. How is the decision arrived at ? 3.1.1 DECISION ENVIRONMENT The difficulty and complexity of the inherent conflicts involved in land evaluation depend to a large degree on the decision environment Figure 5 provides a simplified view of four types of decision environments encountered in land evaluation. As the decision environment moves from sector A to D, the conflicts and dichotomies become more pronounced and difficult to resolve. This is primarily the result of the decision structure. Structuring a decision problem is an essential first step in addressing a decision (Cook and Hammond, 1978). Problem structure is determined by several factors including the degree of value conflict, and the amount of uncertainty involved. Models of decision making identify a continuum of decision types from well-structured to poorly-structured. A well-structured decision problem is one which is well enough understood to be characterized as routine and repetitive. All the information needed to formulate and solve the problem is available. Rittle (in Mason and Mitroff, 1981) used the expression 'tame' problems and listed a number of characteristics including those cited below: Tame problems can be exhaustively formulated and written down on a piece of paper. The solution to a tame problem can be tested. Either it is correct or incorrect Tame problems have closure - a clear solution and ending point 33 Decision Environment Natural Envi ronment Simple (few functions) Complex (mul t i pie functions) Considerable L i t t l e Control W I ^ M ^  Control (single agency) (Multiple agencies i nte rests) A farmer deciding where Rezoning an area of to plant wheat and homogeneous agricultural where to pasture dairy land on the urban catt le on a smal 1, f r i nge. homogeneous farminq unit. Routing of a hiking trai 1 in a local park. a c A forestry agency planning Creation of a Regional for mul t i pi e use in a Plan for a di verse source watershed. region dominated by private land. Regional recreation authority tryinq to provide access to a sensitive estuary. b d FIGURE 5 - Decision Environments (Adapted from Lang, 1978) Every tame problem has an identifiable, certain, natural form: there is no need to argue about the level of the problem. The proper level of generality can be found for bounding the problem and identifying its root cause. (Mason and Mitroff. 1981, pp.10-11) Accordingly, it is possible to specify decision rules or standard operating procedures to formulate the problem, evaluate alternatives, and choose an optimal solution (Keen and Morton, 1978). At the opposite end of the continuum are decision problems which are unstructured. For these problems there is little or no information on the best way to 34 develop a solution. Von Hesler (1984) characterized these as high order problems where the system interdependencies are incomprehensible. Mason and Mitroff (1981) argue that most of the problems facing decision makers today are ones of organized complexity and, using Rittle's typology, are considered 'wicked'. Wicked problems also have a number of identifying characteristics: The problem can be seen in quite different ways depending on circumstances ' and individual or group perspectives. There is no definitive or correct interpretation. There is no single criteria system or rule that determines whether a decision outcome or problem solution is correct Only relative results can be determined. Wicked problems involve dynamic and uncertain environments which creates a need to accept risk. Strong connections link problems creating important external and internal relationships, feedbacks, and costs and benefits. Change resulting from a decision may magnify, or disappear depending on these connections. There is no stopping rule - improvement is always possible. There is no exhaustive, enumerable list of permissible operations to be used in developing solutions. Because of competing claims, there is often a need to trade off 'goods' against 'bads' within the same value system. Social, organizational, and political constraints and capabilities, as well as technoglogical ones, are central both to the feasibility and the desirability of solutions. (Mason and Mitroff, 1981, pp. 10-13) As a result of these characteristics, clear-cut procedures are not available to the evaluator or decision maker. It is necessary to improvise, evolving custom made solutions and creative responses with much of the necessary information generated during the decision making process (Keen and Morton, 1978). In between these two extremes there lies a range of decision problems with varying degrees of structure. The mid-point can be characterized by semi-structured problems (Keen and Morton, 1978). There is enough information available to generally define the nature of a semi-structured decision problem, but uncertainty exists about the current state of affairs, the desired state of affairs, and how to achieve the desired state (Van Gundy, 1981). With semi-structured problems, there is sufficient structure for computer and analytical aids to be of value but intuition, value judgements and creative responses are still essential. The system interdependencies are I 35 diverse but can still be partially understood with careful assessment (von Hesler, 1984). The structure of land evaluation problems is a function of many factors including the number of competing uses and values, the quantity and quality of available information, and the nature of the decision environment as demonstrated in Figure 5. Decision structure may also vary according to the planning or information level as summarized in Figure 6. Lindblom (1959) declared that difficulties increase and competence decreases by an order of magnitude with every level of decision making ascended. Decision structure has important implications for evaluation and information management. Different types of decisions have different information processing requirements and different analytical needs (Bennet, 1983). Few land evaluation problems could be considered to lie at either extremes of this continuum. As shown in Figure 7, some strategic decisions could be very poorly structured and some operational tasks very well structured. Land characteristics affecting land evaluation are reasonably well understood at the operational level. At higher levels however, a true evaluation cannot be made without reference to other resources required to make a land use viable. External supply and demand factors and the competition for land also inject uncertainty and multidimensional concerns into any procedure. Most land evaluation tasks seem to occur somewhere between the two extremes in the realm of semi-structured problems. 3.1.2 DESCRTPTTTVE MODELS If land evaluation is to provide useful guidance for public decision makers, it must be sensitive to the public decisionmaking process. The concept of the decision process can largely predetermine the strategy chosen for design and implementation of any aid to improve the quality of decisions (Keen and Morton, 1984). Therein lies the benefit of descriptive models which provide insight into the decision process. Decision Level Information Required Planning Activity Strategic 4 External information from well beyond the temporal and spatial boundaries of the .problem Predictive, long term Simulated 'What i f of a long term nature National, provincial statistics Strategic or policy planni ng Emphasis on integrating physical and social information to set intermediate or long term goals Tactical Operati onal Descriptive historical i nformati on Current performance Short term future i nformati on Short term 'What i f simulations Regional, local stati sti cs Descriptive and historical i nformation Current performance Reporting on exceptions or violations of standards Regional D l a n n i n g and general land use plans Emphasis on translating policy into spatial plans Local site planning Emphasis on specific implementation and management functions FIGURE 6 - Decision Levels and Information Two types of descriptive models can be identified. Normative models describe the decision process in terms of how an c^timal alternative is, or should be chosen. Positive models attempt to describe how the realities of decision making actually intervene to circumvent normative models. 37 Well Structured Operational Type A Tactical Type B, C "Poorly Structured Strategic Type B,C Type refers to decision environment as i l l u s t r a t e d i n figure three F I G U R E 7 - The Structure of Land Evaluation Decision Problems The rational comprehensive view of decision making is a typical normative model. It describes the decision process as a logical sequence involving the ranking of objectives, listing and evaluation of choices, comparison of consequences, and selection of the choice with the highest net benefit (Richardson and Jordan, 1979; Hudson, 1979). In this model, the decision maker becomes an economic man, acting upon perfect information to determine optimal solutions. Many analysts contend that the rational comprehensive model provides a realistic description of only a few very simple decision processes (Lindblom, 1959; Keen and Morton, 1978; Richardson and Jordan, 1979). In reply, others argue that the standard definition is too restrictive. Simon (1976) developed a definition of bounded rationality, where the comprehensiveness is limited by considering only the most feasible alternatives. Friend (1983) claims that the distinction of a rational method is not 38 particularly helpful, citing a need to be both rationality seeking and sensitive to limitations in the decision process. In the context of this thesis, it is important to recognize the limitations of rational comprehensive analysis. These include: There is usually a lack of the comprehensive 'perfect' information the model ideally requires (Davis and Greenlaugh, 1980; Hall, 1982). The model assumes that actors will hold similar values, but there is rarely common agreement over value scales, measures, or objective functions. There is a lack of satisfactory techniques for evaluating goals and objectives, or for dealing with the vast number of different types of planning problems, not to mention problems of uncertainty (Davis and Greenlaugh, 1980) . There is usually a considerable cost involved in obtaining and analyzing truly comprehensive information. (Davis and Greenlagh, 1980). While these limitations do not necessarily invalidate attempts at rational analysis, they must be recognized and considered during information generation. A wide variety of positive models have been developed (see Allison, 1971; Keen and Morton, 1978; Smith and May, 1980). The identification and definition of individual models is not of particular interest in this thesis. Indeed, they are not actually separate entities and the decision process is not something which can be neatly partitioned and packaged into models representing 'correct' perspectives. As noted by Smith and May (1980), debates about the correctness of various models are largely artificial. Positive models do however introduce a number of observations and concepts which can be very significant in the generation and use of factual and value information for land resource planning. Several of these are noted below. Decision making often involves incremental movement towards satisfactory choices. Rather than identifying an optimal solution, the process examines alternatives that offer different marginal combinations of values to discover a choice which is at least acceptable after only a moderate search (Lindblom, 1959; Simon, 1976; Keen and Morton, 1978). Decision making often involves bargaining or negotiation. As alternatives are compared, bargaining occurs between individuals, government agencies, resource sector interests, and political factions in an effort to arrive at satisfactory policies. ) 39 Bargaining is a give and take process between two or more parties in an effort to arrive at a mutually agreeable combination of trade-offs (Dorcey, 1981). Decisions are often made within, and as a result of, organizational environments. In this environment, decision making rests on the heuristics and standard operating procedures operating within organizational units. Organizations exhibit several characteristics: a. Quasi-resolution of conflict - Rather than maximizing a single goal, organizations consider constraints with critical values. Trade-offs are rarely formally computed but standard operating procedures are used to detect and correct a violation of a critical value. b. Uncertainty avoidance - To avoid accounting for long run uncertainties, organizations deal with short run reaction and response in a negotiated environment c. Problemistic search - Searches prompted by violation of critical values are in the vicinity of both problem symptom and known alternatives. d. Organizational learning - Organizations accumulate experience in the form of positive and negative precedents. Decision making is often dominated by the political process. The process involves interactions between various actors, each with particular interests, priorities, constraints, and power. Advocacy, power, channels of communication, and the rules of the political game become critical (Allison, 1971). The importance of descriptive models lies not so much in the formulation of the models themselves, but in the challenge of adopting a diagnostic approach to ascertain which aspects of the models are most pivotal, or appropriate, in a given decision process (Keen and Morton, 1978). Hudson (1979) builds on this approach and explains that some models are complimentary, others generate fruitful tension, while others are fundamentally incompatible. They will be useful only if the analyst or decision maker can mix approaches or styles, and present them within the specific decision situations. This diagnostic approach begins with the acceptance of several conclusions: A number of factors combine to influence a decision process, making an objective, rational decision very difficult Particular decision processes vary in their use of the descriptive models. The effectiveness of a land evaluation system will depend to a degree on how compatible it is with the descriptive, components prevalent in the decision process. Even the most elegant system becomes useless i f it is incompatible with the 40 decision process. To date, these conclusions seem to be rarely reflected in land evaluation as techniques have seldom been developed with this variety of descriptive models in mind. Any number of implications of taking a diagnostic approach to land evaluation can be formulated. By way of illustration, four important examples are listed below. 1. An optimal answer provided by a rational comprehensive evaluation may be useful but should not be presented to the decision maker as a single infallible reply. An evaluator could present the result while indicating that it may be necessary to consider factors involved in the other descriptive models. To facilitate this, the analyst could supply a description of the components involved in the rational value, and an explanation of how various perspectives or factors in the decision process could affect the components and resulting value. 2. If the decision process is clearly dominated by satisficing, then an evaluation system must have a way to systematically vary and adjust components to reflect incremental value changes. 3. A land evaluation must account for the organizational context If an organization depends on a solidly entrenched review process based on standard operating procedures but the output from an evaluation system cannot be expressed in terms of the review process, the implementation of its recommendations will be difficult at best Evaluations will also be more credible if they can be expressed in- the language of a client agency, and use the resources available within the organizational structure. Ultimately, an evaluation system could strive promote integration within an organization. 4. Recognition of the political model leads to the importance of bargaining in public policy. In this context, land evaluation becomes critical in educating those involved in the bargaining process. Well informed participants is one of the keys to effective bargaining (Fisher and Ury, 1981). Land evaluation techniques could therefore be designed with bargaining in mind, at least as a potential use, if not an explicit component Other implications can certainly be suggested. The important point is the adoption of the diagnostic perspective as represented in the examples given. 3.2 DECISION ANALYSIS The techniques used to derive the measures of value described in Chapter two are not just scientific abstractions. They are expressions of analytical comparison reflecting different approaches to the comparisons, trade-offs, and choices required at the core of a land use decision. 41 Once the decision environment has been explored, the insights gained can be combined with available resources in the selection of appropriate techniques. From a decision making perspective, these techniques form the basis for decision analysis. Decision analysis attempts to formalize and standardize the decision-making process using methods which quantify decision evaluation and outcomes, such as ranking, rating and weighting (Bakus et al., 1982). 3.2.1 TECHNIQUES This section will not include an exhaustive review of decision analysis techniques. Rather, Table 1 serves as a summary listing of those techniques of particular significance to land evaluation. By referring to the table, the options available can be quickly assessed. Decision analysis for land evaluation obviously involves multiple objectives and alternatives. It therefore embraces those techniques designed to handle multiple criteria. Two general techniques are significant in land evaluation. The first includes those methods traditionally associated with multiple objective decision analysis. They provide an ordering of alternatives based on their value as expressed in the multiple outputs and degree of objective attainment Most of these methods require a clear statement of objectives and information on the impact of each alternative on each objective (Sinden and Worrell, 1979). Multiple objective techniques can involve optimization of one objective, conversion of all objectives into a common unit, or weighting of objectives to allow addition on a common utility scale. Some of these techniques are noncompensatory in that one attribute or criteria cannot compensate for another. In compensatory techniques, advantages and disadvantages are integrated into a measure reflecting all the criteria considered (Montgomery, 1983). Four categories of these techniques are shown in Table One. 42 Table 1 Decision Analysis Techniques C la s s i f i ca t i on Methods Characterist ics Application Data needs Mechanics Results Sequential El imination - F lex ib le - Oomi nance Alternatives compared. Choose Alt.A i f i t is better then B on at least one c r i t e r i a and no worse on the others comnarati ve ranking of paired alternatives comparing crops for plantino in a single f i e l d data requi rements - Lexicographic Alternatives are compared based on an ordered set of c r i t e r i a or object ives, start ing with the most important c r i t e r i a . chooses between pai red al ternati ve given the c r i t e r i a wei ghts comparing crops for use on a hobby farm where ease of maintenance is the primary concern. Hard Hu l t i c r i ter ia - requires re l i ab le metric in fo . - begins with an impact matri x with outcomes compared to c r i t e r i a - Concordance Analysi s Data i s normalized and importance of attributes are judged in a weighting system. Must desirable outcome for each c r i t e r i a is stated Rairwise comparison of alternatives and calculat ion of concordance or discordance sets based on distance from desired outcome Selects a best alternative and indicates a preference measure for al1 plans. Appropriate as an elimination or screeni ng procedure Part icu lar ly useful when d i s t inct a l ternat i ves are avai lab le, as opposed to continuous decisions Evaluation of alternative land reclamation projects (Nijkamp, 1980) - Goals Achievement Relates objectives to achievement 1 e ve 1 s. Evaluate costs and benefits of a t t r ibutes , assiqn a re lat ive weiqht to each at t r ibute. Calculate the outcome of a lternat ive actions based on the costs and benefits of attr ibutes. Determine totals for each a l t . Aggregate Achievement index for each al t. Need to be able to express objectives and define units of measurement Evaluation of transportation plans (McAl l ister, 1980) Soft Mul t i c r i ter ia - based on ordinal or qual i tat i ve information - Ordinal Concordance Similar to concordance analysis but uses ordinal info. Begins with pairwise comparison. Concordance index calculated as an aaqregate preference score for those c r i t e r i a where an alternative has better outcomes than a l l other alternatives Does not supply a single best a l t . Eliminates unacceptable a l t s . Part ia l orderinq of acceptable al ternati ves - Mult i -Dimensional Sealing i Ordinal data converted to metric data such that the result ing configuration hears an accentable resemblance to the ordinal data set (geometric representation) Value is judged to be inversely proportional to the distance of an a l t . from the preferred outcome (involves a complicated algorithm) Ranks a l l alternatives based on their aggregate performance Used where quantitative info, is not avai lable. Urban renewal nlans as they relate to qual ity of l i f e . (Nijkamp, 1980) 43 Hard Multiobjective - based on metric Info, on continuous objective functions and constraints - Penalty point - Constraint Optimization Establ ish an Ideal achievement vector. Discrepancy between actual values and ideal values is penalized by means of a D e n a l t y function. One objective is chosen for optimization while other objectives are expressed as constraints. Linear programming calculates optimal values. Measure of di fference from an ideal vector. Produces an optimal package and a quant i tat ive index of comparati ve value. Often used in Environmental -Economic models Strategic land evaluation for agriculture (Smit et a l . , 1981). - Goal Programming Achievement level is spec i f ied for a l l decision c r i t e r i a . An alternative 1s generated which minimizes the weighted sum o f deviations from the decision c r i t e r i a achievement levels. Produces an optimal package and a quantitat ive index of comparative va 1 ue. Envi ronmental -economic decisions where achievement levels can be speci f ied. Assessment of land a l locat ion for poplar harvestinq (Cocklin et a l . , 1981) Standsrdi zation by ranked classes - F lexib le data requirements - Ordinal Combi nati on A measurement of a character i s t ic is a l located to one of a number of ranked classes for a p a r t i a l , not complete or continuous standardi zation. Ranked classes are then compared. Relati ve ordering of al temat i ves. Overlay analysis to locate areas suitable for res ident ia l development. Standardization by uti 1 i ty rat i ngs - Can use a variety of data but requires some functional knowledge - Direct addi t i ve method Each character ist ic is measured. Relationship between character i s t ic and u t i l i t y scale is postulated or establ ished. Characterist ic converted to uti1i ty. Weights obtained for u t i l i t i e s . Sum nf weiqhted u t i l i t i e s generated. Agqreqate score for each al temat i ve. Relative orderina of choices. Used in landscape assessment. Useful for evaluation between choices where character i s t ics are imnortant and indicat ive of value. Selection of prospective campground locations (Sinden and Worrel 1, 1979). 44 The First category shown is sequential elimination methods. These methods involve the use of decision rules to choose between different alternatives. They can be used with quantitative or qualitative information in ordinal or ratio form (Montgomery, 1983). While sequential elimination methods may be used on their own as choice techniques, they may also provide a fundamental rule used in the more elaborate techniques. When used in isolation, sequential elimination methods are noncompensatory. Hard multicriteria models require reliable metric information and begin with the construction of an impact matrix (Nijkamp,1980). This matrix reflects the outcomes of alternatives with respect to the relevant decision criteria (Bakus et al., 1982; Nijkamp, 1980). The techniques assign weights to decision criteria and then use slightly different procedures to generate results. For example, the concordance method makes a pairwise comparison of alternatives and develops concordance and discordance sets. The alternative that is most dominant or concordant is given the highest value. Hard multiobjective optimization models use continuous objective functions and constraints. They require quantitative, ratio information (Nijkamp, 1980). A variety of measures can be used in combination with the optimizing procedure. The penalty points model assigns penalty points based on the discrepancy between an actual value and a predefined ideal value. The alternative with the fewest penalty points is considered best On the other hand, constraint models choose one objective to be maximized and introduce other objectives as constraints. Soft multicriteria methods can be based on ordinal or qualitative information (Nijkamp, 1980). Some soft multicriteria models are simply hard multicriteria models modified and simplified to accomodate qualitative data. Others, like the multidimensional scaling method, attempt to transform ordinal data into metric form. This allows direct comparison or evaluation of alternatives based on aggregate performance. The second group of techniques concentrate more on the characteristics important in obtaining objectives, rather than judging the utility of various outcomes. 45 These methods begin with the assertion that the utility of an alternative is a function of its characteristics. Agreement is therefore obtained on the important characteristics and means of measurement The result is a comparative value, not an absolute measure, and does not provide monetary information (Sinden and Worrell, 1979). Two types of approaches are summarized in Table One. Decision analysis techniques are important in that they explicitly identify the assumptions and decision mechanics behind many operations found in land evaluation. The typology also offers a convenient means of classifying evaluative approaches for easy review and selective search. Selection of a technique depends on the resources available and measures required. If little information is at hand and if only a relative qualitative measure is required, then soft multicriteria methods may be satisfactory. Conversely, if extensive ratio information is available and an absolute measure is required, hard multiobjective approaches may hold the most promise. Sinden and Worrell (1979) continually emphasize the importance of matching available data, measure required, and technique chosen. Figure 8 combines several principles and strategies they have developed into a selection process diagram. To conclude, the challenge is to make the most of available resources, choose a technique which will provide the best measure possible, and use that technique within the confines of its limitations. 3.3 DECISION CRITERIA When assessing the potential effectiveness of an evaluation system, three types of criteria are important The first is scientific accuracy which is largely determined by the technical proficiency of a system. The second is cost-effectiveness which measures the capability against the cost, a very important consideration when resources are limited. The final type is decision making utility or the usefulness of the product within the decision process. These criteria are not mutually exclusive. For example, if 46 Decision Problem Are there multiple c r i t e r i a and variable costs and NO , benefi ts? 1 YES Other Techniques (rarely the case with land evaluation) Objectives can be measured Is a quantitative measure requi red? Characteristics can be measured NO Is a quantitative measure requi red? 1 YES NO YES Is metric data available? NO Is metric data available? YES NO Sequenti al E1imi nation Ordi nal Concord. t Sequenti al EUm. Concord. Analysis Is metric data available? YES NO T Multi Dimensional Scaling Is metric data available? YES NO Ordi nal Combi nation T Di rect Addi t i ve YES Di rect Addi t i ve Hard multiobjective methods FIGURE 8 - Decision Analysis Technique Selection Process 47 a system is not cost effective, it is not likely to be used in a practical decision process. Scientific accuracy, while by no means undeserving of further attention, is covered in the existing literature (see Dorney, 1976; Hills, 1976; FAO, 1977). Cost effectiveness is difficult to measure, although post facto studies have . been done on some specific projects (Dorney, 1976). Cost-effectiveness measures do need development but this development is not meant to be undertaken within the confines of this thesis. While scientific accuracy and cost effectiveness are not explicitly discussed, they are not ignored and will be noted where necessary. It is the third consideration, decision making utility, which will now become the focus. Decision criteria can grow to an unmanageable size if an attempt is made to include all those characteristics noted as important in the literature. At this stage therefore, an effort was made to identify five general principles which were considered to be of primary importance, and representative of the sorts of concerns that are usually ignored or considered secondary in the traditional planning and assessment of land evaluation systems. These principles also reflect the need to incorporate come of the conclusions arising from the positive models of decision making. The principles chosen are detailed below. 1. The decision maker should have the ability to explore the problem. As argued by Lock (1983): Ultimately, the analysis has to be useful to decisionmakers and they have to feel confident about it. The presentation of a single best option does not always inspire this confidence. Strict optimisation is less attractive than the ability to explore the problem. A major role for computer based problems is the facilitation of manager-based sensitivity analyses and the ability to respond to 'What If questions. Decision makers acquire commitment to the solution by feeling both that they have some control over the solution and that they have contributed to its development (p.160). Two types of exploration are envisioned. The decisionmaker or user would benefit from the ability to explore the problem itself, its dimensions and the 48 interelationships between its elements, and the range of possible solutions. Evans et al. (1980) identified this desire in a study of the planning process undertaken in Gray's Harbor in Washington State. A decisionmaker may also need to explore the impact of various alternatives or parameter values. This involves the 'What If questions referred to by Lock (1983). What if questions are important for both factual and value premises. The uncertainty surrounding some factual content, especially long term strategic predictions, requires some exploration. With respect to value content, the political and incremental nature of some decision processes necessitates exploration. This is particularly important where values may change quickly with potentially significant impacts on the subject of an evaluation exercise. 2. Flexibility is important if an evaluation system is to be an effective decision aid for several reasons: a. It facilitates problem exploration. The 'What If questions can only be asked if the system is flexible enough to allow for systematic manipulation of inputs and parameters. b. Users can rarely specify functional requirements in advance in poorly structured decision environments, nor can they always articulate what they want or need. Flexibility is required to allow some reaction-correction or iterative design. c. Variables in the decision environment change over time. This includes changes in societal values, inputs required for land utilization types, and in the information available or actors involved in the decision. Flexibility is required to accomodate such changes without invalidating the complete evaluation. These observations lead to the need for two types of flexibility (Sprague and Carlson, 1982). The first level of flexibility provides the user with the means for creative problem solving. A particular decision can be explored and approached from several perspectives by varying the components or operations within the evaluation system. The second level involves the ability of the system to evolve or accomodate new demands from users or changes in the decision environment 49 3. Land evaluation systems also provide the opportunity for information management, a critical component of improved decision making (Litton and Kieiger, 1971). They can contribute to information management in several ways: a. By organizing and providing quick and easy access to the volumes of existing information relevant to land use decisions. b. By providing a means to assess existing information so a decision maker can determine the value and implications of information in the evaluation process. c. By providing assistance in identifying information gaps. d. By providing direction and possibly an organizational framework for generating new information, descriptive and functional. e. By providing some common ground for discussion, communication and exchange of information as disparate information sources are synthesized by inclusion in the evaluation system. 4. A land evaluation system must be credible. If it is not, and it lets a user down, it is not likely to be influential in the decision process again (Meyers, Kennedy and Sampson, 1979). Credibility is gained largely from the accuracy provided by the technical and scientific standards incorporated into the evaluation. Credibility is also affected by how the evaluation is used in the decision process. Problems often develop simply because the user or decision maker may push a system beyond its limits and assume a level of accuracy which was never intended (Poulton, 1983). On the other hand, if the user is fully versed in the assumptions and limitations inherent in the evaluation, credibility is more likely to be maintained. Assumptions and limitations which should be explicit include the extent to which the evaluation is based on qualitative value judgements, the scale of measurement in terms of ordinal or metric data, the quality of the input data, and the division between the factual and value content The information generation process is also important in that the inputs and procedures used must be acceptable to the user. If the system uses input or a model which is viewed suspiciously by those involved in decision making, then credibility problems may surface. 50 5. Finally, a land evaluation must be understandable or intuitively meaningful to the user. If the results are obscured in technical language, then effective use will be difficult. If, however, it makes sense and its relevance is obvious, the evaluation should receive some use and consideration. 3.4 S U M M A R Y This chapter has provided information which should play a role in the selection and design of land evaluation techniques. It has described the important contents of the decision box from the conceptual model. Concepts presented include decision diagnosis, decision analysis, and decision criteria for design. An understanding of these concepts facilitates the critical linkage between the conversion or information generation and decision components of the conceptual model. If these concepts can be related to land evaluation techniques, then the strengths, weaknesses and limitations of those techniques in the context of the decision process can be identified. This is the task confronted in the following chapter. Chapter 4 ANALYTICAL COMPARISON AND ASSESSMENT OF CURRENT LAND EVALUATION SYSTEMS For the model presented in Figure 1 to function effectively and efficiently, the linkage between the conversion and decision boxes is critical. Having described the internal contents of these boxes in the previous two chapters, attention can now be focussed on the linkage. In this chapter, current approaches to land evaluation are examined to provide a more detailed view of specific approaches and to detail practical illustrations of the linkage. The following analysis is not designed to grade the systems, nor is it primarily a comparative evaluation. Attempts to directly compare approaches have been made (see Lee, 1982) but this is difficult, and in some ways counterproductive, because different systems may have been designed for different purposes, at different scales and in different technical and social contexts. Instead, this analysis concentrates on identifying the structure and strengths and weaknesses of the systems using the theoretical basis established in the previous chapter. This is accomplished by first describing each approach, and then examining its decision making component, including strengths and weaknesses. A comparative summary is provided for both the technical and decision making analysis. The three methods chosen are the ecological planning approach, SIRO-PLAN, and the Land Evaluation Model (University of Guelph). 4.1 TECHNICAL COMPARISON 4.1.1 ECOLOGICAL LAND EVALUATION Ecological land evaluation was popularized by Ian McHarg through his seminal work Design with Nature (McHarg. 1969), and is often referred to as the McHarg 51 52 approach or University of Pennsylvania method to reflect the contribution of that faculty. Being largely based on physical suitability, and qualitative in nature, it could be considered a type of land resource suitability analysis (Steiner, 1983). The methods trademark is the use of map overlays, although the simple use of overlays does not necessarily constitute ecological land evaluation. At the core of the approach is an ecological perspective. It considers the environment as a reflection of its interacting components, with the fittest environment being one in which the bulk of the work required for system survival is completed by the natural environment (Giliomee, 1977). McHarg (1969) explains that: The basic proposition employed is that any place is the sum of historical, physical and biological processes, that these are dynamic, that they constitute social values, that each area has an intrinsic suitability for certain land uses and finally, that certain areas lend themselves to multiple coexisting land uses (p.25). From this proposition comes an attempt to use an ecological profile to determine intrinsic land use suitability. The unitary character of interacting natural processes is reflected in the planning process as a value system. By incorporating these values into a single accounting system, appropriate land uses can be designated (McHarg, 1969). Technically the approach consists of seven steps which are illustrated in Figure 9 and described below. 1. Ecological Inventory This is an interdisciplinary collection process, using existing sources as much as possible and involving search, field checking, and mapping (Steiner, 1981). Physical factors inventoried may include physiography, soils, hydrology, vegetation, climate, and resource values. An inventory of sociocultural factors, like settlement patterns and ethnology, is accomplished by participant observation (Steiner, 1981). 2. Area Analysis The study area is then examined to define structure and function. This is usually accomplished by constructing a two dimensional array which illustrates bivariate relationships between landscape elements, like vegetation and wildlife (Steiner, 1981). 3. Use Analysis 53 S T E P 1 MAP DATA FACTORS BY TYPE f l f t m p l * 1 f i » m p l * t A - e - t o * • - 10 - to% A - tUOKTVV E A O O O • - ftJQKT T O M O O C J U T t f MOOCAATC D - CXTHCMQ.V BROOCD C R O S O N MAT S T E P 2-4 RATE EACH TYPE OF EACH FACTOR FOR EACH LAND USE •»c lcr -lump;* 1 A 1 i • I i C a a * I i • i * c 3 t 0 a 3 1 - PfttJt &JTAB4JTY J - S£-t>C;AAT * - TtRTUflT S T E P 5 MAP RATINGS FOR EACH AND USE ONE SET OF MAPS FOR EACH LAND USE Kapmplp 1 faampla j f iamplt 1 Etamplt a M 0 U 3 M 0 S T E P 6 OVERLAY SINGLE FACTOR SUITABILITY MAPS TO OBTAIN COMPOSITES. ONE MAP FOR EACH LAND USE 1 ^ towt sT ML»«ej»s w c B E S T sumo _. Few u u e uac O V E R L A Y S H M H E S T NUUBER9 A R E L E A S T S U T T E O FOR L A N D U S E A O M C U L T U R E H O U 0 M 0 S Y N T H E S I S FIGURE 9 - Ecological Land Evaluation (Adapted from Steiner, 1981) 54 This involves the identification of a potential set of land uses and the optimal or required natural conditions for the uses. Once again, a matrix can be used with land uses being compared to land use requirements (Steiner, 1981). 4. Needs Analysis Land use needs are then related to natural factors with yet another matrix. This is used to predict consequences of alternative human actions (Giliomee, 1977). 5. Mapping Specific mapped phenomena are now related to land use with the help of the matrices. Constraints and opportunities begin to emerge. 6. Synthesis Opportunities and constraints are then mapped. Rules of combination are formulated to weight the relative importance of mapped phenomena (Steiner, 1983). This is the stage where values begin to play an explicit role (Giliomee, 1977). Derived maps of opportunities and constraints are then overlayed to create a composite suitability map for each land use. These maps show a continuum from areas of greatest opportunity and least constraint to the reverse. (Steiner, 1981; Johnson,Berger, and McHarg, 1979). 7. Alternatives The various suitability maps can then assessed. Overlays and comparisons can be made as dictated by the objectives of the exercise. Needs and desires of users, as well as legal and economic considerations can be injected at this stage. (Johnson, Berger, and McHarg, 1979). The steps described can vary from application to application, but the basic philosophy and operations remain the same. 4.1.2 SIRO-PLAN SIRO-PLAN is actually a complete land use planning methodology but does include a distinctive land evaluation component which will be the focus of this discussion. The method begins with the recognition of the 'policy-balancing' problem in land allocation (Cocks and Austin, 1979). The often cited planning sequence of goals -measurable objectives - control actions seems rational and is often cited as the ideal process. However, when all of a community's goals cannot be met, as is often the case, the process breaks down and becomes more and more political. 55 SIRO-PLAN attempts to work with this problem, providing an indication of optimal land use options while also reducing the intuitive or political component to a series of small explicit decisions on policy sets. Cocks and Austin (1979) consider it a policy oriented approach and explain that: ... it does not start with a preconceived concept of 'good' land use such as McHarg's ecological fundamentalism or some form of economic or environmental determinism. Rather the planner's guidelines must emerge case by case from a political or politically ratified process (p.3). The technical component of SIRO-PLAN has evolved from a plurality of sources including optimization techniques, multiobjective planning, ecological planning, and comparative evaluation literature (Cocks et al., 1983). Two measures of land value are involved. A policy achievement rating for a single use in a particular area in relation to a given policy, and an aggregate measure of suitability between uses in a particular area given a set of policy weights (Ive and Davis, 1985). A computerized tool, LUPLAN, has also been developed to facilitate the method and will be considered part of the procedure in this discussion. The land evaluation component of SIRO-PLAN can be summarized in five steps. 1. Formulation of Land Use Policies Land use options are identified along with explicit management interpretations (McDonald and Brown, 1984). The management interpretations are expressed as policies in such a way that the degree to which any plan (set of controls) satisifies any policy can be measured as a value for a policy achievement indicator (Cocks and Austin, 1979) For example, potential policies could be: a. Give preference to agriculture in areas of high agricultural activity . b. Exclude residential development from flood prone areas. Policies can be identified through working with client or interest groups or existing government plans and policy statements (Cocks and Austin, 1979; Davis and Greenlagh, 1980) 2. Definition of Land Units and Data Collection Spatial planning units ,are created. These units are relatively homogeneous with respect to the attributes or qualities central to the policies defined in step one. The data required to generate the policy achievement indicators is then collected for each unit 3. Calculate Policy Achievement For each policy, it is necessary to rate each land use option for each planning 56 zone. An area of class one agricultural land , for example, would have a high rating for agriculture given the preference policy demonstrated in step one. The ratings depend on two variables. The first is the state of scientific knowledge. Ratings are calculated using established techniques or models which relate land use to land qualities. These vary from subjective, ordinal estimates of importance to complicated models linking land type with crop yield. If there is a well tested, highly developed technique to incorporate functional knowledge, then the rating will be improved. The second variable is the accuracy and completeness of input data (McDonald and Brown, 1984). LUPLAN allows the programming of basic code, which can express most rating models, to calculate policy achievement ratings. 4. Policy Weighting Once policy achievement ratings are calculated, the policies are weighted. As with the definition of policies, there is no set procedure for policy weighting. Weights can be changed to reflect values of various participants in the process, or can be varied in an incremental search for a desired land use pattern (Cocks and Austin, 1979; McDonald and Brown, 1984). 5. Generation of Output The policy weights and policy achievement ratings are combined to produce an aggregate score for each land use in each land area. The preferred use in each area is the one with the highest overall policy achievement This index is calculated by summing the weighted total of the policy satisfaction ratings for land use options. LUPLAN produces this index through a form of linear additive programming. In addition to the plan and aggregate suitability scores, LUPLAN also produces an indication of the degree, expressed as a percentage, that each policy has been achieved. Value weights can then be varied incrementally to move towards a plan with the desired levels of policy achievement After the exercise is complete, LUPLAN produces several outputs including the aggregate suitability ratings, a map of preferred uses, and a summary of policy achievement percentages. 4.1.3 I A N D EVALUATION MODEL Chapter two provided an introduction to the rationale behind the Land Evaluation Model (LEM) developed at the University of Guelph. To reiterate, the approach is based on the concept of integral land evaluation and the assertion that land value must be related to societal needs and demands. The system therefore, does not produce an optimal intrinsic value but attempts to indicate the importance of land areas in the achievement of various demand scenarios, or the affect of physical 57 changes on the ability to meet demands. Currently, the model is limited to agricultural land evaluation although it is felt that the general approach is widely applicable for a variety of uses (Smit, pers comm.). The Land Evaluation Model 2 is a large mathematical programming model which synthesizes a diverse range of information. Information required includes: Inventory of the land resource base. Estimates of the availability of other resources like energy. Indications of alternative land uses. Estimates of productivity of various land use alternatives. Estimates of demands for goods and services. Indications of other constraints on land use. An allocation model is created by expressing information and objectives as constraints which will be satisfied by feasible allocations of land uses to land areas (Smit et al., 1981). From the feasible solutions, a particular allocation is chosen on the basis of maximizing or minimizing an objective function. Objective functions developed are listed below: Minimize the total area of land allocated to uses across the province (Min A) Minimize the summed proportions of each land unit allocated to each use (Min P) Minimize the sum of squared areas of each land unit allocated to each use (Min ASQ) Minimize the sum of squared proportions of each land unit allocated to each use (Min PSQ) (Smit et al., 1981) Each function provides a different measure. For example, Min A provides a measure of overall flexibility in land use subject to the specified constraints, and Min PSQ measures the flexibility of use in particular land units (Smit et al., 1981). The evaluation is completed in six general steps. 1. Creation of Land Types The land base is divided into land units on the basis of (1) climate as The LEM is continually being revised and new applications are being developed. The version decribed here is the second prototype, or LEM 2. Although some specific components may have changed recently, the basic model and analytical operations remain the same. 58 measured in corn heat units (7 zones), (2) land type based on CLI subclass (7 types) and, (3) regional location (6 regions) (LEG, 1983). 2. Identification of Land Uses Eighteen agricultural land utilization types are considered explicitly, with urban and forestry uses incorporated implicitly through a reduction in the available land base (LEG, 1983). 3. Generation of Constraints Data is collected to provide the land base information and the input-output production coefficients which make up the constraints. The constraints specify the conditions under which the evaluation will be completed. The constraint sets can include information on the available productive land base, limitations imposed by input requirements (energy, fertilizer), and other important environmental or economic factors. 4. Scenario Specification A particular set of conditions and demands is selected. Assumptions related to resource availability, societal demands and the associated land use conditions can be varied. An objective function is also selected. 5. Programming The objective function is programmed using both linear and quadratic functions. 6. Output Three types of output can be obtained. The first is simple data on such topics as land availability, yields, and input requirements. The second type of output involves the result of the programming procedure which varys depending on the objective function, constraint set and scenario. This measure will give an indication of the degree to which societal demands can be met given the constraint set The third type involves specialized indices or evaluation measures to indicate strategic importance and flexibility. These vary from 0 to 1 and as the index moves towards one, the system is exhibiting less flexibility. In other words, that particular land use on that specific land type becomes critical if demands are to be met These measures along with a flow chart for the model are included as figures 10 and 11. LEM is an interactive, user oriented system although this use really needs to occur through an operator familiar with the system. Actual use of the evaluation model occurs through five components, each using its own program or software package (Smit et al., 1981). Separate packages have been created for data management, scenario specification, solution algorithms, evaluation measures, and map production. Inhrrmation un the land use system obtained ituni land tesuuice nifnt.sts. crop and livestock specialists, agricultural economists, farmers <nd others Inlurmalinn im in* broad-scale, lung-term obietlives lor the use o( land obtained from pnlicy ind resource analysis, land use planners, economists, demographers, forecasters, modellers ind others I Specifications of lh« QQfjV-*J L A N n EVALUATION MODEL F0* 0N1ARI0 v - ' s ^ v ^ (LEM 2) T Assessment o( available information on land and Hi productivity Itx alternative uses, and development of agricultural production models where necessary I Review ul commodity, and resource consumption pattern;, and protecting of future demand levels, availability of resources and environmental restrictions on land use Measuring and estimating input requirements and product yields in each land area 1 Yields and Input// Linking livestock to land resource via their feed requirements 1 Four data files o n / livestock production Estimating lotal provincial demand and imports for agricultural and forest products Estimating land requirements for urban enpansion 1 Estimating avail-ability of land and energy resources Land Availability/ SCENARIO SPECIFICATION The set of conditions and requirements that the use of land must satisfy TT Specilymg restric-tions on Ine use of land Estimating land requirements for land uses not included in LEM 2 EXECUTING LEM 2 All conditions and requirements that the use of land must satisfy are considered simultaneously — » I I Determining annual productivity lor t ti* ri denned region Determining mininul provincial Und retirements Determ;r.:n» fleiibility in (he use of I ind E l CO Maiinium annual \ \ ^ production lor a use or gmup uf uses and assiiciated energy \ requirements \ ^ \ \ \ 3 Land-ell'oeni land use system and.\^ associated energyvS requirements X V \ ! i LE(£NO LEM 2 Data Output L E M : Model tlutput Internal 11• Land Evaluation Project Eneinal to Land Evaluation Protect Possible leedbaiI or connection Sensitivity to "vsX changes in resource availabilty and \ V production \ \ N N \ requirements \ \ \ ^ Land and energyv required to satisfy all specified \ N V conditions a n d N \ requirements Ovo SensitivitytoTvvS changes in resource availability and \ V production\\X^\ requirements^^ Evaluation Measures Strategic importance of particular areas \ for specific uses \ \ „ | planners investigate^ . | impact of continued | | trends or future f*~ — — — I changes on land I I Other considerations j | in (he lormulatiun L,. I^ ui land use policies I i f Local planning "\_ ^ \ Other Inputs into ^_ ^concerns and issues | j land use planning^ I I I Evaluating the land ) I resource for j I alternative uses | I V.) j Formulation ol | | broad-scale, long-1 | term land use j I policies and • , guidelines . I | LAND USE J | PLANNING I J F I G U R E 10 - Land Evaluation Model Structure (from Smit et aL, 1981) EVALUATION MEASURES FOR HYPOTHETICAL SCENARIO A Climate Zone 7 Urban Proximity Zone 17 Uses L a n d Types 1 2 3 4 5 6 7 8 9 10 11 12 13 Tota l 1 Hay .02 .10 .02 .13 .08 2 Fodder C o r n .26 .25 .27 . .08 3 Improved Pasture .31 .23 .01 4 Unimproved Pasture 5 Gra in C o r n .13 .23 . .03 6 Oats 7 Barley 8 Soybeans 9 White Beans 10 Winter Wheat .12 .24 .31 .16 .39 .13 11 Apples 12 Peaches .60 .60 .06 13 Grapes .21 .25 .21 .25 .24 14 Peas 15 Sweet C o r n .15 .12 .14 .09 16 Tomatoes .09 .16 .05 17 Potatoes 18 Tobacco Agricultural Uses .83 .71 .52 .83 .57 .92 .40 . .53 ~M H .77 19 Forestry 20 Urban .09 .06 .09 .29 .32 .28 .06 Non-Agricul tura l Uses .09 .06 .09 .29 .32 .38 .06 A l l Uses .80 .58 .83 .66 .92 .40 . .82 .63 .61 .83 L a n d Available 2.0 12^0 0.4 0.3 6.2 0.1 0 0 0 0 0.7 0.1 0.1 22.3 (ooo's ha) The closer the index is to one, the more important that use on that land type is to the achievment of the scenario ob.iecti ves. FIGURE 11 - Land Evaluation Model Evaluative Output (from Smit et al, 1981) 61 4.1.4 SUMMARY Table 2 provides a summary of the technical components of the three systems according to data management, evaluative operations, and output 4.2 ANALYTICAL COMPARISON Each of the three approaches will now be examined in relation to the concepts introduced in the previous chapter. 4.2.1 ECOLOGICAL LAND EVALUATION . Decision Environment Ecological Land Evaluation has been used in a variety of decision environments. It has been applied to areas of a few hectares and a few hundred square kilometers as well as urban, suburban, and rural environments (Johnson, Berger, and McHarg, 1979). In Design with Nature . McHarg documents case studies ranging from highway route selection to a general land use plan for Staten Island. Subsequent documented applications include a plan for Woodlands, a new city outside of Houston (Johson, Berger, and McHarg, 1979), a rural housing study for Whitman County, Washington (Steiner, 1981), regional planning in Waterloo, Ontario (Dorney and Hoffman, 1979), a plan for the central waterfront in Toronto (Lee, 1982), and residential' guidelines for Medford in New Jersey (Palmer, 1981). In all of these exercises, ecological land evaluation generated critical information. In examining the methodology and. its application, it seems to be best suited for well-structured decision environments. This conclusion is reached largely by examining the functional information required and the determination of values. The method depends on the identification of bivariate relationships but provides no concepts or techniques to develop this understanding. It includes techniques for analysis of spatial relationships but relies on judgemental determination of bivariate Table 2 Technical Summary of Current Approaches DATA VALUATION OUTPUT McHarg Method - extensive d e s c r i p t i v e data with unsupervised collection - land units created during eval uation - existing sources combined with expert .judgement - depends heavily on expert opi nion - parametric and landscape components - value generated through ordinal combination - bas ic values determined o r i o r to evaluation - matrices showing bivariate relationships - maps showing land qual it ies su i tab i l i t ies for individual uses, and overlays showing composite c a p a b i l i t i e s SIRO-PLAU - very focussed, relatively low data requirements - supervised collection - land units created prior to evaluation - depends heavily on the state of sc ient i f i c knowledae and accuracy i f input data - parametric approach - aggregate values generated with linear additive combination - values can be varied during eval uation - preferred land use plan - ratings for each land use in each area - percentage o f achievement for each p o l i c y LEM - both descriptive and functional requirements' - supervised collection - land units created Drior to eval uation - existing sources plus special research - uses linear programming and mathematical modelling - values inDuted during evaluation in the form of scenarios - maps and l i s t s showing input data and land values - tables showing evaluative measures ( f l e x i b i l i t y , cri t i ca l i ty ) 63 relationships to inject a functional component (Lee, 1982). In a simple system where these relationships are obvious and unlikely to be challenged, this may not be a problem. However, in a complex system characteristic of poorly structured decision environments, it would not be sufficient With regard to value content, the method depends on placing values on these largely natural bivariate relationships. Once again, if the decision is well structured with little value conflict or socio-economic concerns, the approach may be sufficient However, as Gold (1974) illustrated, it is not able to resolve difficult value conflicts. A final consideration involves the availability of resources and information. In a simple system with a fairly comprehensive amount of descriptive information, or if the means exist to collect such information, the McHarg method could prove to be a simple and rapid form of analysis. In a more complex system, functional information may be required and as a result this advantage becomes irrelevant as additional information will be required in any case. In reviewing the case studies, it is apparent that one of the most effective applications involves suburban design where the difficult strategic or tactical questions have already been answered or dismissed. . Under these circumstances, the evaluation is limited to determining physical suitability and performance standards, as was the case in Woodlands and Medford. It has also been recommended for studies with specific objectives at the regional scale, like the routing of a highway corridor. As long as the limitations are clearly understood, it can be effective in providing an initial analysis of descriptive information. Descriptive Models The McHarg approach is not sensitive to many of the principles involved in the various descriptive models. It's emphasis on ecological determinism, intrinsic natural values, and the analytical approach used promotes a rational comprehensive decision process as reflected in normative models. It is recognized that there are other 64 considerations in the decision process which may circumvent the rational comprehensive model, but these are not considered relevant to the evaluation itself (McHarg, 1969). This is not to say that the ecological value judgements are necessarily inappropriate, only that if other factors dominate the decision process, the 'use it as it is or discard i f option offered by the economic determinism inherent in the McHarg method may lead to the latter. Furthermore, effective use may also require that the value premises underlying the approach be explicit and obvious to the decision maker which is not the case with a McHarg analysis (Sinden and Worrell, 1979). Decision Criteria The strengths and weaknesses of this approach can also be related to the five decision criteria introduced in Chapter Three. 1. Problem Exploration Ecological land evaluation does allow exploration of some elements of land use problems. It aspires to illustrate a comprehensive range of relationships and interactions between natural system components and these can be explored to a degree by examining the matrices. However, opportunities to explore a problem beyond a general ecological awareness of the decision enviroment are limited. Alternatives cannot be explored without restructuring the analysis and changing values deemed to be intrinsically good and correct prior to beginning the evaluation. Litton and Kieiger (1971) support this view and indicate that there is some doubt that the method produces more competent planners or decision makers, aside from a general awareness of ecology. 2. Flexibility Ecological land use evaluation is not particularly flexible, constrained by both the mechanics of the approach, and the predetermination of values. The flexibility which does exist is provided by the opportunity to provide different combinations 65 of map overlays in the synthesis phase. The decision maker has however, very little control over the evaluation and few opportunities to pose 'what i f questions. Any new information or changes in the decision environment essentially invalidate the assessment, requiring restructuring of the initial value assessments and suitability ratings. Data Management The McHarg approach deals largely in descriptive inventories. It also provides few information management functions beyond the organization of descriptive inventories. Litton and Kieiger (1971) also criticized the role of information in the McHarg approach, noting its emphasis on collection of vast inventories as opposed to an economic data gathering approach. There is a lack of any ability "to assess the information component of the evaluation, nor does the approach encourage information generation through research aside from the aforementioned descriptive inventories. Credibility The McHarg approach has a certain amount of credibility, partly as a result of its general acceptance and reputation. Of course, any single evaluation may vary considerably depending on the reliability of the inventory information and the dependability of the expert value judgements. The approach has been criticized as simplistic and unsystematic in its use of ecology (Lee, 1982; Litton and Kieiger, 1971). In addition, it does not usually provide explicit explanations of its limitations and qualitative components, at least in the documented applications. Meaningful This is one of the strengths of the method. It does provide a clear evaluation using familar products like maps and matrices which are intuitively meaningful and can be related directly to the landscape. Most users and decision makers should be able to comprehend the basic message conveyed by an ecological land 66 evaluation. 4.2.2 S I R O - P L A N Decision Environment The SIRO-PLAN methodology, with and without LUPLAN, has been used in a variety of projects in Australia. Two representative examples are listed below. 1. Redland Shire Project - LUPLAN was used as an input into a regional planning process in Redland Shire, a rural area on the fringe of a major metropolitan centre. Land use conflicts involved rural residential developments, recreational land use, and conservation. The exercise used 49 policies, 7 land use options, and 413 planning zones. Discussion plans were produced from the perspective of planners, conservationists, and developers (McDonald and Brown, 1984). 2. Dungog Shire Environmental Plan - SIRO-PLAN and LUPLAN were used in producing a land use plan in Dungog Shire, another rural area undergoing development pressures. This was a very large project involving 18 land uses and 200 allocative policies. Evaluation was used in preparation of a zoning scheme for use by the Dungog Shire Planning Committee (Davis and Ive, 1985). The methodolgy has also been used as a policy formulation tool (Davis and Greenhalgh, 1980), as a way of organizing information for planning (Baird, 1981), and as an educational tool (Baird, 1981). As a land evaluation tool, SIRO-PLAN and LUPLAN have been applied primarily in rural or conservation planning contexts. It would seem to be most appropriate in a tactical, semi-structured environment for several reasons. It can accomodate different values and provides some means of understanding value conflicts. These values still require explicit articulation in the form of weights, and no satisfactory method has been developed to generate value weights (Tivy, 1980). Participants in decision enviroments with value conflict may also be reluctant to attempt to express or simply reveal explicit values (Kunruether, 1981). This is a problem which has reduced the effectiveness of the technique (Davis, pers comm.). SIRO-PLAN can accomodate a diverse range of analytical tools but does require a good scientific knowledge base to produce credible ratings. This knowledge 67 may not exist in a poorly structured decision environment Lastly, the LUPLAN evaluation component is limited to spatially oriented, intrinsic evaluation of land units. The aggregate results defy the long range economic analysis and consideration of spatial interdependencies important for poorly structured, strategic land use decisions (McDonald and Brown, 1984). SIRO-PLAN is also poorly suited for operational and well structured decision environments. If no value conflicts exist or if the policy questions and tactical land use questions have already been resolved, then some of the major strengths of SIRO-PLAN have been neutralized. In addition, the scale of resolution and analytical capabilities are not particularly amenable to the site specific design, evaluation, and feature identification required by operational decision making. Descriptive Models Most of the discussion required to relate SIRO-PLAN to descriptive models has already taken place. It includes features characteristic of rational models - the programming involved in LUPLAN - but was also designed to allow incremental changes with a view to developing a satisfactory policy. The implications of various political values can be tested and LUPLAN can be used effectively as an educational device for participants in the process (McDonald, pers. comm.). Problems have developed however during application of the technique as those involved in the political process have resisted the implied modelling, and some have felt replacement of the political process by the model (Davis and Ive, 1985). SIRO-PLAN can accomodate some organizational concerns as standard operating procedures can be expressed in both the policy sets and the rating codes. Decision Criteria Several observations can be made regarding this approaches strengths and weaknesses in the context of the decision centered design criteria. 68 1. Problem Exploration SIRO-PLAN does allow extensive exploration of potential land use decisions. This is especially true of policy values and their impact on various planning solutions. McDonald and Brown (1984) commented on the transparent nature of the approach and argued that the technique forces the decisionmaker to consider a wide range of factors involved in suitability and alternative land uses. Interactive use of LUPLAN allows 'what i f questions related to the impact of different value profiles, the sensitivity of the system to values, and the areas of greatest potential land use conflict It is slightly less effective at exploring the dimensions of the problem. Construction and manipulation of data sets can provide important insights into the planning context of the decision environment Identification and rating of land units provides some insight into the character of the study area and the link between land use information and policy requirements. It does have a weakness in that it does not provide many general insights into the ecological nature of the study area as was characteristic of the McHarg method. 2. Flexibility The flexibility of the system is one of its greatest strengths. It can accomodate a wide range of analytical techniques, data types, value perspectives, and policy sets. The approach is also transparent in that these inputs and controls are explicit and easy to identify, thus allowing manipulation as directed by decision makers. SIRO-PLAN can also handle changes in the decision environment by altering policy sets and weights. New data, policies, and control instruments like a newly developed standard operating procedure, can be incorporated easily into the framework (Cocks et al., 1983). The one limiting factor of note is the reliance of LUPLAN on a single aggregating function, linear additive 69 programming, which restricts the analytical options somewhat Information Management SIRO-PLAN can be an effective tool for information management especially when used in conjuction with LUPLAN. It organizes existing data into a single framework and directs data collection and future information generation by providing a link to the land use policies to be served by the information (Cocks et al., 1983). By using the approach, it also becomes clear what information is in fact critical in establishing values in the decision process. One disadvantage is that if a user requires some general information beyond the policy set it would have to be accessed through a separate exercise. To resolve this deficiency, there has been progress made towards linking LUPLAN to other information systems, like the Canada Land Data System (Yapp et al., unpublished). Credibility Credibility is difficult to judge without a longer history of use. Technically, its credibility depends on the knowledge available to calculate suitability ratings. It was found in completing the second part of this thesis that credible rating techniques are not available for many of the types of policies identified. The major advantage is the transparent nature of the technique. The subjective component is explicit leading Cocks et al. (1983) to declare that there is no mystique involved and people can see how the decision is made. On the other hand mistrust of the computer model and its handling of value sets has been experienced (Davis and Ive, 1985). Meaningful SIRO-PLAN is a new approach and its land evaluation component requires some active thinking on the part of users and decision makers. Linking land evaluation to explicit policy statements should be understandable to those involved in the 70 process. The output of suitability ratings is not particularly meaningful if the numbers are taken as isolated, absolute values in themselves. Careful interpretation is required and once again, some may resist the use of numbers produced by a computer model. Some of these problems could be overcome by the inclusion of interpretive aids, like land use compatibility matrices as found in the McHarg approach. The use of preferred use maps and zoning schemes is well within the experience and understanding of most land use decision makers. 4.2.3 LAND EVALUATION MODEL Decision Environment The LEM is best suited for strategic broad scale, long term planning and evaluation tasks (Smit et al. , 1981). To date, it has been used to explore issues at the provincial level in Ontario. Evaluation exercises completed include: 1. Examination of the effects of growth in the Agri-Food sector on land use options. This study examined the impact of increased food production targets on future options and strategic value of land units for particular agricultural activities. The model indicated that it should be possible to satisfy demand to 2001 but there would be virtually no flexibility remaining in the land use system (Dyer et al., 1983). 2. Examination of the effects of urban expansion on agricultural land needs. In this application, the model was manipulated to reflect urban expansion anticipated by 2001 and the impact on importance of land areas was predicted. It was found that agricultural flexibility would be seriously reduced and that only marginal land would not be critical for food production (LEG, 1983). The modelling approach deals with the uncertainty and lack of structure involved in these decision problems by allowing for the explicit generation of different scenarios and by calculating flexibility and criticality, rather than suggesting optimal uses. The objective is one of measuring the range of options available given the specified conditions (Smit and Flaherty, 1984). The programming techniques also deal with the variety of information inputs involved in these poorly structured problems by producing a measure of value which is dimensionless. 71 Descriptive Models It is difficult to characterize the LEM approach in regards to the descriptive models. The programming technique itself is a rational or optimizing approach. Nevertheless, the system, with its use of scenarios and value measures, is not designed to produce a single best answer. Smit et al. (1981) explain that the system "does not replace human decision makers; it assists in making land use policy and planning problems more manageable for decision makers" (p.7). When taken as a whole, the LEM system could be used in harmony with a number of characteristics of positive models. It can support an incremental approach through the systematic manipulation of constraints. Organizations can select input which reflects the programs of concern in their operating procedures. While the general approach may be used in a number of ways to assist different descriptive models, it is a complex research project The modifications necessary to generate scenarios could preclude its use in many cases. Decision Criteria The decision criteria emphasizes several strengths and weaknesses of the LEM. 1. Problem Exploration The opportunity to use the LEM to explore a decision problem is promoted as a significant strength. The user is not committed to any single assumed values and can test different scenarios to judge their impact on land importance (Smit et al., 1981). In this way, 'what i f questions related to urban growth, erosion policies, and even climatic change can be answered. The amount of flexibility in alternative options and the range of acceptable solutions can also be explored. With respect to the nature of the problem addressed by any particular exercise, considerable understanding must be gained in order to formulate the correct constraints. This understanding is difficult to communicate to a user although basic input data related to a decision problem can be extracted independent of 72 the analysis. 2. Flexibility The LEM is also promoted as being very flexible with countless conceptual possibilities for the inclusion of constraints (Smit et al., 1981). A user can exert some control to utilize the flexibility for creative problem solving. This is limited in that the model structure is fixed so that the analytical operation itself cannot be varied. New or expanded problem sets can be added by updating the data base and providing new constraints (CRD., 1978). Once again, any new constraints do require information which may be difficult to obtain or express within the confines of the model. 3. Information Management The LEM can be a very effective information management tool. It organizes and provides user access to a variety of information including productivity models. It has also been linked to other data banks like the Canadian Soil Information System (CANSIS) (Dyer et al., 1983). Development of the system has also indicated data deficiencies and directly initiated research on such inputs as productivity models (Smit, pers comm.). In this way, development of the constraint sets have focussed research and generation of functional information at the University of Guelph. 4. Credibility The LEM has not been used as a practical tool beyond applied research on a provincial level. The results produced thus far would seem to validate the model as an accurate decision tool at the strategic level. Some problems can be seen in more practical planning applications in that even if it is conceptually understood, the complex programming function may be viewed with some suspicion by decision makers. 73 5. Meaningful The LEM may encounter some problems with this criteria simply because it is new and beyond the experience of most decision makers. The evaluative measures being expressed on a scale from 0 to 1, seem to make sense but to be meaningful, they must be accompanied by a good understanding of the conceptual basis of the model. This may be beyond the ability or desire of a decision maker busy with day to day concerns. 4.3 SUMMARY Table 3 provides an evaluative summary of the strengths and weaknesses of the three approaches. Consideration of these sorts of strengths and weaknesses when choosing between existing land evaluation systems within the context of the decision environment involved in an exercise is important (McAllister, 1980; Poulton, 1983). By viewing evaluative techniques in this manner, an evaluator can create an approach which is a more effective decision aid. While such analyses of available techniques would supply important guidance, and the ability to select approaches from a 'menu* of this sort may be sufficient in some cases, it is still necessary to address fundamental design problems. Decision Support Systems (DSS) provide a logical design progression from current appraoches, and several DSS concepts have already appeared in techniques like the LEM. The following chapter introduces Decision Support Systems as a design approach. Table 3 Comparative Summary of Current Approaches • Decision Descriptive Decision Criteria Envi ronment Component 1 2 3 4 5 McHarg Method Best for well structured, operational decisions or tactical decisions wi th well defi ned objecti ves. Rational values using an ecological perspective. O • • O SIRO-PLAN Best for tactical decisions with spatial variables. Allows incremental satisficing and can accomodate a variety of decision making approaches. O O o LEM Best for poorly structured, long term, strategic decisions. Provides rational analysis but not an ootimal answer. Allows variance of value input. o o * o • ^ Weakness Strenoth Nuetral or dependent on additional ^^on additional work 1. Explorative capability 2. Flexibility 3. Data management 4. Credibility 5. Intuitively meaningful Chapter 5 DECISION SUPPORT SYSTEMS Several salient points have been introduced in the thesis to this point It has been observed that even with the wide range of technical approaches to land classification and evaluation, problems in using the information supplied persist To improve the use and design of land evaluation systems, it must be recognized that land evaluation is ultimately meant to support decisionmaking. If this claim is accepted, it follows that a number of concepts important in decision making have significant implications for land evaluation and are in fact implicit in many approaches. These concepts include the decision environment descriptive models, and decision analysis. For the conceptual model introduced in Chapter One to function properly, these concepts must be linked with the technical considerations during design of the system. As a means of explicitly incorporating these concepts into the design of evaluation systems, this chapter outlines a decision centered approach to design based on decision support system development In a decision centered approach, land evaluation is viewed not as an end product in itself, but rather as a decision support system combining with intuition for more effective land use decision making. Fabos (1978) expressed one of the central tenets of this approach as follows: Ideally, intuition would work hand in hand with an evaluation system, selecting the criteria to be included in the evaluation, and interpreting the output into feasible and efficient recommended uses. Evaluation should be viewed as a means to reinforce, strengthen and facilitate intuition, not replace it (p.26). This is the exact perspective found at the core of decision support systems as revealed in the following description of decision support systems by Thierauf (1982): The emphasis is on enhancing the decision making process by allowing the individual to utilize mathematical and statistical models that are appropriate for the problem being solved and leaving the remainder to the manager for reaching a final decision. The key point is the enhancement of the manager's decision making ability by allowing the computer to do what it does best (i.e. objective, quantitative measurement) and letting the manager introduce subjective qualitative factors (p.59). 75 Decision support systems will now be explained by first outlining the evolution of the concept, then providing a definition including important characteristics, and concluding with an outline of the decision support system design process. 5.1 E V O L U T I O N There is some controversy surrounding the origins of decision support systems. Some claim that they are simply a type of management information system (Ein-Dor and Segev, 1981) while others consider them a new form of technoglogy (Keen and Morton, 1978). One view which provides a clear explanation and logical sequence of events is that decision support systems have evolved from electronic data processing and management information systems. They have not however replaced these technologies, but are in fact a distinct class of management information system which often works in concert with the other two more traditional tools (Sprague and Carlson, 1982). As the need for, and volume of, information increased at an incomprehensible rate, and with the advent of computer technology, electronic data processing (EDP) became the driving force in the advancement of information technology. It allows the storage and rapid retrieval of large volumes of data for routine and clerical tasks (Moore and Chang, 1983). While EDP may improve the efficiency of some decision making tasks, they lack the manipulative and interpretive capabilities desired for information management To improve on this deficiency, management information systems (MIS) have developed. MIS supply predefined data interpretation, aggregation, and reporting capabilities (Moore and Chang, 1983). The main impact of traditional MIS has been on relatively structured tasks in which standard operating procedures, decision rules, and information flows can be defined prior to the decision process (Thierauf, 1982). 77 Applications of MIS have been expanding in the land classification and planning field (see Meyers, Kennedy, and Sampson, 1979; Nijkamp, 1983). One representative example is ECOSYM, a classification and information system for wildland resource management (Henderson, Davis and Ryberg, 1981). ECOSYM uses a comprehensive data base and decision rules to perform a variety of land classification tasks from vegetation analysis to range productivity ratings. In Canada, the Canadian Geographic Information System and Canada Land Data System provide the capability to access environmental baseline data and manipulate that data to produce map overlays and summary tables (Thie et al., 1984). Figure 12 illustrates the Canada Land Data System which is typical of large information systems dedicated to processing descriptive information. Organizations have also appeared devoted to MIS. These include the International Society of Soil Science working group on Soil Information Systems (Sadvoski and Stein, 1978) and the Urban and Regional Information Systems Association. Information systems add a new dimension to information management but are still somewhat limited. Decision makers began to seek service oriented systems to support problem solving and planning at a variety of levels and for a variety of decision types. MIS allows decision makers to extract information, but not explore what would occur under a range of options (Huber, 1982). To counter this limitation, some information systems have developed an added dimension. PREPLAN, a resource management information system developed in Australia (Kessel et al., 1984) combines an integrated data base with predictive modules for vegetation, fire effects, and fauna. Besides organizing and integrating information, PREPLAN also allows the user to change input parameters to simulate 'what i f situations. The increasing frequency of smaller scale, interactive additions to management information systems such as the capability in PREPLAN has prompted the recognition and development of decision support systems. INPUT-iSYSTEM OVERVIEW! - •MANIPULATION • • O U T P U T r n J D I G I T A L D A T A J J - STANDARD FORMAT* L.-.-- . . . . J UK ;~Yt» * •* • DRUM SCANNER ' i 1- DATA BASE CREATE 2- AREA & PERIMETER CAU^ANCiN 5 - SUBSETS 6 - DISSOLVE 7- DATA CONVERSION POINT OATA DIGITAL DATA LINKS SEGMENT FILES DESCRIPTIVE FILES POLYGON FILES • SPSS FILES • STANDARD FILE ASSESSMENT LANGUAGE GRID CONVERSION FILM IMAGE RECORDER TABLES & MAPS ON A CRT SCREEN DIGITAL PLOTTER GERBER 42 COLOUR SEPARATION UNDER DEVELOPMENT R&D STAGE. FIGURE 12 - Canada Land Data System (from Thie et aL, 1984). 79 5.2 DEFINITION A decision support system is difficult to define or label in that it can take many forms. It may deal with a specific problem or a general problem area and usually involves a human-machine information processing system (Bennet, 1983). Decision support systems are best understood by examining their purpose and characteristics. The purpose of decision support systems (DSS) is to improve the performance and effectiveness of decision makers' (Sprague and Carlson, 1982). More specifically, decision support systems: can serve a variety of decision types but are most effective where there is enough structure for computer and analytical aids to be of value but where judgement is an essential component of the decision process. extend the range and capability of a decision makers decision process to help them improve their effectiveness (Sprague and Carlson, 1982). create a supportive tool, under the control of the decision maker, which does not attempt to automate the decision process, predefine objectives, or impose solutions. Several characteristics are important in achieving this purpose and these are listed below: Flexibility and adaptability to accomodate changes in the environment and decision making approach of the user (Sprague and Carlson, 1982). Inclusion of interactive dialog components for easy use by noncomputer people in a human/machine interface. Utilization of a management by perception approach designed to be forward looking and to promote an understanding of the system first (Thierauf, 1982). Inclusion of integrated data base and query capabilities to extract information (Thierauf, 1982). Lack of commitment to any one distinctive technology. (Keen and Gambino, 1983). Integrated subsystems are often involved to provide technology and decision process options (Thierauf, 1982). Parallels with decision centered land evaluation requirements should now be obvious. Decision support systems are designed to provide the flexibility, the capability to explore decision problems and incorporate value judgements, and facilitate important information management functions. DSS are also best suited to the type of 80 semi-structured decision characteristic of land evaluation tasks. 5.3 DESIGN Decision support systems are not developed in one design cycle with fixed steps and well-defined boundaries (Hurst et al, 1983). Rather, the design process is one of adaptive evolution and iteration (Keen and Gambino, 1983). This usually involves supplying a small, flexible system, allowing decision makers to react to the system, and then making necessary adjustments or, if need be, major revisions. Adaptive evolution is a continuous process with development continuing as understanding of the natural and decision environment changes (Keen and Gambino, 1983). This design strategy is accompanied by a distinctive analytical approach. Three general analytical options are available (Hurst et al., 1983). The 'bottom up' approach involves creating individual tools which are later combined and applied to an evaluation problem. It suffers from a lack of decision focus as tools become ends in themselves with no appropriate application. The 'top-down' approach attempts to define a global, wide-range plan with a prespecified problem structure. With this approach, an exorbitant amount of time is spent doing global designs without ever addressing concrete details (Hurst et a l , 1983). Decision support systems usually adopt a 'middle-out' approach. This approach begins much closer to the level of the immediate decision problem. Once some experience is gained with the problem at hand, the designer can learn from the user and generalize specific tools (bottom-up) and specify general policies (top-down) as required (Hurst et al., 1983). Within the process, a difficult trade-off should be acknowledged. This involves the effort to support the existing decision process while encouraging movement towards more effective decision making. Just how far, and how quickly, to move towards innovative approaches must be decided. Keen and Gambino (1983) suggest that the 81 first priority should be to support the existing decision process, and then slowly improve the process as user learning occurs. It is not possible to provide a recipe for constructing decision support systems which would be generally applicable in all cases (Keen and Morton, 1978). It is nevertheless possible to suggest a procedure and related organizational components which will help to combine decision requirements, information resources and current technology into an effective DSS. Four design considerations will be used to identify a design framework: Decision framework, Technology level, Tactical option, and Design guideline. 5.3.1 DECISION FRAMEWORK The design process should begin with a decision framework. One of the most popular and generally applicable models is the intelligence-design-choice paradigm as intially developed by Simon and used in some of the DSS work to date (Sprague and Carlson, 1982; Thierauf, 1982). This framework identifies three phases of decision making which are applicable regardless of the decision environment These phases are as follows 1. Intelligence - This involves searching the environment for conditions calling for decisions. Raw data are obtained, processed and examined. 2. Design - Inventing, developing and analyzing possible courses of action. This involves processes to understand the problem, generate solutions and test solutions for feasability. 3. Choice - Selecting a particular course of action from those available. Figure 13 illustrates some of the functions involved at each stage and representative examples from land evaluation. By using this paradigm, the design process can become focussed on decision making requirements. 5.3.2 TECHNOLOGY LEVELS Decision support systems do not instantly appear as a single entity, as is often the case with a management information system. They are more often a synthesis of 82 INTELLIGENCE Gather Data - Soils, Resources Identify Objectives - Increased Productivity, Conservation Diagnose Problem - Land Use Conflicts Validate Data - Accuracy, Scale Structure Problem - Land Tenure, Institutions DESIGN Manipulate Data - Technical Mechanics Quantify Objectives - Production Level Generate Alternatives - Land Uses, Policies Assign Risks or Values to Alternatives - Land Value Measures Generate Reports - System Output CHOICE Generate Statistics on Alternatives - Comparative Productivity Simulate Results of Alternatives - Land Use Plans Explain Alternatives - Land Use Policies Choose Among Alternatives - Political, Value Content Explain Choice - Final Plan FIGURE 13 - Intelligence - Design - Choice Paradigm (Adapted from Sprague and Carlson, 1981). three levels of technology (Sprague and Carlson, 1982). 1. Specific DSS A specific DSS actually accomplishes the evaluative operation that allows a decision maker to deal with specific sets of related problems. 2. DSS Generators A DSS generator provides a set of capabilities to build specific DSS quickly and easily. Computer software and instructional manuals used in building models for specific evaluations could be DSS generators. 3. DSS Tools DSS tools facilitate development of specific DSS through DSS generators or, in some cases through direct application. Examples would include analytical models and related information for dealing with a component part of an evaluation exercise. 83 Figure 14 provides a graphic illustration of the relationship between these parts including examples extracted from land evaluation. 5.3.3 TACTICAL OPTIONS Different circumstances will require different tactical strategies. Some problems would not warrant construction of an elaborate or complete DSS as an initial objective. If the proper tactical option is not selected, the system may prove to be either incomplete, or too expensive and comprehensive for effective application. Three tactical options are available: 1. Quick Hit If it is not clear that a general DSS capability is needed, but there is a recognized benefit for decision support, it is possible to develop a specific DSS directly from DSS tools, capture the benefits, then consider future options. 2. Staged Build one specific DSS, but with some advance planning, so that part of the effort in developing the first system can be used in developing the second. With proper planning, a DSS generator evolves from the development of several successive specific DSS. 3. Complete Before building any specific DSS, develop a full service DSS generator, and the organizational structure for managing it The lack of a tactical strategy has created some problems for management information systems. By attempting to develop a complete, comprehesive information system in one design exercise, designers of information systems have found themselves with large, expensive products which turn out to be of limited use in the decision process. Decision support systems avoid this problem by explicit consideration of tactical options. 5.3.4 DESIGN GUIDELINES When designing a decision support system, it is important to have some concept of the components of the system and how they relate to the decision process. 84 SPECIFIC DSS SPECIFIC DSS DSS TOOLS DSS TOOLS FIGURE 14 - Technology Levels for Decision Support Systems (adapted from Sprague and Carlson, 1981). This guidance is provided by the Representation-Operation-Memory Aid-Control Aid design paradigm (ROMC). The R O M C model is used to structure the thoughts of a designer when formulating the components of a decision support system. Representations are meant to simulate the ways in which a decision maker conceptualizes the 85 information involved in the evaluation. The actual model or algorithm which completes the evaluation is termed the operation. Memory aids represent the way in which the decision maker stores information, while control aids allow the decision maker to control the system. Table 4 provides a list of examples of each along with some equivalent activities in land evaluation. By developing a design around these components, specific aspects of a DSS can be related directly to decision making. 5.4 RQLES IN DSS There are a number of roles in decision support systems as demonstrated below: Manager/User Decision maker Intermediary R a m Planner/Clerical Staff Builder IB9u|V' Scientist/Planner Technical Supporter Scientist Programmer/Scientist Programmer/Scientist Of particular interest is the implication of decision support systems for the role of the land use planner. The role of the planner has never been easy to define. A planner can be seen as a technical expert, an advocate, a mediator, or simply a bureaucrat implementing goverment policy. He or she can also be a decision maker, especially at the operational level. This effectively includes all of the roles involved in DSS. However, within the decision support systems framework, the planner can become an important intermediary, albeit one with considerable expertise and a legitimate professional opinion. The role becomes one of an analysts who assists those with the responsibility to make a decision, and ensure that they are aware of the assumptions behind any value judgements as well as the implications of the decision (Sinden and Worrell, 1979) This role becomes especially significant with the increased use of computers. Planners are not computer experts, but are often computer literate and able 86 Table 4 R O M C Framework Oecision makers use Decision support systems provide Conceptualizations - s p a t i a l units - system interactions - c o n f l i c t s i tuations Representations - maps - flow diagrams - Ecological Characterizations - Compatibility or impact matri ces Analysis - Compare alternatives - Consider implications of change Operations - Alternative evaluation - S e n s i t i v i t y Analysis - Scenario Generation Memory - Land c h a r a c t e r i s t i c s - Past results Memory Aids - Data banks - Recording and updating functions Personal Control - value prediction - selection of c r i t i c a l parameters - interpretation of resul ts Control aids - a b i l i t y to change value sets - interaction with data banks - interpretive a i d s , guidelines and explanations 87 to translate decision making needs to those responsible for computer software and hardware development A planner can ultimately find him or herself using a DSS as a decisionmaker, or working with other experts to help construct the system itself. 5.5 SUMMARY This chapter has introduced the concept of decision support systems within the context of a decision centered approach to land evaluation. This concept is not offered as a panacea or as a solution to all the problems plaguing the conceptual model. It is also a new concept which is poorly developed, especially with regards to land evaluation. What the DSS concept does provide is an alternative approach which seems to hold considerable promise as a way to address the creation of land evaluation systems. Chapter Seven will begin to demonstrate this potential by using an actual land evaluation task m combination with the framework provided in this chapter. Chapter 6 A DECISION SUPPORT SYSTEM FOR LAND EVALUATION This chapter provides an illustration of a preliminary decision support system for land evaluation in a specific study area. The area chosen is the eastern portion of the Central Fraser Valley Regional District in the Lower Mainland of British Columbia. Three objectives can be identified for the chapter: 1. Demonstrate the Decision Support System design principles and use. 2. Demonstrate how to examine the strengths, weaknesses and mechanics of an evaluation system to ensure proper use. 3. Provide an initial land evaluation tool for the study area. After this brief introduction, a design process is demonstrated using the Intelligence - Design - Choice paradigm. This is a simplification in that all three activities occur throughout construction and use of a decision support system, not in three discrete steps. Notwithstanding this qualification, the model will be used as an. organizational tool for documentation of a macro design cycle. The resulting decision support system consists primarily of LANDPLAN, a Canadian modification of the LUPLAN package reviewed in conjuction with SIRO-PLAN in Chapter Four. 3 Figure 15 provides a flow diagram of the demonstration exercise 6.1 STUDY AREA The study area consists of the District of Matsqui, Town of Abbotsford, and Electoral Area A (Sumas Mountain), all located in the eastern half of the Central Fraser Valley Regional District (CFVRD) as shown on Figure 16. This area is bounded on the north by the Fraser River, on the south by Washington State, on the west by the District of Langley, and on the east by the Regional District of Fraser-Cheam. LANDPLAN is a version of the LUPLAN model described in Chapter Four developed at the Lands Directorate of Environment Canada. Appendix one contains a complete reference and other details. 88 INTELLIGENCE DESIGN • CHOICE Decision Problem Natural Envi ronment Decision Envi ronment I *~— Available Data Decision Requirements Development Option - Staged Approach Representations Operations Memory Ai ds Control Aids Select Technique LANDPLAN Land Uses I Policy Set Data Bank •*-\ E val uati ve Operation - v Rating Codes / Data Analysis 1 Compatibility Matri x Decision Support System - LANDPLAN - Data Analysis - Matrix FIGURE 15 - Design Process for the Decision Support System Demonstration FIGURE 16 - Location Map of Study Area 91 Abbotsford - Clearbrook acts as a major regional centre while the Trans Canada Highway provides a primary transportation corridor. The remainder of the study area is predominantly a rural resource land use environment. Physically, there are three distinct types of landscapes. Sumas Mountain provides a forested environment, there are several low-lying prairie areas under cultivation, and there are upland areas with a mix of rural uses. Land use conflicts include most of those found in a near urban rural environment 6.2 EVALUATIVE TASK The decision problem to be confronted is the allocation of rural land use within the study area. This requires some form of land evaluation. In this demonstration, the evaluation occurs at the tactical level and is a general purpose evaluation in that several land uses are rated and compared for each land area. Several limitations and assumptions used to bound this task should be explained. As a general comment, it needs to be emphasized that the decision support system provided represents the intial stage of the iterative design cycle, a system meant as a preliminary tool provided quickly and efficiently to act as a step towards a complete system. This point becomes clear as the chapter progresses. With this in mind, other limitations are as follows: The evaluation must use either existing mapped information, or data that can be easily obtained from published reports, interpretation of base maps, or examination of aerial photography. The DSS is limited to existing computer tools and evaluative methods. The expertise for original software development does not exist The DSS is not being built from within an agency or a particular decision process. Although this would be desirable, the level of commitment and resources required from those involved is prohibitive. The evaluation is based largely on physical and spatial information. Some economics are incorporated implicitly through consideration of current use and spatial patterns but there is no attempt at integrated analysis. It is also intrinsic in nature. These limitations are discussed further while explaining the selection of the method. 92 These limitations are largely a result of the constraints involved in the current research project. While they do serve to bound this study, and some are regrettable, they also serve to mimic reality. Most are not uncommon in many land evaluation tasks today. One of the benefits of the DSS design process is that it begins close to this reality and points the way to an improved and more effective decision process. 6.3 INTELLIGENCE The intelligence stage is centered around three activities. Decision diagnosis involves an examination of the natural system and the decision environment to determine decision structure and account for descriptive models. The second important activity is the assessment of available information. Finally, the first two activities are combined to outline system requirements, objectives, and selection/design guidelines. 6.3.1 DECISION DIAGNOSIS Intelligence begins with the identification of conditions leading to decision problems. In the study area, conditions identified include ongoing competition for available land, conflicting resource use pressures, and land conservation concerns. These conditions lead to a decision problem which could be expressed as: How to provide for an evolving mix of resource and land use activities in a near urban rural environment? This decision problem clearly requires some form of land evaluation. To provide the direction required for decision support, the problem conditions and resulting structure should be reviewed. Physical Environment The subject area can be divided into nine landscape units based on the three natural landscape types mentioned earlier and these have been labelled on Figure 17. These units create unique combinations of physical attributes and land use. In each of 93 Nathan Lowlands organics and Fraser River Sediments agricultural use FRASER RIVER Langley Uplands glaciomarine deposits agriculture and mixed rural uses Matsqui Prairie Fraser River sediments 2nd organics agricultural land UM1 PSunus Mountain pre-tertian bedrock with a thin mantle of glacial, colluvial and eolian sediments; steeply sloping forested landscape forestry, recreational uses SUMAS RIVER Resource Fringe mix of gtaciobnarine and topography. mi\ of rural uses glacialfluvial deposits, roiling and broken Sumas Prairie sand to sand; loam lacustrine deposits agricultural land use Abbotsford Uplands glacialfluvial sand and gravel agricultural land uses (berry cultivation) Sumas. Corridor X organic deposits industrial and rail development mixed with agriculture I Vedder Mountain y pre-tertiary bedrock, stecplj sloping forested landscape m forestry, recreational uses "| Clearbrook-Abbotsford 2 Abbotsford Airport 3 Sumas Mountain Provincial Park • N kms. 1 o kms. FIGURE 17 - Study Area Landscape Map 94 the nine landscape units, well-defined physiography and drainage characteristics have encouraged a particular land use pattern. As a result, the units provide a spatial perspective to organize and structure the decision problem, much like a landscape classification at a reconnaisance scale. Prairie landscapes are represented by the Sumas and Matsqui Prairies, a portion of the Nathan Creek lowland, and an area of organic deposits in the Sumas corridor. These areas are low-lying, approximately ten metres above the Fraser River under normal conditions, and are mostly poorly or imperfectley drained. The prairie landscapes are underlain by either Salish sediments which are lacustrine silt, clay, or sand deposits (Sumas Prairie) or Fraser River sediments consisting of channel fill and floodplain deposits (Matsqui and Nathan). Areas of organic Salish Sediments are also dispersed throughout lowland areas (Armstrong, 1984) These physical attributes have encouraged agricultural land uses. Dairy and crop farming are predominant with some mixed farming and market gardening (Manning and Eddy, 1978, District of Matsqui, 1980). Upland landscapes consist of rolling plateaus which dominate the western half of the study area. The Langley Uplands are underlain by the Fort Langley Formation which are glaciomarine deposits of stony silt to loamy clay (Armstrong, 1984). Moving east from the Langley Uplands, there is an area of rougher topography consisting primarily of glacial till deposits from the Sumas Drift formation. The Abbotsford Uplands are also underlain by Sumas Drift glaciofluvial outwash deposits but are relatively level. These areas are also mainly agricultural. However, largely as a result of topographical limitations, there are numerous patches of natural vegetation. Agricultural practices tend to be more diverse with intensive uses, like poultry operations and rural development, alongside dairy, pasture and crop farming (Disrict of Matsqui, 1980). Of special note is the extensive cultivation of berry crops in the Abbotsford Upland area. 95 Other rural uses spread throughout the uplands include residential lots, commercial and institutional uses, and transportation facilities as represented by the Abbotsford Airport Within the agricultural community, there are also a number of small landholdings operated as hobby farms (District of Matsqui, 1980). The final landscape unit consists of mountain environments, primarily Sumas Mountain along with a small portion of the northern slopes of Vedder Mountain. Sumas Mountain is the most prominent physical feature in the Fraser Valley. Physically, the mountain areas are moderate to steeply sloping (generally greater than 30%) with a shallow soil layer (less than 1.2 metres) of sandy to fine silty loams over sedimentary bedrock. Both areas are forested with Coastal Douglas Fir, Western Hemlock, and Western Red Cedar the dominant species (Ministry of Lands, Parks, and Housing, 1985). The lower slopes of Sumas Mountain adjacent to Abbotsford- Clearbrook support a mix of residential uses along with some small agricultural holdings. The remainder of the mountain is used for extensive recreation and a significant, albeit small scale commercial forest operation. Several additional natural features deserve mention. The Fraser River abuts the study area with a number of sloughs and creeks draining the adjacent areas and outletting into the river. The other major drainage feature is the Sumas River which is fed by drainage courses throughout the- Sumas Prairie and some mountain streams flowing down the southern slopes of Sumas Mountain. Less significant drainage systems include the Salmon River and Nathan Creek, both finding their source in the Langley Uplands. Pepin creek and Fishtrap creek are two small, isolated watersheds draining south across the international border. These landscapes do not contain any atypical or untouched natural systems. To use a typology developed by Odum (in Rodiek, 1978), the study area consists largely of production and compromise environments. The first category is characterized by 96 growth systems with high production rates and simplified speciation. Compromise environments are just that, a compromised mix of uses typical of, for example, urban fringe areas. Some areas of Sumas Mountain do display the maturity and longer life cycles found in protection environments but the entire area has been changed and manipulated by human activity and productive systems. While such a rural environment may not be considered as sensitive or complex as more highly productive or undisturbed ecosystems, there is an underlying complexity created by interactions both within and between landscape units. For example, forest harvesting on Sumas Mountain does affect surface drainage, creating some concern for the potential impact of increased runoff on adjacent agricultural prairie lands (Ministry of Lands,Parks and Housing, 1985). Complicated interactions occur within units. Agricultural activities, especially the use of manure and fertilizers in raspberry cultivation, are leading to increased nitrogen levels in groundwater. Groundwater levels themselves have also been dropping as a result of agricultural use. Although the problem and interaction can be recognized, there is still considerable uncertainty over the nature of the groundwater resource, and the magnitude and implications of the problem. Even land use patterns can take on a complexity which defies simple analysis. Raspberry production has increased in recent years, a trend that has been accompanied by relocation of pig and poultry operations from more developed areas in Langley to the relative security of the Abbotsford agricultural community. With the increased demand for raspberry land, and the intrusion of other operations onto prime raspberry land even though they are not dependent on the soil itself, some raspberry cultivation has been attempted on marginal land. This creates problems of economic viability and land degradation as the row crops lead to erosion on the steeper slopes (Peters,pers. comm.; Schreier, pers. comm.) With respect to resource values, the study area is important in several sectors. The predominance of agriculture has already been noted. Sixty-five percent of the 97 District of Matsqui is devoted to agriculture (District of Matsqui, 1980, p. 11) while eighty-eight percent of the Sumas Prairie is agricultural (Manning and Eddy, 1978, p.45). In the Abbotsford area, the raspberry industry is particularly significant . Ninety percent of Canada's and thirty percent of North America's raspberries are produced in Southwestern British Columbia (Daubney, 1980). There are also important aggregate deposits in the study area. Major deposits underlie southern Matsqui with less significant concentrations along the western edge of the Town of Abbotsford. The Matsqui area currently produces the largest output in the Lower Mainland with a total of 1,500,000 cubic metres in 1978 with a further 9,000,000 cubic metres under existing leases (Hora and Bosham, 1981, p.18-19). A third important resource can be found in the mountain landscapes. Sumas Mountain is an important area for local commercial harvesting in the winter, when harvesting in more extreme areas is not possible. It is estimated that there is currently 147,000 cubic metres of harvestable timber on Sumas Mountain (Lands, Parks and Housing, 1985, p.ll). Whonnock Industries, the major commercial entity, plans to harvest in the order of twenty hectares or 8,500 - 11,300 cubic metres annually for the foreseeable future (Ministry of Lands, Parks, and Housing, 1985, p.ll). In addition to these productive functions, the available land in the study area also acts as a source of residential and recreational land. In recent times, the economic recession has reduced value conflicts created by these uses simply because of reduced demand. Growth is however forecasted to continue, if at a slower rate than anticipated prior to 1981. In the District of Matsqui, the 1981 population of 42,001 is forecasted to increase to 88,000 by 2001 with Abbotsford growing from 12,745 to 25,600 and Electoral Area A from 306 to 600 (Central Fraser Valley Regional District, 1984). It has also been noted that while there is no immediate shortage of industrial land, there is a trend to locate further east as land prices increase closer to Vancouver (Underwood McClelland Ltd., 1984). 98 This combination of resource values and functional interactions leads to land use value conflicts. These conflicts vary in nature, number and intensity between landscape units. Table 5 provides a summary of the conflicts by landscape unit Decision Environment To deal with the problems and conflicts noted in the previous section, ongoing policy and allocative decisions are made. This is the other half of the equation which determines decision type and structure. The majority of the land in the study area with the exception of Sumas Mountain is privately owned. The ultimate expression of land value is therefore found in the decision made by an individual landowner. This individual will follow a process which includes consideration of economic return, labor required, and even lifestyle desired. The result may not be an optimal use, but it will express the operative value attached by the landowner. Of course, there is a regulatory process which complicates this simple decision environment by introducing societal values. The decision environment on private land is actually one of many interests and diffuse control. This can be seen in the range of concerns reflected in relative policy documents. It can also be seen in Table 6 which summarizes the parties involved, their interests, and their mandates. Some central structure is provided in the area by the Agricultural Land Reserves. This control device dominates rural planning and vests considerable control in the hands of the local councils, the Agricultural Land Commission, and the Provincial Government. On public land, control rests with the Ministry of Lands, Parks and Housing. This control is however partly administrative in that there are other sectoral interests which are accounted for in the planning process and promoted by other agencies. This creates a decision environment dominated by diverse values and controls, but with some control provided by either the Agricultural Land Reserves or the exsistence of Table 5 Land Use Conflicts - CFVRD Matsqui Prairie Agriculture 4 » Forestry - Concern with the Impact of harvesting adjacent land when drainage problems already exist Sumas Mountain Residential 4 •» Forestry - visual impact of logging and the impact of logging trucks - Environmental Impact of logging as opposed to development - servicing costs of development as opposed to revenue from logging Recreation 4 , » Forestry - Compatibility of uses in close proximity - Impact of logqing on recreational resources includinq freshwater fishery Langley Uplands Agriculture 4) 1 > Rural Development - Pressure for exclusion of marginal land from the A.L.R. - Conlicts between farm and non-farm use Agriculture 4 » Aggregate Extraction - Use of agricultural land for gravel pits - Problems in rehabilitating pits to agricultural use Abbotsford Uplands Agriculture 4 > Rural Development - Existing institutional and commercial development attracts further develooment pressure - pressure for exclusion Agriculture 4 i> Aggregate Extraction - Use of agricultural land for gravel pits - Problems in rehabilitating Dits to agricultural use Agriculture 4 » Conservation - Erosion of marginal land - Groundwater pollution Berry Cultivation 4 » Pig and Poultry Operations - Prime raspberry lands being used for operations which do not depend on land quality, possibly forcing cultivation of marginal land. Sumas Prairie Agriculture 4 > Conservation - Potential pollution of the Sumas River system Agriculture 4 + Forestry - Concerns with the potential imoact on surface runoff of harvesting on adjacent forest land when drainage 1s already a problem 100 Table 6 Decision Participants - CFVRD Participant Interest Control Landowner - Profit - Lifestyle - Ability to use land to further personal goals within certain limits - Ability to sell or apply for exclusion from regulatory controls Local Gove rnment - Social Cost/Benefit - Tax Revenue - Political concerns (reelection) - Ability to enact Official Community Plans and implementing by-laws under the Municipal Act R.S.B.C. 1979 C.290 - Approval of Subdivisions, zoning amendments and other variance appl ications . - Ability to screen exclusion applications from the Agricultural Land Reserve. Agricultural Land Commission - Preservation of Agricul tural Land - Ability to designate agricultural land reserves - Ability to render decisions on exclusion aoplications. - Authority obtained from the Land Commission Act RSBC 1979 c.9 Environment and Land Use Commi ttee - Land use and Envi ronmental Quality - The Environment and Land Use Act RSBC 1979 c. 110 allows the ELUC of cabinet to promulgate regulations aimed at protecting the natural environment - This authority can supercede all other public controls in the case of signi ficant issues. Ministry of Lands , Parks and Housing - Administration of Crown Land - Controls initial allocation of Crown Land under the Land Act Ministry of Forests - Forest Management - Controls the issuance of Forest harvesting licences pursuant to the Forest Act Ministry of Transportation and Hi ghways - Regional Transportation - Can establish and alter highways and review development proposals within 800 metres of a highway under the Highway Act RSBC 1979 c. 172 • —• Mi nistry of the Envi ronment - Environmental and Water Resource Management - Can control uses having environmental impact throuqh the Water Act, Wildlife Act, Fisheries Act and Pollution Control Act - Preparing long term environmental management plans under the Environmental Management Act (1981). 101 Crown Land. A detailed descriptive study of the decision process has not been made. The nature of the decision process seems to reflect a number of descriptive observations to some degree. Rational planning principles are in evidence in, for example, the use of capability information in establishing Agricultural Land Reserves. Even so, exclusion decisions have been known to be politically motivated. Although the Ministry of Lands, Parks and Housing may have jurisdictional control over Electoral Area A, the planning process involves negotiation with other government agencies, each with sectoral interests and organizational procedures to consider (VanderHorst, pers. comm.). The Sumas Mountain planning process relies heavily on a study team drawn from several provincial agencies as well as the local governments. 6.3.2 AVAILABLE INFORMATION One of the limitations imposed on this demonstration is the use of existing information. This also represents a reasonable constraint on the first stage of any iterative design exercise. To reiterate, one of the major purposes at this stage is to provide an initial system for decision makers to react to. With this in mind, the next stage in the intelligence process is an initial review of available data and information resources. This review will provide an indication of what type of evaluation system can realistically be considered as well as an initial survey of information requirements and available information. Ideally, a wide range of information should be included in land evaluation. There are numerous sources detailing information requirements. Some refer to agricultural evaluation (FAO, 1983), some to general suitability (McHarg, 1969), some to soils in particular (Griffin, 1977; Nowland, 1976), some to definitions of prime land (Giles and Koeln, 1983), and others to integral land evaluation (LEG, 1978). A comprehensive review of all suggested requirements will not be provided here. These 102 sources were reviewed and compared to what was available in the study area. Generally, the information available at the tactical level in the study area consists of descriptive inventories and level two classifications (capability mapping). There is consistent mapping coverage available at scales of 1:50000 and 1:25000 of surface features like soils and geology. Spatial inventories of such resources as aggregates and signficant features are also readily available. Interpretive mapping includes agricultural and forest capability and erosion potential. Reference should be made to appendix two for a complete description of the available information. 6.3.3 SYSTEM RF.OI JIREMENTS The insights obtained from decision diagnoses and information assessment can now be used to outline system requirements. In this case, a tactical land evaluation system for the study area would have to account for the following realities and requirements: 1. The decision environment is semi-structured in that there are different value conflicts and fragmented authority, but some factors do help to provide a degree of structure at the tactical level. 2. There is a lack of functional or economic models to apply at the tactical level. Information is largely descriptive and in ordinal or interval units, rather than continuous ratio form. 3. Characteristics can be assessed and related to land performance, but alternative outcomes can not be valued due to a lack of quantitative measures (generally produced at the strategic level). 4. Values and requirements change with time, so an absolute measure, if it can be produced, will be open to question and quickly obsolete. The evaluation system will have to deal with these realities while producing a measure which can provide some support for tactical decision making. 103 6.4 DESIGN Having completed the intelligence phase, it is now possible to focus on the design of a particular decision support system. As outlined in Chapter Five, this begins with a choice of design approach options, followed by consideration of the Representation-Operation-Memory Aid-Control Aid framework. A specific system is then designed. In this case, a micro computer tool, LANDPLAN, is at the core of the system and will be the major topic of this section. 6.4.1 DEVELOPMENT OPTIONS Three development options are available. The first is to begin with a complete system in mind. This approach has been dismissed for several reasons. Firstly, a complete system would require external information including provinicial production targets and coordination with adjacent regions which is beyond the resources of this exercise. Secondly, it is difficult to define exactly what a complete system should involve given the current state of the art in land evaluation in the Lower Mainland area. This is particularly true when uncertainty surrounding the economy and future values is considered. A third reason involves the current state of government funding. In a time of restraint, the resources involved in constructing a complete system are difficult to justify without at least some evidence of what can be accomplished. The other extreme involves the construction of a quick-hit system with a single specific decision function. Once again, this option has been eliminated. Land evaluation in the study area is a task which will evolve and change over time. There is also an obvious need to encourage and advance land evaluation beyond its current practice. A quick-hit approach may become rapidly obsolete and also tends to encourage the status quo by demphasizing continued development and future innovation. This leaves a staged approach as the preferred option. The evaluative needs and requirements will evolve as pressures and values in the study area change. This 104 encourages a system which can initially operate within current realities, namely limited resources and no current evaluation system to build on - a working group has been formed to examine agricultural land evaluation in the Lower Mainland but has yet to make substantial progress - but provide a base for the continued development that will be required to meet future demands and create a defensible support system. 6.4.2 R O M C F R A M E W O R K Table 7 lists some of the representations, operations, memory aids and control aids specific to this case study. The table also indicates what provisions have or have not been made for the various activities or concepts in the decision support system designed for this demonstration. 6.4.3 T E C H N I Q U E S E L E C T I O N Considerable time was spent reviewing and comparing various techniques to build a decision support system around. Part of this process is documented by Chapter Four and Figure 8. In the final analysis, an interactive micro-computer tool, LANDPLAN, was selected as the operational core of the DSS. Selection of an interactive micro-computer tool was intentional. The opportunities presented by micro-computing technology have been instrumental in expanding information management capabilities and in allowing the realization of many of the DSS principles. LANDPLAN offers several features which fulfill the system requirements and ROMC needs identified previously. These include: A linear additive model which bases evaluation on assessment of characteristics. Provides some measure of quantitative analysis, but does not require continuous ratio information and can accept a variety of types of information. Supplies a mapping capability, data bank, and ample opportunity for development of additional features. To supplement and complement LANDPLAN, other tools were incorporated into the Table 7 R O M C Framework - CFVRD 105 Decision Maker Uses System Provides Representations - Spatial Perception - Interactions - Land Uses - Landscape uni ts Maps - Rating codes - Land use matrix Operations - Evaluation - Consideration of political values - Linear additive programming model - Pol icy wei ghti ng Memory Aids - Land use data - Recording past results - Data bank Control Aids - For system use - For posing value questions - LANDPLAN Manual - Abi 1 i ty to vary pol icy we i gh ts system. In particular, information analysis tables and compatibility matrices helped to fulfill some ROMC requirements and improve the effectiveness of LANDPLAN. 6.4.4 LANDPLAN The LANDPLAN model involves three major tasks: establishment of a policy set, collection and organization of data, and formulation of suitability rating codes. 106 Policy Set The initial stage of constructing a LANDPLAN model is to identify land uses and the pivotal policies directing these uses. At the tactical level, land uses are of a general nature and in the interests of constructing a manageable set, eight uses were identified as follows: 1. Agricultural - land dependent cropping 2. Agricultural - land independent (pig and poultry farming) 3. Rural development - residential, commercial 4. Mountain residential - low density development in the mountain areas 5. Urban expansion 6. Aggregate extraction 7. Amenity conservation - signficant features, landscape protection 8. Forestry - commercial and private woodlots For each land use, policies on which to base suitability ratings must be identified. There is no 'correct technique' for establishing a policy set, nor is there a means for determining when a set is complete. What a policy set should do is reflect the principle considerations involved in the land use decisions. Policies used in this exercise were extracted from existing planning documents, from personal contact with planners in the area, and from reviewing literature relating land qualities to evaluative measures. Three general types of policies were identified. Exclusion policies provide an absolute rule whereby a use is not allowed if certain conditions exist Preference policies encourage a use if an advantageous land quality or condition exists. There are also policies discouraging uses in areas where a detrimental condition is recorded. After eliminating those policies which could not be expressed within the limitations of LANDPLAN or policies which do not involve an element of land evaluation, twenty-nine policies were chosen for the demonstration. These are listed 107 below: Agriculture 1. Give preference to agriculture on high capability land. 2. Give preference to agriculture on the most productive soils. 3. Discourage agriculture in areas of high erosion potential. 4. Discourage agriculture in areas of the greatest environmental impact 5. Encourage large continuous areas of agricultural production. Agriculture - land independent 6. Encourage agriculture not requiring high quality land to locate on land of marginal capability. 7. Discourage use for poultry and animal husbandry in areas of incompatible land use. Rural Development. 8. Encourage concentration of rural development in areas already developed. 9. Discourage development in flood susceptible areas. 10. Encourage development in areas currently abandoned or in scrubland. 11. Give preference to development in areas most suitable for septic sewage disposal. Mountain Development 12. Encourage development on areas of suitable slope. 13. Give preference to development in areas with existing access. 14. Discourage development in areas of high forestry potential. Urban Development. 15. Give preference to urban development in areas currently reserved for expansion. 16. Encourage urban development in areas adjacent to the existing urban centres and the Trans Canada Highway. 17. Give preference to development of areas with existing major transportation facilities, rail and road. Aggregate Extraction 18. Discourage extraction in areas of incompatible land use. 108 19. Encourage extraction in .areas with high quality deposits. Amenity Conservation 20. Encourage conservation of significant freshwater features. 21. Encourage conservation of significant wildlife habitat 22. Encourage conservation of significant vegetation. 23. Encourage the conservation of significant and vulnerable visual landscapes. 24. Encourage conservation of surface water resources. Forestry 25. Give preference to forestry in areas of high management potential and quality. 26. Discourage forestry in areas of potential downstream impact Exclusion Policies 27. Exclude urban development from areas included in the Agricultural Land Reserve. 28. Exclude rural development from the highest quality agricultural land. 29. Exclude agriculture from areas with steep slopes. There are techniques which can be used to further refine and develop a policy set by working closely with decision makers or experts. This can include such descriptive techniques as content. analysis or through expert judgement using a Delphi technique (see Sinden and Worrell, 1978). Agreement on a policy set can serve as an important agenda, thus allowing negotiation around criteria, rather than values, as is recommended by Fisher and Dry (1981). Data It is difficult to separate data collection from the calculation of suitability ratings as the codes used to calculate the ratings dictate the data items required. In this demonstration, a general data bank was created and subsequently modified as rating codes were developed. 109 The first step in data organization is the creation of spatial or planning units. Two basic options, a grid or a polygon system, were considered. The grid system has its advantages including easy electronic data manipulation and comparison. A polygon system was nevertheless chosen for several reasons. Land use decisions are well enough defined to identify important data items on which to base spatial units. Polygons also allow a range of variable sizes to reflect the requirements of the various landscape units. In very homogeneous areas, large polygons can be used while more detail and smaller polygons are required in diverse areas. In addition, previous exercises using LUPLAN had also used polygon systems. Polygons were formed by first examining each of the landscape units shown on Figure 17. In each unit, the information items which dominated the anticipated land use decisions were identified and used to subdivide the landscape unit into what will be referred to as planning units. For the prairie and upland areas, agricultural land capability was used. In the mountain areas, slope and drainage basin were considered must influential and revealing. An initial set of 98 polygons were used for the LANDPLAN prototype. This initial set was simplified on the basis of preliminary model results which provided virtually identical scores for some adjacent polygons. Any units with consistently identical ratings were combined. Upon completion, there were 81 polygons, each being relatively homogeneous with respect to the information that is most likely to have a bearing on the decision process. These polygons become the planning units which form the spatial base of the LANDPLAN analysis. Figure 18 is an uninterpreted display of the planning unit base map. A total of 48 data items were included in the data bank, although some were not related directly to policies and others are essentially dummy items to provide for future flexibility. Appendix two lists the data items along with their source and means of collection. I l l To provide some insight into the use and importance of each item, a decision support system should provide some form of data analysis. Two measures have been provided in this demonstration. The first, represented by Table 8 simply relates data items to the rating policies, providing an indication of which items have been used in the assessment of each land use policy. From this initial table, a simple information analysis can be completed along the lines of that suggested by McCurdy and Myers (1978). Table 9 indicates the relative importance of each item is the result Table 9 also highlights some assessment measures for the data items. Four assessment measures are significant: 1. Hardness - The reliability of the information and the degree of confidence it can be given (Fabos, 1978). 2. Accuracy - The accuracy of mapped boundaries, uniformity of any regions, and the overlay compatibility. 3. Completeness/Availability - How complete is the data set and what would it require to fill any gaps (Fabos, 1978). 4. Acceptability - This refers to the source of the data. Governmental structure may suggest that certain organizations, and particular data sources be used to ensure later acceptance, even if these sources are not the most refined (Davis and Ive, 1985). These measures are important if an evaluation is to be fully understood. Such analytical guides provide a basis both for refinement of the data bank, and for assessment of reliability. If, for example, a particular item is shown as being very important, but is also considered to have a low degree of accuracy or is particularly subjective and qualitative, then the resulting land use rating must be viewed with this in mind and used very carefully. It may also be desirable to use this analysis in refining and weighting policies. An effort could be made to give the most weight to those policies using the most reliable data items, or distribute weights to emphasize a particular data item. Table 8 LANDPLAN Data Items Policy Data I t e m 1 2 1 4 5 6 7 8 9 10 11 12 13 1« 15 16 17 IB 19 20 21 22 23 24 25 26 1 2 3 Data I tens 1. Planning unit 2 . Landscape unit 3. Polyqon size 4 . Polygon purity class 5. Dominant surficial class 6. Soil management Group 7. Drainage 8. Slone class 9. Slope type 10. Asoect 11. Minimum Altitude 12. Maximum Al ti tude 13. Erosion Potential 14. Climate - Effective Degree Growing Days 15. Climate - Climatic Moisture Deficit 16. Agricultural Capability class 17. Limiting factor 18. Cover - Matrix 19. Cover - patch 20. Cover - Corridor 21 . Forest capability 22. Forest Management Class 23. Agaregate Potential 24. Existing Permits 25. Recreation - Hater 26. Recreation - Development Potential 27. Significant Features - Freshwater 28. Significant Features - Wildlife 29. Significant Features - Vegetation 30. Sigaificant Features - Landscape 31. Significant Features - Historic 32. Diversity 33. Water Resources - Order/Oensity 34. Hater Resources - Reach 35. Hater Resources - Downstream use 36. Hater Resources - Subsurface 37. Flood Potential 38. Transportation - Road 39. Transportation - Rail •O. Distance - From Hwy.l 41 . Distance - From Regional Center •2. Land use - Dominant 43. Land Use - Secondary 44. Additional Land use 45. Adjacent use - Dominant 46. Adjacent use - Secondary 47. P a s t u s e 48. P lanning D e s i g n a t i o n 113 Table 9 L A N D P L A N Data Assessment Item Slope Land use Secondary Land Use Agricultural Capability Cover - Matrix Forest Management Class Surficial Class Soil Management Group Drainage Cover - Patch Importance Score* 6 4 3 3 3 3 2 2 2 2 Assessment 2 3 Water Resources - Order Densi ty Water Resources -Downstream Use Transportation - Road All others 2 2 h m h h m m h h m 1 m m m h m h 1 m 1 m m 1 m m m h m m m m h m m m h h m m 1 1 h m h m m 1 m m h m h m 1. Hardness 2. Accuracy 3. Completeness Avai 1 abi 1 i ty 4 .Acceptabi1ity h- High m- Medium | - Low * This importance score is based on equal weights for all policies. The score therefore has the form: V(Ij) _CDi where V(Ij)= Score Di = Use in a policy For a variable weighting scheme, the score could be calculated according to the policy weight, thereby making the formula: V(Ij)= 1 Wi*Di where Wi= the weight for policy i (McCurdy and Meyers, 1978) 114 Rating Codes Suitability ratings are calculated by expressing each policy in basic computer code. Ideally, the development of each policy code should include consultation with a wide range of experts as well as the government agencies responsible for implementing the policy in question. In this manner, development of policy codes could serve to survey, integrate and focus the available knowledge and information resources in each field. With twenty six policies, such a rigorous approach was not possible in this exercise. The codes involved in this model were derived from existing literature and informed judgement Chapter Four explained that the codes in this type of exercise are dependant on the state of scientific knowledge and how well this knowledge can be converted in to basic code. The most advanced code would involve a formula providing a continuous function relating the data item to performance or limitation with a value between 0 and 1. This is unfortunately, rarely available. The codes in this exercise are based largely on discrete rating schemes wherein an information item is simply assigned a value between 0 and 1. Wherever possible, this value is based on a known relationship and thus, the difference between ratings reflect the degree of performance or limitation. For example, soil management groups are rated on the basis of yield data. In other cases, there is no direct measure available. There were not empirically tested values for the impact of conflicting land uses in an area or for the impact of proximity to the regional urban centre. Subjective judgements were required to provide a relative measure in these cases. A description of each policy code is given in Appendix Three. To complete an evaluation, each policy must be given a weight The model then uses a form of linear addition to maximize the weighted sum of suitabilities and provide a preferred use and suitability rating for each planning unit An algebriac description of this program can be found in Appendix Four. 115 To test the model, each policy was given the same weight and the result is shown on Figure 19. Appendix Four contains the actual evaluative output and should be referred to at this time. Two measures are produced by the program. Each land use is given a suitability rating for each planning unit For example, forestry has a rating of .1 in planning unit 72. These ratings provide a relative ordering with the land use with the highest score considered the preferred use. When the policy weights are all equal, the maximum score is .22 and the minimum is 0 indicating no rating while the lowest rating with some measurable suitability is .5. LANDPLAN also produces an exercise summary with totals for each land use. Figure 19 becomes the reference point for future applications and development It is difficult to assess the correctness of this base model at this stage. The subject area v/as carefully reviewed and the base map seems intuitively accurate and indicative of current policy and land use. No major aberrations are apparent Continued use and observation would lead to further refinement but considerable development has already gone into the current version. It was judged to be sufficient for the purposes of this demonstration. 6.4.5 INTERPRETIVE MATRIX To assist in use and interpretation of LANDPLAN, a compatibility matrix has been constructed. This matrix has been included as Figure 20. By examining the matrix in conjunction with the suitability ratings for each planning unit conflicts can be identified to guide policy application. This will be demonstrated in section 6.5. 6.4.6 SYSTEM SUMMARY A simple decision support system has now been created. It consists of LANDPLAN, an associated data bank, data assessment and policy set, and a 117 2 3 5 6 7 Land Uses 1. Land Dependent Agriculture 2. Land Independent Agriculture 3. Rural Development 4. Mountain Development 5. Urban Development 6. Aggregate Extraction 7. Amenity Conservation 8. Forestry Symbol s Compatible P o t e n t i a l l y ComDatible with special o o l i c i e s or design standards Incompatible FIGURE 20 - Land Use Compatibility Matrix compatibility matrix to assist in interpretation. The decision support system concept however applies as much to the way in which a system is used as it does to the design of the system itself. This is explored in the following section. 6.5 CHOICE The LANDPLAN model completes the choice function by generating a preferred use with a linear additive model. Notwithstanding this imposed operation, the decision maker has some personal control through the specification of policy weights. This control aid is an integral component of the evaluation. Three applications using this control aid to support decision making are demonstrated in this section. Additional choice functions based on other operations would presumably be developed as iterative design progresses. Future design possibilities and needs are also discussed. Prior to applying the system, it is important to review the assumptions and limitations inherent in the evaluative operation. 6.5.1 SYSTEM ANALYSTS The Decision Support System created includes an operation which produces suitability scores and indicates a preferred use. Several assumptions and limitations must be understood if this choice mechanism is to be properly applied. The ratings do include some ordinal and qualitative data. Some ratings codes are based on cryptic judgements. Average land management with no major improvements (access, drainage) is assumed. The final scores are not absolute values or meaningful in isolation. It is not therefore a system which will supply a single best answer. Even if an exhaustive effort is put into developing the rating codes or factual content, questions relating to the policy set and weights or value content are likely to persist The experience with LUPLAN in Australia has been that use of the system to produce a final map has not entirely been successful due to suspicion and a lack of acceptance 119 (Davis, pres. comm.). As a Decision Support System however, it can be a very useful and effective land evaluation tool. The next section reviews three applications which reflect the limitations of the system but still provide useful choice functions. 6.5.2 APPLICATIONS As noted earlier, the ratings indicate the suitability of a use on the basis of the rating codes and policy weights. Although they do not represent absolute values, they do provide useful information of a relative and comparative type within each planning unit This represents the first of the potential applications. In some cases, one land use is preferred by a wide margin. Unit 63, an area of high quality agricultural land in the Sumas Prairie, registers a rating of .17 for land dependent agriculture, .07 for land independent agriculture, and substantially less for any other uses. In unit 80, an area of high management potential for forestry, the rating for forestry is more than double any other use. Even the most biased weighting schemes continue to produce the same preferred use. The general magnitude of the rating can also be meaningful. Unit 64 contains the main channel of the Sumas River, a wetland and lake area, and significant vegetation and wildlife habitat The preferred use, conservation, has a relatively high rating of .22. At the opposite extreme, unit 68 has a preferred use rating of only .05. Upon examination, unit 68 turns out to be an area in the Sumas Mountain Landscape unit of steeply sloping exposed bedrock, a marginal area for any form of use. These interpretations are largely intuitively obvious. Of more interest is the identification of those marginal areas which will be the subject of land use value conflicts. In many cases there is only a minimal difference between the ratings for the first two or three land uses. Five such cases are listed below: Unit 72 -Forestry(.l),Conservation(.09), Residential(.08) 120 Unit 8 - Development (.11), Conservation (.1) Unit 29 - Rural development (.13), Agriculture (.1) Unit 70 - Conservation (.11), Residential (.1) Unit 35 - Conservation (.16), Agriculture (.11) The compatibility matrix becomes important in these situations. Two cases should be recognized. In the first, the land uses which are closely related are in fact shown on the matrix as compatible or complementary. These are potentially harmonious mixed use environments. Although special policies, practices, design guidelines may be required to coordinate multiple use, no fundamental value conflicts exist Agriculture and conservation provide a case in point Special incentives or strict enforcement of codes of practice may be required to protect significant features or prevent land degradation, but the uses may actually generate constructive tension, not destructive conflict On the other hand, there are those areas where the ratings are, a clear indication of fundamental conflict For example, there is little difference between the top three uses in unit 72 which is on Sumas Mountian in the fringe of current development Both of these cases can be summarized directly on the compatibility matrix by indicating scores of incompatible uses in the relevant cell. Figure 21 shows unit 6 (no conflict) and unit 72 (significant conflict). An additional measure can be provided by summing the differences between incompatible uses to produce a conflict score. A high score would indicate a large difference in the rating between incompatible uses, thereby indicating that there is clearly a dominate use in potential conflict situations. Where there is only a slight difference and a high potential for conflict, as in unit 72, the score is much lower. A more formal analysis can be made of marginal planning units by exercising the control function and varying the weights. The base weights can be systematically changed and shifts in preferred use recorded. A sensitivity analysis can then be 121 L a n d Uses 1. L a n d Dependent A g r i c u l t u r e 2 . L a n d I n d e p e n d e n t A g r i c u l t u r e 3 . R u r a l D e v e l o p m e n t 4 . M o u n t a i n Development 5 . Urban D e v e l o p m e n t 6 . A g g r e g a t e E x t r a c t i o n 7 . A m e n i t y C o n s e r v a t i o n 8 . F o r e s t r y Syefcols - Compatible P o t e n t i a l l y C o m o a t l b l e w i t h s p e c i a l o o l l c l e s o r d e s i g n s t a n d a r d s - I n c o m p a t i b l e Unit 72 •"/.o» C o n f l i c t Score: •01+.02+.01 3 = .013 Unit 6 1 t » 4 t £ , •*/.» .07. C o n f l i c t Score: .06 + .01 ~ 2 = .035 FIGURE 21 - Use of the Land Use Compatibility Matrix 122 performed. This can be qualitative as the changes in individual polygons are simply observed, or more quantitative. A quantitative assessment could be made by either recording the sensitivity of individual polygons to shifts, or an analysis of the impact of the changes in weights of the various policies on the complete set of polygons. Each type was completed in this demonstration but only the second is documented. The sensitivity of individual planning units is best judged qualitatively but the sensitivity analysis is required to examine the policies. Table 10 indicates the sensitivity of the model to changes in pairs of policies representing each land use. It is interesting to note that the evaluation is particularly sensitive to the rural development policies. This is not surprising in that many of the units consist of marginal class four and five agricultural land, and a minimal value shift is sufficient to register rural development as the preferred use. In the event that political values begin to shift away from agricultural land preservation, the defence of agricultural land will require firm resolve and a clear rationale. As a second application, the Decision Support System can be used to illustrate the implications of land use decisions to a decision maker. This involves beginning with a decision to change a land use, then varying the weights to make the model reflect the decision, and finally, examining the exercise totals to judge the impact of applying the same values throughout the complete area. In effect, the model is operated in reverse. In the Central Fraser Valley Regional District, this need may arise when dealing with exclusion applications from the Agricultural Land Reserve. Two applications for development of rural residential lots on land typical of , for example, units 14 and 20 may be made. A decision maker or makers may be moving towards exclusion and approval of the lots. Before approval, the weights for rural development are increased until each unit shows rural development as the preferred use. The aggregate results are shown in Table 11. After this simple but effective demonstration, a decision maker may 123 Table 10 Sensitivity Analysis Policy qrouD Number of Area Probability of Percentage of Polygons Changed Change Area Changed Changed Agriculture Policies 1,3 4.4* increase* 2 3 04 .03 1 05 8.5°* 4 8 09 .06 2. 79 12.2% 13 30 94 .20 10 66 (does not include Sumas Units Conservation Policies 22,23 4.4°* 9 41 26 .11 11 07 8.52 17 107 52 .21 28 85 12.20; 20 122 77 .25 32 95 Rural Development Policies 8,11 4.42 6 11 79 .09 4 06 8.5? 12 45 79 .19 15 77 12.22 22 93 55 .34 32 22 Aggregates Policies 18,19 4.4% 1 2 0 .02 67 8.52 2 9 0 .03 3 03 12.22 9 29 44 .14 9 90 Forestry Policies 25,26 4.42 5 15 .40 .29 18 71 8.52 5 15 40 .29 18 71 12.22 7 30 27 .41 36 77 (Inculdes only M ountain Units) Mountian Resi dential 4.42 4 15 75 .24 19 13 8.52 6 26 .56 .35 32 26 12.22 6 26 56 .35 32 26 * Increase refers to the increased potential influence of that policy in the complete policy set 124 Table 11 Impact of Value Scenarios Development Weights Equal Weights Planning Unit 20 Planni ng Unit 14 Use Area Area% Area . Area% Area Area% 1 • 234.18 62.84 • 227.54 61.06 • 181.46 48.69 ?. • 9.07 2.43 9.07 2.43 9.07 2.43 3 • 31.58 8.47 • 45.22 12.13 • 91.30 24.50 4 15.40 4.13 15.40 4.13 15.40 4.13 5 0.0 0.0 0.0 0.0 0.0 0.0 6 0.0 0.0 0.0 0.0 0.0 0.0 7 37.77 10.14 . 30.77 8.26 30.77 8.26 8 44.65 11.98 44.65 11.98 44.65 11.98 Total 372.65 372.65 372.65 reconsider exclusion in unit 14. Finally, the Decision Support System can be used as an educational device meant to improve the knowledge of those involved in the decision process. Construction of the system itself is very educational. It provides an opportunity to carefully examine policies, what occurs when policy must be translated into decisions, and the information required in the translation. Both of the previous two applications can be simulated by an individual decision maker or bargainer to inform him/her of conflict areas and allow the analysis of 'What i f questions. It is also possible to speculate on the values of others involved in a bargaining process. Fisher and Ury (1981) emphasize the importance of focussing on interests, not positions, as well as the establishment of objective criteria. In this case, the policies can represent an objective criteria while the weights reflect the various 125 interests. For example, weights were developed reflecting the likely interests of an agriculturalist from the Ministry of Agriculture and Food, a conservationist from a local interest group, and a developer. Figures 21,22 and 23 are the result A participant can then examine the maps, and in conjunction with the compatibility matrix, begin to understand how his interests interact with others involved in the process in placing a value on land. If the DSS is introduced to all those involved in a negotiation or bargaining process, discussions will focus on areas where intersts result in genuine conflict, rather than areas where positions create general arguments even though a particular use is preferred regardless of the weights given to the criteria. These scenarios can also be used to demonstrate the use of data analysis. Table 12 indicates the importance scores for each scenario. As could be predicted, the importance scores reflect the emphasis of the scenario weights. Production values make more use of forestry and agricultural data while land use, transportation, and planning designation become more important in the scenario using development values. 6.5.3 ITERATIVE DESIGN As was emphasized at the beginning of the chapter, this has been meant only as an initial system. This was made explicit in the choice of a staged approach to development To direct further system development, suggestions regarding both the support required at the other decision levels (middle-out concept) and refinements or additions to the system supporting tactical decision making. The DSS provides valuable guidance for land evaluation at the operational level. This is accomplished partly by identifying areas of conflict An exhaustive land evaluation study on a planning unit where there is little question about the preferred use is not an efficient distribution of resources. Other units, like units 20, 72 and 18, will require more site specific evaluations prior to operational decisions. FIGURE 23 - Map of the Development Value Bias Scenario L E G E N D I I Agrfcnltiire ^Lij Rural D c i d o p u i u l H Urban D C T C I O P D X M EjctracthM tjV-l Conserratfoo Forestry • N k m . 129 Table 12 Data Assessment - Value Scenarios Agri culture Development Conservation Data Item , Score Data Item Score Data Item Score 22. Forest man. 8. Slope 42. Land Use 16. Agric. Cap. 33. Water Resources 35. Water Resources 22.0 17.4 14.0 11.0 9.6 9.6 8. Slope 42. Land Use 1 43. Land Use 2 22. Forest man. 44. Land Use 3 1. Poilygon 18. Cover 38. Transport 20.5 20.2 15.5 12.8 11.0 11.0 9.5 9.1 8. Slope 27. Significant 28 2 9' Features 22. Forest Man. 42. Land Use 1 43. Land Use 2 33. Water Resources 19.2 13.2 13.2 12.8 9.9 8.8 43. Land Use 18. Cover 6. Soil Group All others 9.0 9.0 . 8.0 <8 48. Planning Des. All Others 9.1 <9 35. Water Resources 5. Surficial Unit 7. Drainage' All others 8.8 7.7 7.7 <7 Data requirements are also identified. By organizing the data or information around policy statements, those items of particular importance are highlighted. If the data analysis has shown key deficiencies in data which is instrumental in the calculation of ratings for a particular policy, then improvement in this data item should be a prerequisite for any operational application of the policy. For example, the policy relating to the environmental impact of agriculture is clearly incomplete due to the lack of reliable groundwater profiles. 130 The tactical analysis can also be used as a checklist for site specific work. Slope, for example, is an important item in many policies and should be determined in the initial stages of operational decision making. The Decision Support System is hardly necessary to identify slope as a critical factor, but other data items may be less obvious. The Decision Support System provides less clear direction for strategic evaluation. Using the tactical DSS can provide an indication of the type of direction required from the strategic level. As a start, some indication is needed of what amount of land is sufficient for the different uses. It would then be possible to order and place a value on outcomes as represented by the aggregate results of a value scenario. If outcomes could be evaluated, other analytical tools would become available. The Land Evaluation Model supplies this information by developing supply and demand scenarios. Considering the level of uncertainty involved in strategic planning, this seems to be a reasonable approach. Some carefully developed scenarios are needed, even if there is some doubt as to their economic certainty. As a tactical Decision Support System, the system has several significant advantages, especially when the decision criteria introduced in Chapter Three are considered. For those involved in the building of the LANDPLAN model, an excellent opportunity to explore the problem is created. LANDPLAN forces those involved to explore a wide range of land use and policy options. By programming policy codes, it is necessary to explicitly express and evaluate the factors involved in policy satisfaction. Keen and Morton (1979) argue that programming a problem is one of the most effective learning techniques and with the simple format inherent in LANDPLAN, wide participation in model construction should be possible. Finally, the use of the DSS provides insight into the nature and extent of potential problem conflicts and the sensitivity of the policy set 131 The DSS is also very flexible. It is possible to test an infinite variety of value scenarios in an explicit fashion. The rating codes, policy sets and land uses are also easy to modify to provide different types of analysis or more specific policy considerations. For example, a specific system could be formulated to perform a single use evaluation using a more detailed land use and policy set for agriculture. One of the greatest strengths of the system is in its use of data. By creating the policy set, collection of data becomes a focussed task. It also organizes the data and provides a simple means for conversion to higher level information. The system can be critisized on several general points but these are tempered by some positive replies as shown below. 1. The information base does include some qualitative judgements as well as items assessed as inadequate in some respects But The system produces a reasonable result which reflects current use, planning designations and land capability. It also acts as an initial tool to identify data gaps. 2. The policy set is not comprehensive But It does include the main points and represents a trade-off between a manageable program and comprehensiveness. 3. The system does not provide a single best choice which is likely to be defensible in the face of vigorous opposition. But It does initiate an iterative process, prompts reaction and learning, and can be used effectively as a decision support system in a number of ways. There are several areas in which improvements would be desirable. Technically, the data management capabilities are deficient One of the limitations of this demonstration was the lack of programming expertise available so a major attempt at improvement was not undertaken. To function as a proper DSS, some additional capabilities should be added including: 132 A data management module to allow direct extraction from and manipulation of the data bank. Automated mapping functions (LUPLAN does contain a mapping capability). A module for recording and updating results automatically so several scenarios can be stored and compared. Sensitivity analysis could then become an automated option. These capabilities would not be difficult to incorporate by a programmer who could integrate existing data management, computer graphics, and electronic spread sheet functions into the existing LANDPLAN frame. There are two major conceptual weaknesses which need to be addressed at the tactical level. The first is the lack of economic analysis while the second is the lack of spatial, extrinsic analysis between units. Economic analysis could be improved if better information and data was available to relate land or spatial charcteristics with economic return. Other economic analysis would require either supply and demand information from the strategic level, or another evaluative option. Strategic supply and demand information has been discussed previously. Additional evaluative options would involve estimates of economic feasibility which could then be applied at the tactical level by varying policy weights (greater weight for more profitable activities) and evaluation of outcomes. A simple linear programming model could, for example, be incorporated to evaluate the preferred uses from an economic perspective. A second desirable option would be the ability to compare and aggregate units. This is accomplished to a certain degree by consideration of adjacent or downstream use in the LANDPLAN model. Improved capabilities are however necessary to allow comparison of nonadjacent units. For instance, a watershed must be examined from the source area to the mouth if potential impacts are to be incorporated into the evaluation. Contiguous parcels of a certain size may also be required for economic feasibility but the ability to aggregate and keep spatially organized totals does not exist 133 Finally, research is required to develop models on which to base rating codes. Continuous functions relating land performance to land characterisics or capability classes are particularly deficient In summary, the Decision Support System is just an initial step in an iterative design process. It provides some useful decision support but is deficient in some areas. Guidance has however been obtained to direct development of a more complete system incorporating additional options at the tactical level as well as evaluation required at the strategic and operational levels. Chapter 7 CONCLUSIONS The effective use of information is of critical importance in the management and use of available land resources. Land evaluation is a fundamental process in the conversion of environmental data into useful information. In the context of land evaluation, effective use depends not only on continued refinement to produce more specific and defensible technical information, but also on the integration of such information into the decision process. This thesis presents a conceptual model which involves a cyclical process of information generation and use, with adaptive design resulting from the interaction between technical components and decision making components. The model has not however been functioning effectively due to the inadequate integration of these two components. Integration of information into the decision process requires an understanding of several concepts important in the decision process. These include decision structure, decision level, applicable descriptive models, and decision analysis. Each of these concepts have implications for the use of existing information generated by land evaluation and the design of new land evaluation systems. These implications have been demonstrated by examining three current approaches to land evaluation. Several current land evaluation systems have begun to incorporate interactive, flexible evaluative models. These represent a logical progression towards Decision Support Systems, which are presented as a decision centered approach to the design of land evaluation systems. A case study was used to demonstrate the design of an intial DSS using existing information. The DSS was applied to land evaluation in the Central Fraser Valley Regional District and was also used to provide guidance for continued development in an iterative design cycle. 134 135 Out of the above examination, several important conclusions emerge as listed below: Decision concepts should be used in applying and designing land evaluation systems in an effort to avoid information degeneration. An understanding of these concepts will encourage use of existing information in its proper decision context and with an awareness of its potential role and limitations. Decision Support Systems supply an approach which emphasizes the use of information in the decision process. Even if a single best answer is not really produced, as is often the case if systems are examined closely, a DSS approach allows development of useful choice applications. Iterative design should include continued development of several capabilities: a. Development of supply and demand scenarios at the strategic level. b. Specific improvement of selected information for use at the operational level. c. Continuous functions for developing rating codes. d. Development of evaluative operations for economic feasibility and spatial aggregation. Decision Support System technology is becoming increasingly common in the land resource planning field. The availability of micro-computing systems facilitates widespread dissemination of decision support system software and packages. As more and more decision makers and planners take advantage of these opportunities, Decision Support Systems will become an integral part of information management and land evaluation. Such a scenario leads to a decision exercise supported by interactive decision aids throughout the process, all of which are linked for a continuous flow of information from strategic to operational levels. When this becomes reality, the linkage between information and decision making so critical to the functioning of the conceptual model should improve to the point where we would no longer have to ask: Where is the wisdom we have lost in knowledge ? Where is the knowledge we have lost in information ? T.S. Eliot REFERENCES Allison, Graham T. 1971. Essence of Decision: Explaining the Cuban Missile Crisis. Little, Brown and Company. Boston, Massachusetts. Alston, Richard M. and David Freeman. 1975. The natural resources decision-maker as political and economic man: toward a synthesis. Journal of Environmental Management Vol.3: 176-183. Anderson, Lee, Jerry Mosier and Goeffrey Chandler. 1979. Visual Absorption Capacity, in Our National Landscape: Proceedings of a Conference on Applied Techniques for Analysis and Management of the Visual Resource. April 23-25, Pacific Southwest Forest and Range Experiment Station, Berkley. Armstrong, John E. 1984. Environmental and engineering applications of the surficial geology of the Fraser Lowland, British Columbia. Geological Survey of Canada. Paper 83-23. Audet R. and A. Le Henaff. 1983. Land planning framework of Canada: An overview. Working Paper no. 28. Lands Directorate, Environment Canada, Ottawa. Baird, I.R. 1981. The application of SIRO-PLAN to rural planning: Recent research and development activity of the Land Use Planning Group. Technical Memo 81/31. CSIRO, Canberra. Bakus, Gerald, W.G. Stillwell, Susan Latter, and .Margaret Wallerstein. 1982. Decision making: With applications for environmental management Environmental Management 6(6): 493-504. Bastedo, James D. and John A. Theberge. 1983. An appraisal of inter-disciplinary resource surveys (Ecological Land Classification). Landscape Planning. 1983(10): 317-334. Beek, K.J. and J. Bennema. 1974. Land evaluation for agricultural land use planning: An ecological method, in Approaches to land classification. Publication 22, FAO, Rome. Beek, K.J.. 1978. Land Evaluation For Agricultural Development Pub. No. 23. International Institute for Land Reclamation and Improvement Wageningen, The Netherlands. Bennet, John L. (ed.). 1983. Building Decision Support Systems. Addison-Wesley Pub. Co., Reading, Massachusetts. Berger, J. 1976. The Hazelton Ecological Land Planning Study, human adaptions in a rural and urban environment: The basis for possible future land use patterns. Landscape Planning. 1976(3): 303-335. Briggs, D.J. and LFrance. 1981. Classifying landscapes and habitats for regional environmental planning. Journal of Environmental Management 17(1983): 249-261. Brinkman, R. and AJ. Smyth (eds.). 1973. Land evaluation for rural purposes: Summary of an expert consultation, Wageningen, The Netherlands. October 6-12, 136 137 1972. International Institute for Land Reclamation and Improvement British Columbia Ministry of Environment, and Ministry of Agriculture and Food. 1983. Land capability classification for agriculture in British Columbia. MOE Manual 1. Victoria. British Columbia Ministry of Lands, Parks and Housing. 1984. Sumas Mountain Crown Land Plan, unreleased draft Carpenter, Richard A. 1980. Using ecological knowledge for development rating. Environmental Management 4(1): 13-20. Clough, J.Donald. 1984. Decisions in public and private sectors: Theories, practices and processes. Prentice-Hall Inc., Englewood Cliffs, New Jersey. Central Fraser Valley Regional District 1984. Growth Report (draft). Central Fraser Regional District Dewdney Alouette Regional District, Greater Vancouver Regional District Regional District of Fraser-Cheam. 1980. Plan for the Lower Mainland of British Columbia. Vancouver, British Columbia. Coastal Zone Resource Subcommittee. 1978. The management of coastal resources in British Columbia. Vols. 1 and 2. B.C. Land Resources Steering Committee. Cocklin Chris, S.C. Lonergan and B.Smit 1984. A goal programming model to assess options in resource use for renewable energy: The case of poplar development in Eastern Ontario. Paper presented to the 15th annual Pittsburgh Conference on Modelling and Simulation. Pittsburgh, April 19-20, 1984. Cocks, K.D. and M.P. Austin. 1979. A policy-oriented approach to land use planning: A pre-print Tech memo. 79/14. CSIRO. Canberra. Cocks K.D. et al. 1983. SIRO-PLAN and LUPLAN: an Australian approach to land use planning: 1- The SIRO-PLAN land use planning method. Environment and Planning B:Planning and Design. 1983(10): 331-345. Coombs, Donald B. and J.Thie. 1978. The Canada Land Inventory System, in Planning the uses and management of land. M,T.Beatty, G.W.Petersen, and L.D.Swindale (eds.). Soil Science Society of America. Daubney, Hugh. 1980. Red Raspberry cultivar development in B.C. with special reference to pest response and germplasm exploitation. Acta Horticulturae. 1980(12): 59- 67. Davidson, Donald A. 1980. Topics in applied geography: Soils and land use planning. Longman Group Ltd., London. Davis, J.R. and P.V. Greenhalgh. 1980. Policy analysis: A test in the Hunter Valley. Technical Memo. 80/21. CSIRO. Canberra. Davis, J.R. and J.R. Ive. 1985. Dongog Shire Environmental Plan: An application of SIRO-PLAN to local government planning. Divisional Report 85/2. CSIRO. Canberra. Davis, Richard. Research Scientist CSIRO. Personal interview. Guelph, Ontario, 18 July, 138 1985. Davis, J.R., P.M.Nanninga, and K.D.Cocks. 1985. The usefulness of computer aids that capture expert knowledge about land management A paper presented to the International Conference on Management of Rural Resources: Problems and Policies. Guelph, Ontario, July 1985. Dearden, Philip. 1978. The ecological component in land use planning: A conceptual framework. Biological Conservation 1978(14): 167-179. Dent, David, and Anthony Young. 1981. Soil survey and land evaluation. George Allen and Unwin. 1981. DiCastri, R, F.W.G. Baker, and M.Hadley, eds. 1984. Ecology in practice: The social response. UNESCO and Tycooly International Publishers. Paris. District of Matsqui. 1980. Matsqui Official Community Plan. Dorcey, Anthony H.J., and Kenneth J. Hall. 1981. Setting ecological research priorities for management: The art of the impossible in the Fraser Estuary. Westwater Research Centre, University of British Columbia, Vancouver. Dorcey, Anthony H.J.. 1983. Coastal Management as a Bargaining Process. Coastal Zone Management Journal. 11(1-2): 13-39. Dorney, R.S.. 1976. Biophysical and cultural-historic land classification for Canadian urban and urbanizing land, in Ecological(biophysical) land classification in urban areas. E.B. Wiken and G.R. Ironside (eds.). ELCS No.3. Environment Canada, Ottawa. Dorney, R.S., and D.W. Hoffman . 1979. Development of landscape planning concepts and management strategies for an urbanizing agricultural region. Landscape Planning. 1979(6): 151-177. Dyer, Allan, M. Brklacich, D. Bond, and B.SmiL Recent advancements to and applications of a land evaluation model: A decision support system for land use analysts and decision-makers. 1983: in Papers from the Annual Conference of the Urban and Regional Information Associations: Decision Support Systems for Policy and Management R.Schmitt and H.Smolin (eds). August 14-17, Atlanta, Georgia. Ein-Dor, Phillip, and Eli Segev. 1981. A paradigm for management information systems. Praeger Publishers, New York. Epp, P.F.. 1984. Soil constraints for septic tank effluent absorption. MOE Manual 5. B.C. Ministry of Environment Kelowna, British Columbia. Evans, Nan, Marc J. Hershmann, George Blomberg, and William Lawrence. 1980. The search for predictability: Planning and conflict resolution in Grays Harbor Washington. University of Washington, Seattle. Fabos, Julian Gy, CM. Greene, and S.A. Joyner Jr. 1978. The METLAND landscape planning process: Composite landscape assessment, alternative plan formulation and plan evaluation; Part 3 of the Metropolitan Landscape Planning Model. Research Bulletin No. 653. University of Massachusetts. 139 Fabos, Julian Gy. 1979. Planning the total landscape: A guide to intelligent land use. Westview Press, Boulder, Colorado. Fisher, Roger, and William Ury. 1981. Getting to yes: Negotiating agreement without giving in. Penguin Books Ltd., Middlesex, England. Fiaherty, Mark, and Barry Smit. 1982. An assessment of land classification techniques in planning for agricultural land use. Journal of Environmental Management 1982(15): 323-332. Food and Agriculture Organization of the United Nations. (FAO). 1974. Approaches to Land Classiication. Soils Bulletin 22. FAO, Rome. FAO. 1977. A framework for land evaluation. International Institute for Land Reclamation and Improvement Wageningen, The Netherlands. FAO. 1983. Guidelines: Land evaluation for rainfed agriculture. Soils Bulletin 52. FAO, Rome. Forman, Richard T. 1981. Interaction among landscape elements: A core of landscape ecology. Proceedings of the International Congress Netherlands Society of Landscape Ecology. Veldhoven, The Netherlands: 35-47. Friend, J.K.. 1983. Reflections on rationality in strategic choice. Environment and Planning B: Planning and Design. 1983(10): 63-69. Ghiselin, Jon. 1982. Reaching environmental decisions: Making subjective and objective judegments. Environmental Management 6(2): 103-108. Giles, Robert H., and Gregory T. Koeln. 1983. Land and cropland primeness: Concepts and methods of determination. Environmental Management 7(2): 129-142. Giliomee, J.H. 1977. Ecological planning: Method and evaluation. Landscape Planning. 1977(4): 185-191. Girt, J.L. (project director). 1976. The evaluation of alternative methodologies for rural land evaluation. Pub No.82. Centre for Resources Development, University of Guelph, Guelph, Ontario. Gold, Andrew J. 1974. Design with nature: A critique. AIP Journal, July 1974: 284-286. Goodchild, Michael F. 1976. The determinants of land capability. Occassional paper No.7. Lands Directorate, Environment Canada, Ottawa. Goulter, Wenzel, and LHopkins. 1983. Watershed land-use planning under uncertainty. Environment and Planning A. 1983(15): 987-992. Government of Ontario. 1978. Food land guidelines: A policy statement of the Government of Ontario on planning for agriculture. Goverment of Ontario, Toronto. Griffin. D.W.. 1977. A technical guide for determining land use suitability. Publication 47. College of Agriculture, University of Illinois at Urbana-Champaign. 140 Healey, Patsy. 1983. Rational method as a mode of policy formation and implementation in land use policy. Environment and Planning B: Planning and Design. 1983(10): 19-39. Henderson, Jan A., Lawrence S. Davis, and Edward Ryberg. 1981. ECOSYM: A classification system for wildland. resource management Department of Forestry and Recreation, Utah State University, Logan, Utah. Hickling, Allen. 1975. Aids to strategic choice. Centre for Continuing Education, University of British Columbia, Vancouver. Hill, Morris. 1968. A goals achievement matrix for evaluating alternative plans. Journal of the American Institute of Planners. Jan, 1968: 19-29. Hills, G. Angus. 1976. An integrated, iterative holistic approach to ecosystem classification, in Ecological(biophysical) land classification in Canada. ECLS no.l, J.Thie and G.Ironside eds. Lands Directorate, Environment Canada, Ottawa. Hoffman, Douglas W.. 1971. The assessment of soil productivity for agriculture. ARDA report no.4. Holling, CS. (ed.). 1980. Adaptive environmental assessment and management International series on applied systems analysis no.3. John Wiley and Sons, New York. Hopkins, Lewis D. 1977. Methods for generating land suitability maps: A comparative evaluation. Journal of the American Institute of Planners. 43(4): 386-400. Hopkins, L.D. 1974. Plan projection, policy - mathematical programming and planning theory. Environment and Planning A. 1974(6): 419-430. Hora, Z.D., and F.C.Basham. 1981. Sand and gravel study 1980 British Columbia Lower Mainland. British Columbia Mineral Resources Branch, Victoria. Huber, George P.. 1982. Decision Support Systems: Their present nature and future applications, in Decision making: An interdisciplinary inquiry. G.R. Ugson and D.N. Braunstein eds. Kent Publishing Co., Boston, pp. 249-263. Hudson, Barclay M.. 1979. Comparison of current planning theories: Counterparts and contradictions. Journal of the American Planning Association. 45(4): 387-399. Hurst Gerald, D.N.Ness, T.J. Gambino, and T.H. Johnson. 1983. Growing DSS: A flexible, evolutionary approach, in Building decision support systems. J.LBennet (ed.). Addison-Wesley Co., Reading, pp. 111-133. Inbar, Michael. 1979. Routine decision-making:The future of bureaucracy. Sage Publications, London. Ive, J.R. and K.D. Cocks. 1983. SIRO-PLAN and LUPLAN: An Australian approach to land use planning; 2. The LUPLAN land use planning package. Environment and Planning B: Planning and Design. 1983(10): 347-355. Ive. J.R. and K.D. Cocks. 1984. SIRO-PLAN and LUPLAN: Notes for local government clients. Tech memo. 84/6. CSIRO, Canberra. 141 Johnson, A.H., J.Berger, and I.L. McHarg. 1979. A case study in ecological planning: The Woodlands, Texas, in Planning the Uses and Management of Land. Keen, Peter G., and Michael S. Morton. 1978. Decision support systems: An organizational perspective. Addison-Wesley Publishing Co., Reading, Massachusetts. Keen, Peter G. and T.J. Gambino. Building a Decision support system: The mythical man-month revisited, in Building decision support systems. J.L.Bennet (ed,). Addison-Wesley Pub. Co., Reading, pp. 133-173. Kessel, Stephen R., Roger B. Good, and Angas Hopkins. 1984. Environmental Management 8(3): 251-270. Kunruether, Howard. 1983. A mulu-attribute multi-party model of choice: Descriptive and prescriptive considerations, in Analysing and aiding decision processes. P.Humphreys, O Svenson, and A.Vari eds., North Holland Publishing Company, Amsterdam, pp. 69-91. Land Evaluation Group (LEG). 1978. Summary report: Methodology study and development of a data base for rural land evaluation in Ontario. Pub. No. 92. Centre for Resources Development, University of Guelph, Guelph, Ontario. Land Evaluation Group. 1983. Av overview of the LEM 2 system: Vol. 1: Modifications and applications to land use issues. Report No. 6/81-83. University School of Rural Planning and Development, University of Guelph. Land Evaluation Group. 1985. Socio-economic assessment of the implications of climatic change for food production in Ontario. Pub. No. LEG-22. University School of Rural Planning and Development University of Guelph. Lang, Reg. 1978. Environmental information in a planning/management context in Applications of Ecological(biophysical) Land Classification in Canada. C.D.A. Rubec ed.. ELCS report no.7. Lands Directorate, Environment Canada, Ottawa. Lang R, and Audrey Armour. 1980. Environmental planning resourcebook. Lands Directorate, Environment Canada, and Multiscience Pub. Ltd., Montreal. Lee, Brenda J.. 1982. An ecological comparison of the McHarg method with other planning initiatives in the Great Lakes Basin. Landscape Planning. 1982(9): 147-169. Lindblom, Charles E. 1969. The Science of Muddling Through. Public Administration. 1959(2): 79-88. Litton, Burton Jr., and Martin Keieger. Design with nature: A review. AIP Journal, Jan 1971: Lock, Andrew R. 1983. Applying decision analysis in an organizational context in Analysing and aiding decision processes. P.Humphreys, O.Svenson, and A.Vari (eds.). North-Holland Pub Co., Amsterdam. Luttmerding, H.A.. 1980. Soils of the Langley-Vancouver map area. RAB Bulletin 18. British Columbia Ministry of Environment Victoria. Mabbutt J.A.. 1968. Review of Concepts of Land Classification, in Land Evaluation. 142 G.Stewart ed... MacMillan, London. pp.11-27. MacDougall, Bruce E. 1975. The accuracy of map overlays. Landscape Planning. 1975(2): 23-30. MacDougall, Bruce E.. 1983. Microcomputers in Landscape Architecture. Elsevier, New York. Mack, Ruth P.. 1971. Planning on uncertainty: Decision making in business and government administration. John Wiley and Sons Inc., New York. Manning, Edward, and Sandra Eddy. 1978. The agricultural land reserves of British Columbia: An impact analysis. Paper No.13. Lands Directorate, Environment Canada, Ottawa. March, James G, and Zur Shapira. 1982 Behavioral decision theory and organizational theory, in Decision making: An interdisciplinary inquiry. Gerardo Ugson and Daniel Braunstein eds.. Kent Publishing Co., Boston, pp.92-116. Marsh, William. 1983. Landscape planning: Environmental applications. Addison-Wesley Pub. Co., Reading, Massachusetts. Mason, Richard O., and Ian I. Mitroff. 1981. Challenging startegic planning assumptions: Theory, cases, and techniques. John Wiley and Sons, New York. McAllister, Donald M.. 1980. Evaluation in environmental planning: Assessing environmental, social, economic, and political tradeoffs. MIT Press, Cambridge, Massechusetts. McCormack, D.E., and R.W. Johnson. 1982. Soil potential for onsite sewage disposal in Leon County, Florida. Soil Survey and Land Evaluation. 2(1): McCurdy, Dwight R., and Charles CMyers. 1978. Methodologies for designing inventories to support management information systems, in Integrated inventories of renewable natural resources: Proceedings of a workshop. Jan 8-12, 1978, Tucson, Arizona. Rocky Mountain Forest and Range Experiment Station, Arizona, pp. 150-154. McDonald, Geoff T., and A.L. Brown. 1984. The land suitability approach to strategic land use planning in urban fringe areas. Landscape Planning. 1984(11): 125-150. McDonald, Geoff. School of Australian Environmental Studies. Personal Interview. Guelph, Ontario. 18 July, 1985. McHarg, Ian L. 1969. Design with Nature. Doubleday and Co., Garden City, New York. McRae, S.G and CP. Burnham. 1981. Land evaluation. Clarendon Press, Oxford. Meyers, Charles, Michael Kennedy, and R.N. Sampson. 1979. Information systems for land use planning, in Planning the uses and management of land. M.T. Beatty, G.W. Petersen, and LD.Swindale (eds.). Soil Conservation Society of America. Moore J.H. and M. G. Chang. 1983. Meta-Design considerations in building DSS. in Building decision support systems. J.L. Bennet (ed.). Addison-Wesley Pub. Co., 143 Reading, pp. 173-203. Moss, Michael, and W. Nickling. 1980. Landscape evaluation in environmental assessment and land use planning. Environmental Management 4(1): 57-72. Murdick, Robert G.. 1980. MIS: concepts and design. Prentice-Hall Inc. Englewood Cliffs, New Jersey. Naveh, Zev, and A.S. Lieberman. 1984. Landscape Ecology: Theory and Application. Springer-Verlag, New York. Nijkamp, Peter. 1980. Environmental policy analysis: Operational methods and Models. John Wiley and Sons, New York. Nijkamp, Peter. 1983. Information systems for regional development planning: A state of the art survey. Environment and Planning B: Planning and Design, 1983 (10): 283-302. Nix, H.A.. 1968. The assessment of biological productivity, in Land evaluation. G.A. Stewart (ed.). MacMillan, London, pp. 77-87. Nowland, John L. 1977. Soil survey concerns in urban areas, in Ecological (biophysical) land classification in urban areas. E.B.Wiken and G.R. Ironside eds.. ELCS no.3. Lands Directorate, Environment Canada, Ottawa. Olschowy, G. 1975. Ecological landscape inventories and evaluation. Landscape Planning. 1975(2): 37-44. Ortolano, Leonard. 1984. Environmental planning and decision making. John Wiley and Sons, New York. Petak, William J. 1980. Environmental planning and management: The need for an intergative perspective. Environmental Management 4(4): 287-295. Peters, W.S.. B.C. Ministry of Agriculture and Food. Phone interview. 26 November, 1984. Poulton, M.C.. 1983. The limits to and effective use of evaluation methods. Environment and Planning B: Planning and Design. 1983(10): 179-192. Rodiek, J.E.. 1978. Landscape analysis: A technique for ecosystem assessment and land use planning. Landscape Planning. 1978(5): 27-44. Robertson, Glen. Planner, District of Matsqui. Personal interview. Clearbrook, B.C.. 20 November, 1984. Roome, Nigel J. 1984. Evaluation in nature conservation decision making. Environmental Conservation. 11(3): 247-252. Rees, W.E.. 1977. The Canada Land Inventory in perspective. Lands Directorate, Fisheries and Environment Canada, Ottawa. Richardson, J.J., and A.G. Jordan. 1979. Governing under pressure. Martin Robertson and Co. Ltd., New York. 144 Riquier, J.. 1974. A summary of parametric methods of soil and land evaluation, in Approaches to land classification, soils bulletin 22. FAO, Rome. pp. 47-52. Sadvoski, Dr. Alexander, and Dr.S.W. Bie. (eds.). 1978. Developments in soil information systems: Proceedings of the second meeting of the ISSS working group on soil information systems, Varna, Bulgaria, may 30 - June 4, 1977. Centre for Agricultural Publishing and Documentation, Wageningen, The Netherlands. Schreier, H. and M.A. Zulkifli. 1983. A numerical assessment of soil survey data for agricultural planning and management Soil Survey and Land Evaluation. 3(2): 41-53. Simon, H.A.. 1976. Administrative Behavior. 3rd Edition. The Free Press, New York. Simon, H.A. 1969. The sciences of the artificial. MIT press, Cambridge, Massachusetts. Sinden, John A., and Albert C. Worrell. 1979. Unpriced Values: Decisions without Market Prices. John Wiley and Sons, New York. Slaymaker, Olav, and L.M. Lavkulich. 1978. A review of land use-water quality relationships and a proposed method for their study. Westwater research Centre, University of British Columbia, Vancouver. Smit, Barry, M.Brklacich, LDumanski, K.B.MacDonald, and M.H.Miller. 1984. Integral land evaluation and its application to policy. Canadian Journal of Soil Science. 64(4): 467-480. Smit, Barry, and Mark Flaherty. 1984. Assessing land use options for future food needs. Paper presented at the annual meeting of the Association of American v Geographers, Washington, D.C, April 1984. Smit, Barry, and the Land Evaluation Project Team. 1981. Procedures for the long term evaluation of rural land. CRD Pub. No. 105. University School of Rural Planning and Development University of Guelph, Guelph, Ontario. Smit, Barry. Land Evaluation Group, University of Guelph. Personal interview. Guelph, Ontario, 4 August 1984. Smith, G. and D.May. 1980. The artificial debate between rationalist and incrementalist models of decision making. Policy and Politics , 8(2): 147-161. Sprague, Ralph Jr. and Eric D. Carlson. 1981. Building effective decision support systems. Prentice-Hall Inc., Englewood Cliffs, New Jersey. Stanley Associates Engineering Ltd. 1983. Draft Report: Sumas mountain rural residential feasibility study. District of Abbotsford and Central Fraser Valley Regional District Steiner, Frederick. 1981. Ecological Planning for Farmlands Preservation. American Planning Association, Chicago. Steiner, Frederick. 1983. Resource suitability: Methods for analyses. Environmental Management 7(5): 401-420. 145 Tjallingii, S.P.. 1974. Unity and Diversity in Landscape. Landscape Planning . 1974(1): 7-34. Thie, J., E.B. Wiken, and C.D.A. Rubec. 1984. Ecological land survey as a basis for land resource planning and management in Canada, manuscript Lands Directorate, Environment Canada, Ottawa. Theiruaf, Robert J.. 1982. Decision Support Systems for effective planning and control: A case study approach. Prentice-Hall Inc, Englewood Cliffs, New Jersey. Underwood McLellan Ltd. 1984. Lower Mainland industrial project: Industrial land strategy. Central Fraser Valley Regional District Vander Horst Dan. Ministry of Lands,Parks and Housing. Personal Interview. Vancouver, 12 March, 1985. VanGundy, Arthur B.. 1981. Techniques of structured problem solving. Van Nostrand Reinhold Co., New York. Vink, A.P.A.. 1975. Land use in advancing agriculture. Springer-Verlag, New York. Vink, A.P.A.. 1983. Landscape ecology and land use. Longman, New York. Ward, Neville E.. 1976. Land use programs in Canada: British Columbia. Lands Directorate, Environment Canada, Ottawa. Yoemans, W.C. 1983. Visual resource assessment: A user guide. MOE Manual no.2. British Columbia Ministry of Environment Victoria. APPENDIX ONE LANDPLAN is a version of LUPLAN which was modified by the Lands Directorate, Environment Canada. At the Lands Directorate, information is available from: Ron Gelinas Lands Directorate, Environment Canada Ottawa, Ontario K1A 0E7 Information regarding LUPLAN can be obtained from: Richard Davis Division of Water and Land, resources CSIRO P.O. Box 1666 Canberra City, ACT 2601 Australia LANDPLAN is written in basic computer language for the Apple 2e or the Apple 2 + . LUPLAN is available for a variety of operating systems. 146 APPENDIX TWO This appendix lists the data items included in the data bank. Sources and method of recording are noted where required. 1. Planning Unit (81 units): This item identifies which of the 81 planning units is being recorded. 2. Landscape Unit (9 units): This is a more general identifier using the nine landscape units. 3. Planning Unit Area: Refers to the area of each unit in square kilometers to one decimal place. 4. Polygon Purity Class (3 classes): This item indicates the purity or internal variability of the unit based on agricultural capability or forest management class. High - Pure capability or one forest ecosystem. Medium - Up to an 80%-20% split with 2 agricultural capability classes or a mix of mature and immature forest ecosystems. Low - More than an 80-20 split with a greater range than 2 agricultural capability classes or a mix of forest ecosystems with at least one more prominent land use. 5. Surficial Geology (16 classes): 1. S l o p e D e p o s i t / C o l l u v i u m 8> Sand t o L o a m / L a c u s t r i n e 2. Lowland s t r e a m / A l l u v i a l 9- f i n e Sand/Beach 3. Mountain S t r e a m / A l l u v i a l l 0 - S i l t / G l a c i o L a c u s t r i n e 4. F i n e Channel F i l l / A l l u v i a l 1 1 • G r a v e l and S a n d / G l a c i o f 1uvia1 5. Sandy Channel F i l l / A l l u v i a l 1 2 - I c e C o n t a c t G r a v e l / G l a c i o f l u v i a l 6. W i n d b l o w n / A e o l i a n l 3 - F 1 ° w T i 1 1 / G l a c i o f l u v i a l 7. S i l t t o C l a y / L a c u s t r i n e Stony S i l t - L o a m y C l a y / G l a c i o m a r i n e 15. Bedrock 16. O r g a n i c s 6. Soil Management Groups (7 Groups): These are management groups identified by Schreier (1983) and delineated from Soils of the Langley Vancouver Map Area. 1980. Report No.15. British Columbia Soil Survey. Scale of 1:25000. 7. Drainage (5 classes): Determined from Ministry of Environment, 1984. Soil Drainage Maps 92G/lc, 92G/ld, 92G/lb, 92G/lg, and Soils of the Langley Vancouver Map Area. 1980. Report N0.15. British Columbia Soil Survey. Classes: 1- rapid 2- Well to Moderately Well 3- Imperfect 4- Poor 5- Very Poor 8. Slope Class (8 classes): Determined from the Soils of the Langley-Vancouver Map Area. 1980. Report No.15. British Columbia Soil Survey. 147 148 0 - 0-.5% 4 - 9-15% 1 - .5-2% 5 - 15-30% 2 - 2-5% 6 - 30-60% 3 - 5-9% 7 - +60% 9. Slope Type (3 classes): Determined from the Soils of the Vancouver Map Area (see previous item) 1 - Simple 2 - Complex 3 - Mixed 10. Aspect (9 directions): Determined from 1:50000 Topographical mapping. Surveys and Mapping Branch. 1980. Mission, map 92G/1. 1 - North 6 - Southeast 2 - Northwest 7 - East 3 - West 8 - Northeast 4 - Southwest 9 - Combination (valley or peak) 5 - South 11. Minimum Altitude: Extracted from 1:50000 Topographical Mapping (see previous item) 12. Maximum Altitude: Extracted from 1:50000 Topographical Mapping (see aspect). 13. Erosion Potential (5 classes): Determined from the Surveys and Mappimg Branch, B.C. Ministry of Environment 1984. Surface Soil Erosion Potential. Maps 92G/lc, 92G/lb, 92G/lg, 92G/lf, 92G/le, Scale=1:25000. Class 1 - negligible (less than 6 tonnes/ha/year) Class 2 - slight (6-11 t/ha/year) Class 3 - moderately severe (11-22 t/ha/year) Class 4 - severe (22-33 t/ha/year) Class 5 - very severe (greater than 33 t/ha/year) 14. Climate, Effective Degree Growing Days (4 classes): Determined from the Surveys and Mapping Branch, B.C. Ministry of Environment 1980. Effective Degree Growing Days ( C days), 1:100000, Langley B.C. Map 92 G/SE. 1. 1100 Effective Degree Growing Days 2. 1000-1100 Effective Degree Growing Days 3. 900-1000 Effective Degree Growing Days 4. 800-900 Effective Degree Growing Days 15. Climate, Climatic Moisture Deficit/Surplus (3 classes): Determined from the Surveys and Mapping Branch, B.C. Ministry of Environment 1981. Climatic moisture deficit/surplus (mm.). 1:100000. 1 - 40mm. 2 - 40-115 mm. 3 - 116-190 mm. all are surpluses 149 16. Agricultural Capability (7 classes): Determined from the Ministry of Agriculture and Food, and Ministry of Environment, Land Capability Classification for Agriculture in British Columbia. 1983. Moe Manual 1, Central Fraser Valley Map Sheet, 17. Limiting Factor (5 factors): Determined from Land Capability Mapping (see above). 1 - Excess Water 2 - Terrain 3 - Stoniness 4 - Low Perviousness 5 - Bedrock/Slope 18. Cover, Matrix (8 classes): Matrix refers to the dominant or background cover (Forman, 1981). Cover, including the following two data items, is determined from 1:50000 topographical mapping and 1:50000 Black and White Air photography. Classes are: 1. R u r a l / D e v e l o p e d 2. F i e l d / R u r a l 3 . Grassland/Open 4. Wetland 5. Brush 6 . O r c h a r d or P l a n t a t i o n 7. Woodland 8 . F o r e s t 19. Cover, Patch (7 classes): Patch cover refers to the fragmented patches of cover within the matrix (Forman, 1981). Classes are: 1. R u r a l / D e v e l o p e d 2. F i e l d R u r a l 3 . Grassland/Open 4. Brush 5. O r c h a r d or P l a n t a t i o n 6 . Woodland 7. F o r e s t 20. Cover, Corridor (5 classes): Corridor cover refers to linear corridors of cover traversing a planning unit 1. D e v e l o p e d / T r a n s p o r t 2. Open 3 . V e g e t a t e d 4. S t r e s s 5. S h o r e l i n e CUee 1. Class 2. Class 3. Class 4. Class 5. Class 6 . Mo s i g n i f i c a n t l i m i t a t i o n * . M o d e r a t e l i m i t a t i o n s . M o d e r a t e l y s e v e r e l i m i t a t i o n s . S e v e r e l i m i t a t i o n s . V e r y s e v e r e l i m i t a t i o n s . O n l y p e r e n n i a l f o r a g e c r o p s ; n o f e a s i b l e I m p r o v e m e n t s . N o c a p a b i l i t y . Class 7. 150 21. Forest Capability (7 classes): Determined from Canada Land Inventory Mapping . Forest Capability. 1973. Map 92G/1 1:50000. C l a a a 1 - 111-130 c u b i c f e a t per a c r e par year growth, - aa r e q u i r e d can be f u r t h e r a u b d l v l d a d aa c l a a a l a (131-150), c l a a a l b (151-170), c l a a a l c (171-190), e t c . , I n c r e a s i n g by 20 c u b i c f e e t f o r c l a a s e a t h e r e a f t e r . 91 - 110 c u b i c f e e t per a c r e per year 71 - 90 c u b i c f e e t per a c r e per year 51 - 70 c u b i c f e e t per a c r e per year 31 - 50 c u b i c f e e t per acre per year 11 - 30 c u b i c f e e t per a c r e per year C l a a a C l a a a C l a a a C l a a a C l a a a Claaa - 0 - 1 0 c u b i c f e e t per a c r e per year 22. Forest Management Class (3 classes): In the forested mountian landscapes, forest site classification was available from the Ministry of Forests as included in the Sumas Mountain Study (Ministry of Lands Parks and Housing, 1984). Class 1 - Good to Medium Class 2 - Medium Class 3 - Meduim to Poor There is no classification available for woodlands in agricultural areas so three qualitative classes were assessed as follows. Class 1 - Class one C.L.I. woodlands which provide the matrix for the planning unit Class 2 - Areas with class one potential currently in scrub or abandoned. Class 3 - Areas currently cleared and established in agricultural or other rural land uses. 23. Aggregate Potential (presence or absence): Determined from Hora and Basham. Sand Mainland. Ministry of Energy, Mines and 1:50000. 1 - Potential aggregate source 0 - No aggregate sources and Gravel Study 1980, B.C. Lower Resources. Figure 3-lb. Scale of 24. Existing Pits and Permits (5 cases): These have been identified from Hora and Basham (1980). Cases include: 0 - no pits or permits 1 - inactive pit or permit 2 - area under permit with no pit 3 - active pit 4 - quarry 25. Recreation - Water features (presence or absence): This item has a value of 0 (no feature) or 1 (feature present) based on the former Central Fraser Valley Regional Official Plan. 1981. Schedule E. Regional Outdoor Recreation Assets Map. 151 26. Recreation - Development potential (presence or absence): Same as above. 27. Significant Freshwater Features (6 classes): These have been extracted from the CFVRD Official Regional Plan (former). Schedule D. Biologically Important Natural Assests Map. Scale of 1:50000. Classes are: 0 - no features 1 - Sloughs 2 - tributary streams 3 - wetlands, marshes 4 - fish bearing streams 5 - lakes 28. Significant Wildlife Features (5 classes): Same source as above. Classes are: 0 - no feature 1 - wildlife breeding area 2 - seasonal wildlife habitat 3 - important areas of habitat diversity 4 - combination of above features 29. Significant Features - Vegetation (4 classes): Same source as item 27. Classes are: 0 - no feature 1 - sensitive representative vegetation 2 - unique vegetation 3 - combination of above 30. Significant Features - Landscape (presence or absence): Same source as item 27 with 0 (absence) and 1 (presence) scores. 31. Significant Features - Historic (presence or absence): Same as above. 32. Diversity (4 classes): Diversity is based on the number of landtypes per sq. km.. A land type is either a natural ecosystem (Mature Forest) or a land utilization type (Agriculture, Rural development). It has been determined from land use maps, 1:50000 B&W air photographs, and 1:50000 topographical mapping. Classes are: 1 - 1 land type per sq. km. 2 - 2 land types per sq. km. 3 - 3 land types per sq. km. 4 - More than 3 land types per sq. km. 33. Water Resources - Order/Density (5 and 3 classes): Order refers to a standard stream classification and is the first number in the data item. These are numbered from 1 to 5 with two first order tributaries combining to create a second order and so on. Density refers to the number of sq. km. units (using a km. grid) containing streams as opposed to the total area of the planning unit These have been determined using 1:50000 topographical 152 mapping. 34. Water Resources - Reach (6 classes): This item indicates the type of stream habitat in a planning unit Classes are: 1 - Sumas River 2 - Flat stream 3 - Incised stream and valley 4 - Agricultural drainage 5 - Slough 35. Water Resources - Downstream Use (5 classes): This item indicates the dominant downstream use as determined from land use and topographical mapping. Classes are: 1 - Fraser River 2 - Undeveloped 3 - Agricultural 4 - Forest 5 - Developed 36. Water Resources - Subsurface (2 classes): This is a very general item taken from the Surficial Geology Map (Armstrong, 1984). Class 1 indicates the presence of significant aquifers while Class 0 includes areas with no significant groundwater resources. 37. Flood Potential - Fraser River (2 classes): This item is taken from the former CFVRD Official Plan. Schedule C. Floodplain Map. A value of 0 indicates no flood potential while 1 means that the area is in the 200 year floodplain of the Fraser River. 38. Transportation (4 classes): This item has been derived primarily from 1:50000 topographical mapping. The following classes have been defined: 1 - Excellent - dual Highway 2 - Good - two lane hardtop with complete coverage of the planning unit 3 - Fair - all weather roads with limited coverage of the unit 4 - Poor - no or very limited (logging road) access. 39. Transportation - Rail (3 classes): This item has been derived from 1:50000 topographical mapping. Classes are: Class 0 - No facilities Class 1 - Track Class 2 - Track and switching/loading facilities 40. Distance From Highway One (4 classes): This item has been measured off of the 1:50000 topographical mapping. The 4 classes are: 1 - greater than 80% of the unit within 1 km of highway 2 - Greater than 50& within 1 km. and more than 80% within 2 km. 3 - Portion of the unit within 2 km. 4 - Not within 2 km. 41. Distance from Abbotsford - Clearbrook (4 classes): 153 This item is identical to the previous item except that the distance is from the urban center. 42. Land Use - Dominant (14 classes): This item lists the dominant land use in each unit as taken from CFVRD. Farm Use Study. 1980. scale of 1:50000; CFVRD Land Use Mapping. 1985. scale of 1:50000; 1:50000 B&W air photographs. Classes are listed below. 1. A g r i c u l t u r e I . 1 H o r t i c u l t u r e 1.2 P o u l t r y / A n i m a l Husbandry 1.3 F i e l d Crops 1.4 Market G a r d e n i n g 1.5 D a i r y 1 . 6 M i x e d 2 . R u r a l 2.1 E x t r a c t i o n 2.2 I n s t i t u t i o n a l 2.3 R e c r e a t i o n a l 2.4 R e s i d e n t i a l 43. Land Use - Subdominant (14 classes): This is the same as the previous item except that it lists uses which are not dominant but still account for 20-49% of the unit area. 44. Land Use - Secondary (14 classes): This item is the same as the previous two but lists uses which account for only 10-20% of the unit area. 45. Adjacent Use (4 classes): Adjacent use includes two numbers. The first indicates the use class: 1 - Agriculture 2 - Rural Development 3 - Urban Development 4 - Undeveloped 46. Adjacent Land Use - Secondary This item is the same as above but notes the secondary, rather than dominant land use. 47. Past Use -This is essentially a dummy item as no use was predicted for it in the policy set The changes in use at a regional scale have not been significant in recent years. 48. Planning Designation (6 classes): This item indicates the general planning designation as determined from local official plans. The classes are: 1 - Agricultural land reserve 2 - Rural resource 3 - Rural residential 4 - Part A.L.R., part urban (urban fringe). 5 - Urban development zone 3. Urban 3.1 i n d u s t r i a l / C o m m e r c i a l 1.2 R e s i d e n t i a l 4. Undeveloped 4.1 p r o d u c t i v e Woodland 4.2 U n p r o d u c t i v e or open 154 6 - Industrial APPENDIX THREE This appendix provides a listing and explanation of the rating code. Each code provides a rating for each land unit from 0 to 1 depending on the diagnostic attributes chosen and the values assigned. The code is written in Basic computer code. The explanation also includes a listing of the data items involved, key sources for the approach or figures used, and an indication of the scale of measurement and the type of input data. R A T I N G S C O D E F O R T H I S E X E R C I S E , 4 0 0 0 REM * * 4 0 1 0 REM * * * * * * 4 1 1 0 FOR J = 4 1 2 0 FOR K = 4 1 3 0 A ( J , K ) = 4 1 4 0 N E X T K * * * * C F V R D E M S * * K * » K » » » * * » « * 1 TO B 1 TO 26 0 4 1 50 N E X T J 4 1 60 R E M 4 1 7 0 R E M ** F I E L D : R O P S * * 4 1 8 0 R E M * * * * 4 1 9 0 I F W ( l ) > 65 T H E N 4 5 6 0 4 2 0 0 R E M * A G . C A R . - P C ) L . 1 * 4 2 1 0 I F W ( 1 6 ) = 1 T H E N A ( 1 , 1 > = 1 : G O T O 4 2 6 0 4 2 2 0 I F W <16) = 2 T H E N A ( 1 , 1 ) 8 : G O T O 4 2 6 0 4 2 3 0 I F W ( 1 6 ) = 3 T H E N A ( 1 , 1 . > = 6 4 : G O T O 4 2 6 0 4 2 4 0 I F W ( 1.6) -- 4 T H E N A ( 1 , 1 ) = 4 9 : G O T O 4 2 6 0 4 2 4 5 I F W ( 1.6) = 5 T H E N A ( 1 , 1. ) = 2s G O T O 4 2 6 0 4 2 4 8 I F W<1.6> = 2 . 5 T H E N A ( 1 , .1.) = 7 2 : G O T O 4260 4 2 5 0 I F W(16) = 3 . 5 T H E N A ( 1 , 1) = 5 5 : G O T O 4 2 6 0 4251 I F W < 1.6) = 4 . 5 T H E N A ( 1 , 1) = 3 5 : G O T O 4.260 4 2 5 5 I F (W <16) < 1 ) A N D (W<16) > ,5) T H E N A ( l , 1 ) = . 7 4 2 6 0 I F W(17) = 1 T H E N A ( 1 , 1. ) = A ( t , 1 ) -»• . 1 : GOTO 4271. 4 2 7 0 I F W(17) = 2 THEN A ( i , 1 ) = A < 1. , I > + . 05 4271 I F A (1,1. > < = .1 T H E N 4 2 8 0 4 2 7 2 IF W<15> = 2 T H E N A ( 1 , 1 ) = A ( 1 , 1 ) - . 0 5 : GOTO 4 2 8 0 4 2 7 3 I F W(15) = 3 T H E N A ( 1 , 1 ) = A( 1 , 1 ) - . 1 Give preference to agriculture on high capability land. Capability classes (data item 16) are assigned values from 0-1 based on general findings from Hoffman (1971) who correlated productivity with capability classes. Addition of .1 is made if the limiting factor is excess water and .05 if limiting factor is terrain due to the ability to compensate for these limitations by special practices. 155 156 4280 REM * PRODUCT CV t TY - PHI... 2 * 4 2 7 0 IF <W(2) = 3 :) OR (W <: ?.) = = 4) THEM 4360 * 3 0 0 IF !'H6) = !. THEN A < :i ,2) = . 5 4 : GOTO 4390 4310 IF W(6) = 2 THEN A( 1 ,2) 5 1 : GOTO 4370 4320 IF W(6) = 3 THEN A ( 1 ,2) = . 5 3 : GOTO 4390 4330 IF W(6) = 4 THEN A ( I ,2) = . 5 0 : GOTO 4390 4340 IF W(6) = 5 THEM A ( i ,2 ) — . 4 7 : GOTO 4390 4350 IF W(6) -- 6 THEN A ( 1 ,2) 5 0 : GOTO 4 3 9 0 4360 IF W(6) = 1 THEN A ( 1 ,2) 5 5 : GOTO 4 3 7 0 4370 IF W(6) = 2 THEN A ( 1 ,2) - . 5 0 : GOTO 4390 4380 IF W(6) = 4 THEN A ( 1 ,2) * EROS ION POT. - P 45 4 3 8 5 REM 0 L . 3 * 4370 IF W(13) = 1 THEN A ( 1 , 3 ) == 1: GOT 0 4440 4410 IF W(13> -•= 2 THEN A CI. ,3) = . 8 3 : GOTO 4440 4420 IF Wi13) = 3 THEN A ( 1 , 3 ) = . 5 5 : GOTO 4 4 4 0 4430 tF W<13) = 4 THEN A ( I , 3 ) = Give prefence io crops on most productive soils. Soil groups (item 6) are assigned a value from 0-1 based on yield data from Schreier (1983). This policy is measured on a metric scale with quantitative data. Discourage agriculture in areas of high erosion potential. Erosion classes (item 13) are assigned a value from 0-1 based on soil losses indicated on the source map from the B.C. Ministry of Environment Policy is measured on a metric scale using quantitative dau. 4440 REM * ENV. I.MPAC T -- POL. 4 * 4450 IF W(7) > = 2 THEN 4 5 2 0 4460 IP W(8) < 2 THEN A ( J. , 4) ~ . 5: GOTO 4470 4470 IF (W(8) > 2) AND (W(5) 1. 6) THEN A ( 1 , 4 ) = . 5 : GOTO 44 70 4 4 8 0 IF W(8) > 2 THEN A ( 1 , 4 ) . 4470 IF W(8) < = 2 THEN A ( 1 , 4 ) = A ( 1 , 4 ) + . 5 : GOTO 4560 4500 IF (W(8) > 2) AND (W(5) = 1 4) THEN A ( 1 , 4 ) = A ( 1 , 4 ) + . 3 : GO TO 4551 4510 IF W(8) > 2 THEN A ( 1 , 4 ) = A ( 1 , 4 ) + . 1 : GOTO 4551 4520 IF W(B> < 2 THEN A ( 1 , 4 ) = . 1 : GOTO 4540 4 5 3 0 IF W(8) > 2 THEN A ( 1 , 4 ) = . Discourage Agriculture in areas of high impact potential The general approach used in this code is documented in Slaymaker and Lavkulich (1978). Planning units are first grouped by slope and drainage (items 7 and 8) then assigned a value between 0-1 on the basis of surficial geology. The policy anempts to express the relationship between these attributes and impart on local water quality. The code is measured on an ordinal scale with largely qualitative data. 4 5 4 0 I F W(8) < = 2 THEN A ( .1 , 4) = A ( 1 , 4 ) + . 3 : GOTO 4551 4550 IF W(8) > 2 THEN A ( 1 , 4 ) - A ( 1 ,4) + . 1 4551 REM * L O C A T I O N P O L . 5 * 4 5 5 2 IF W (3) < 1.0 THEN A ( 1 , 5 ) == . 1 : GOTO 4555 4553 IF (W(3) > 1.0) AND <W(3) < 5 .0 ) THEN A ( 1 , 5 ) = . 3 : GOTO 4555 4554 IF WI3I > 5 . 0 THEN A ( 1 , 5 ) = . 5 4555 IF W(45) = 1.4 THEN A ( 1 , 5 ) = A ( 1 , 5 ) +• . 5 : GOTO 4 5 6 0 4 5 5 6 IF W(45> = 1.3 THEN A O , 5 ) = A ( 1 , 5 ) + . 4 : GOTO 4560 4 5 5 7 IF W(46) = 1.2 THEN A ( 1 , 5 ) = A ( 1 , 5 ) + . 3 : GOTO 4560 4 5 5 8 IF W(46) = 1 . 1 THEN A ( 1 , 5 ) = A ( 1 ,3) + .2 Encourage large continuous areas of agriculture. This policy gives added value to large parcels of continuous agricultural use. Lines 4552 - 4554 assign a value between 0-.5 based on the size of the unit while lines 4555 - 4558 add additional points for adjacent agricultural use. 157 4560 REM 4 5 7 0 REM * * ANIMAL HU SBANDRY * * 4 5 8 0 REM *.• * * * 4 5 9 0 REM * LAND CAP . -• P O L . 6 4600 IF W(16) < = 3 THEN A ( 2 , 6 ) = . 4 : GOTO 4 6 4 0 4610 IF W(16 > = 4 THEN A ( 2 , 6 ) = 6: GOTO 4 6 4 0 4620 IF W ( .16) = 5 THEN A ( 2 , 6 ) = 8 : GOTO 4 6 4 0 4621 IF W(16) = 3 . 5 THEN A ( 2 , 6 ) = . 5 : GOTO 4 6 4 0 4622 IF W(16) = 4 . 5 THEN A ( 2 , 6 ) = . 7 : GOTO 4640 4630 GOTO 4660 4640 IF (W(4) = 1) AND (W(8) > = 5) THEN A<2 ,6 ) = A<2 ,6 ) - .1 : GOTO 4 6 6 0 4 6 5 0 IF (W(4) > = 2) AMD (W(8) > = 6) THEN A ( 2 , 6 > = A ( 2 , 6 ) -. 3 4660 REM * LAND USE -P O L . 7 * 4670 A ( 2 , 7 ) = 1 4 6 8 0 IF (W(42) > = 2 ) AMD (W(42 ) < = 3 ) THEN A ( 2 , 7 ) = . 7 : GOTO 4 7 0 0 4690 IF (W(42) > = 3 ) AND <W(42 ) < == 4) THEN A ( 2 , 7 ) = .3 4700 IF W( IB) = 8 THEN A ( 2 , 7 ) = A ( 2 , 7 ) - . 5 : GOTO 4 7 2 0 4 7 1 0 IF W U B ) = 7 THEN A ( 2 , 7 ) = A ( 2 ,7 ) - . 2 Encourage agricultural activities not requiring high quality land to locate on land of marginal capability. This policy assigns the greatest value ot land that is marginal (class 4 to 5. item 16) but subtracts value for steeply sloping areas (item 8) due to the difficulty and expense involved in constructing support facilities. It is an ordinal measure with qualitative and qualitative data. Discourage poultry operations and animal husbandry in areas of incompatible land use. The value is initially set at 1 with penalty points subtracted for incompatible uses (urban or rural residential) and forested landscapes (items 42 and 18). This is an ordinal measure with spatial descriptive data. * * RURAL RES 4 7 2 0 REM 4730 REM I DENT IAL * * 4740 REM * • « . * * 4750 IF W(J. ) > 65 THEN 4 9 7 0 4760 REM * LAND I IP If P O L . 8 * 4770 IF (W(42) = 2 . 2 ) OR (W(42) 2 . 4 ) THEN A ( 3 , 8 ) = .Bs GOTO 4790 4780 IF W<42) = 4 . 2 THEN A ( 3 , 8 ) . 4 4 7 9 0 IF W(43) = 2 . 4 THEN A ( 3 , 8 ) A ( 3 , 8 ) + . 2 : GOTO 4B00 IF W(43) = 4 . 2 THEN A ( 3 , 8 ) A (3 , 8) + .1 Encourage rural residential development in areas of existing development or unproductive land 8 This code assigns ratings for land use (items 42 and 43) with areas of existing rural residential development and unproductive land being rated from 1 to .4. 4810 REM * FLOODPLA IN - P O L . 9 * 4 8 2 0 IF W(37) = 0 THEN A ( 3 , 9 ) = 1 Discourage development in flood susceptible areas. This policy code assigns a value of 0 to flood susceptible areas and 1 to flood free land. 4830 REM L . 10 * 4840 IF W(18) = 1: GOTO 4860 4 8 5 0 IF W(19) = 4 THEN A ( 3 , 1 0 ) . 3 " COVER - PC.) 5 THEN A ( 3 , 1 0 ) = 10 Encourage rural residetial development in areas of scrubland. This policy considers scrubland to be both an attractive setting fur estate developoment. and relatively unimportant for other uses. A value of 1 is assigned to areas with a scrubland matrix, and .3 to areas with patches of scrubland. 4fc)60 KEM * S E P T I C C A P A B I L I T Y - P O L . 1 1 * 4 8 7 0 I F W(8) > = 6 T H E N 4 9 7 0 4 8 8 0 I F W(5) = 16 T H E N 4 9 7 0 4 8 9 0 A ( 3 , 3 1 ) = 1 4 9 0 0 IP (W(6) > = 3) AND (W(6) < = 5) T H E N A ( 3 , 1.1 ) = A ( 3 , 1 1 > - . 1 ! GOTO 4 9 2 0 4 9 1 0 IF W (6) = 6 T H E N A ( 3 , l l ) = • A < 3 , 1 1. > - . 2 4 9 2 0 IF W(B) = . 5 T H E N A ( 3 , 1 J > = A ( 3 , 1 I. ) - . 4 : GOTO 4 9 4 0 4 9 3 0 IF W(1J ) = 4 T H E N A C S , 11) = A ( 3 , 1 1 ) - . 2 4 9 4 0 IF (W(7) = 1> OR <W<7> = 3) T H E N A (3 , 1 1.) = A < 3 , 11 > - . 2 : GOTO 4 9 7 0 4 9 5 0 I F W(7) = 4 T H E N A ( 3 , 1 1 ) = A ( 3 , 1 1 ) - . 3 : GOTO 4 9 7 0 4 9 6 0 IP W(7) = 5 T H E N A ( 3 , 1 1) = A < 3 , 1 1 ) - . 4 11 158 Give preference to development in areas must suitable for septic sewage disposal. Septic ratings are based on absolute constraints and soil potential ratings as outlined in the approach by McCormack and Johnson (1982). The rating is 0 if slope is greater than 30% or if the unit consists of organics. After these absolute constraints the score starts at one and subtracts points for conditions calling for corrective measures. Land attributes used are texture (item 6), drainage (item 7). slope (item 8) and depth to groundwater (estimated from height above the Fraser River, item 11). * » M O U N T I A N 4 9 7 0 REM 4 9 8 0 REM R E S I D E N T I A L * * 4 9 9 0 REM * • * * » 5 0 0 0 I F W ( 1 ) < == 6 5 T H E N 5 0 8 0 5 0 1 0 REM * S L O P E S - P O L . 1 2 * 5 0 2 0 I F W<8) < = 4 T H E N A ( 4 , 1 2 ) = . 9 : GOTO 5031 5 0 3 0 I F W(B) = 5 T H E N A ( 4 , 12) ---ct •m u 5031 I F WU.O) > 5 T H E N A ( 4 , 12) = A ( 4 , 1 2 ) + . 1 : GOTO 5 0 4 0 5032 I F ( W d O ) < = 1> OR (W(10) = 5) T H E N A ( 4 , 1 2 ) = A ( 4 , 1 2 ) + . 05 12 Encourage mountain development in areas of moderate slope. In mountain landscapes, the policy code assign decreasing value as slope moves from 0 to 30% (item 8) and also provides additional points for areas with favourable aspect (itme 10 with southeast aspect most preferred). This code is on an ordinal scale with descriptive spatial data. 504 0 REIT O L . 1 3 * 5050 :!F W(30) ) I * A C C E S S - P ! T H E N A ( 4 , 1 3 13 Give preference to development in areas with existing access. This code gives areas with existing access (item 38) a value of 1. Also measured on an ordinal scale with descriptive spatial data. 5060 PL ' I AS - P O L . 1 4 * 5070 IF N Cr/.') > 1 = .1 * F O R E S T A R E T H E N A ( 4 , 13) 14 Discourage development in areas of high forestry potential. Areas of marginal value for forestry are given a value of 1 (item 22) and 0 for areas of high forest management potential. Measured on an ordinal scale with qualitative data. 5 0 9 0 REM #* U R B A N / I N D U S T R I A L . * * Give preference to urban development in areas currently 6 0 0 0 REM **••»• reserved for expansion. 6 0 1 0 REM * P L A N . D E S . 15 U A - POt 15 * of 1 is given to those areas designated for z.,-.->,-> TIT i"i'//in> . „ T - , „ ... urban expansion (item 48). This is measured on an 6 0 . 0 I F W (48) 4 T H E N A ( 5 , 15) = o r d i n a l Z , e ^ d e s c r i o u v e ^ ^ 1 159 6 0 3 0 REM * P R O X I M I T Y - P O L . 1 6 * 6 0 4 0 IF W(41) = 1 T H E N A ( 5 , 1 6 ) = . 6 : GOTO 6 0 6 0 6 0 5 0 IF W(41. ) = 2 T H E N A ( 5 , 1.6) = . 3 : GOTO 6 0 6 0 6 0 5 5 GOTO 6 0 8 0 6 0 6 0 I F W(40) = 1 THEN A ( 5 , 1 6 ) = A ( 5 , 1 6 ) + . 4 : GOTO 6 0 8 0 6 0 7 0 I F W(40) = 2 T H E N A ( 5 , 1 6 ) = Encourage urban development in areas adjacent to existing urban centers and the Trans Canada Highway. This policy gives higher value to areas in close proximity to Abbostford/aearbrook and the Trans Canada Highway (item 40 and 41). Measured on an ordinal scale with descriptive spatial data. 6 0 B 0 REM * TRANSPORT A I"ION - P O L . 1.7 * 6081 IF W(38) = 1 T H E N A ( 5 , 1 7 ) = 17 . 5 6 0 8 2 I F W(39) = 1 T H E N A ( 5 , 1 7 ) = A ( 5 , 1 7 ) + . 5 Give preference to development of areas with existing major transportation facilities. The benefit and role played by transportation facilities in urban growth are recognized in this poucy. Areas with major transportation advantages are given a value from J to 1 (items 38 and 39). Measured on an ordinal scale with descriptive spatial data. 6 0 8 3 REM 6 0 9 0 REM EX T R A C T 10 N #* 6100 REM *#** 6 1 1 0 REM •• LAND USE -P O L . 1 8 * 6 1 2 0 I F W(42) = 2 . 1 T H E N A ( 6 , 1 8 ) = 1: GOTO 6 1 6 0 Discourage extraction in areas of incompatible land use. 6 1 3 0 I F W<43> = 2 . 1 T H E N A ( 6 , 1 8 ) = . 7 : GOTO 6 1 6 0 ™* oode provides greater value for extractive uses in 6 1 4 0 I F W(44) = 7. i T H E N A ( 6 , 16) 18 u e u with e" s t i n8 P i B 1 0 ( 1 penalizes areas with _ ^ """ ' some degree of development (items 42.43.44). Measured on 6 1 5 0 GOTO 61 BO 3 0 o r ^ i n a ^ sa^t descriptive spatial data. 6 1 6 0 I F (W(43) > 2 . 3 ) AND (W(43) < 4) T H E N A ( 6 , 1 8 ) = A ( 6 , 1 8 ) 6 1 7 0 I F (W(43) > 2 . 3 ) AND (W<43) < 4) T H E N A (6 ,1 .8) ~ A ( 6 , 1 8 ) - . 1 6 1 8 0 REM * A G O . P O T . Encourage extraction in areas with high quality deposits. - P O L . 1 9 * 6 1 9 0 I F W(73) = 1 T H E N A ( 6 , 19) = IQ A v a I u e o f 1 i s a^'Pied » *ose underlain by 1 aggregate deposists (item 23). Measured on an ordinal scale with presence/absence data. 160 6 2 0 0 REM *#•* 6 2 1 0 REM *# C O N S E R V A T ION * * 6 2 2 0 REM » * * » 20 6 2 3 0 REM • S I G P E A TOR E S - P O L . 2 0 , 2 1 , 2 2 * 0 * 6 2 4 0 IP W(27) > 0 THEN A ( 7 , 2 0 ) = *• * 6 2 5 0 I F W(28) > 0 T H E N A ( 7 , 2 1 ) = 22 6 2 6 0 I F W(29> = 1 T H E N A ( 7 , 2 2 ) = . 5 : GOTO 6 2 0 0 6 2 7 0 I F W<29> > = 2 T H E N A ( 7 , 2 2 ) = 1 Encourage conservation of significant freshwater features. This code assigns a value of 1 to areas with significant freshwater features (item 27). Measured on an ordinal scale with descriptive spatial data. Encourage conservation of significant wildlife habitat Areas that contain significant wildlife habitat (item 28) are assigned a value of 1. Measured on an ordinal scale with presence/absence data. Encourage conservation of significant vegetation. This code assigns a value of .5 to representative natural vegetation and 1 to rare or unique features. This is an ordinal measure based on descriptive spatial information. 6 2 8 0 REM * L A N D S C A P E - POL . . 2 3 * 6 2 9 0 S = 0 6 3 0 0 IF W(30) = 1 T H E N S = 1 5 : GOTO 6 3 7 0 6 3 1 0 I F W U S ) = 8 T H E N S = 5 : GOTO 6 3 4 0 6 3 2 0 I F W(18> = 7 T H E N S = 3 : GOTO 6 3 4 0 6 3 3 0 I F W(18) = 5 T H E N S = .1. 6 3 4 0 IF W<25) = 1 T H E N S - S + 5 6 3 5 0 I F W(26) = 1 T H E N S = S + 3 6 3 7 0 S - S / 30 6 3 7 5 V = 0 6 3 8 0 I F W(18) = 8 T H E N V = 1 0 : GOTO 641.0 6 3 9 0 IF W U B ) = 7 T H E N V = .10: GOTO 6 4 1 0 6 4 0 0 I F W U B ) = T H E N = 5 6 4 1 0 I F W U 9 ) > 0 T H E N V r.: V + 2 6 4 2 0 IP W O ) = 5 T H E N 1 V = V -i- 10 : GOTO 6 4 4 0 6 4 3 0 I F W (8) > 5 T H E N 1 V = V + 20 6 4 4 0 I F W(32) = 1 T H E N V - V + 5 : GOTO 6 4 6 0 6 4 5 0 IF W(32) = THEN V - V + 2 6 4 6 0 D = W(12) - W ( 1 1 ) 6 4 7 0 I F (I) > -- 30) AMD (D < = 100) T H E N V = V + 2 t GOTO 64 90 64BO IF !) > 100 THEN V = V + 5 6 4 9 0 V => V / 80 6 5 0 0 A ( 7 , 23) = S V Encourage the conservation of significant and vulnerable visual landscapes. This policy code provides a landscape assessment The rating is divided into two, with .5 dedicated to a measure of landscape quality, and .5 for landscape sensitivity or visual absorption capacity. Sources for the general approach and identification of parameters include Moss and Nickling (1980). Anderson et al (1979). and Yeomans 23 d983). Landscape significance begins by assigning a value of .5 to areas identified as significant landscapes (item 30). For other units, signficance is a combination of land cover (item 18). water features (item 25) and potential recreational use (item 26). Visual Absorption Capacity (VAC) is based on cover (items 18 and 19). elevation change (items U and 12), and Diversity (item 32). Steep areas with a large elevation change and a uniform forest cover would rank as very sensitive. This rating is measured on an ordinal scale with presence/absence, descriptive spatial, and qualitative data. 161 6301 REM * WATER RESQ U R C E S - P O L . 2 4 * 6 5 0 2 :tF (W(33) > 2 . 1 ) AND (W(33) < 3 . 3 ) T H E N A ( 7 , 2 4 ) = . 2 : GOTO 6 5 0 7 6 5 0 3 I F W(33) = 3 . 3 T H E N A ( 7 , 2 4 ) = . 4 : GOTO 6 5 0 7 6 5 0 4 I F W(33) = 4 . 1 T H E N A ( 7 , 2 4 ) 24 = . 3 : GOTO 6 5 0 7 6 5 0 6 I F W(33) > 4 . 2 T H E N A ( 7 , 2 4 ) = . 5 6 5 0 7 I F W(35> = 5 T H E N A ( 7 , 2 4 ) = A ( 7 , 2 4 ) + . 5 : GOTO 6 5 0 9 6 5 0 8 I F W(35> = 3 T H E N A ( 7 , 2 4 ) = A ( 7 , 2 4 ) + . 3 6 5 0 9 I F W(34) = 5 T H E N A ( 7 , 2 4 ) = A ( 7 . 2 4 ) - . 1 Encourage conservation of surface water resource. This policy code assigns increasing value as the order and density of surface water increases. Value is decreased for agricultural drainage but increased for developed watersheds. Measured on an ordinal scale with spatial descriptive and qualitative information. 6 5 2 0 REM * * F O R E S T R Y ** 6 5 3 0 REM * * * * 6 5 4 0 REM • F O R E S T PUT E N T I A L - P O L . 2 5 * 6 5 5 0 IF W(22) « 1 . 3 T H E N A ( 0 , 2 5 ) = . 5 : GOTO 6 6 0 0 6 5 6 0 IF W(.22> < = 1 . 2 T H E N A ( 8 , 25) = 1 6 5 8 0 I F W<22> = 2 . 1 T H E N A ( 8 , 2 5 ) = . 6 : GOTO 6 6 0 0 6 5 9 0 IF W(22) = 2 . 2 T H E N A ( 8 , 2 5 ) = . 3 : GOTO 6 6 0 0 6 5 9 5 GOTO 6601 6 6 0 0 I F W(.2t> > 1 T H E N A ( 8 , 2 5 ) = A ( 8 , 2 5 ) - . 3 Encourage forestry in the areas of highest management potential. A value of .5 is assigned to areas of medium potential and 1 to areas of high potential (item 22). In the case of woodlots, values are .3 and .6 to reflect the secondary nature od the use in an agricultural mosiac Value is decreased if the C L I . rating is less than class one (item 21), a rarity in this areas. Measured on an ordinal scale with qualitative data. 6601 REM * DOWNSTREAM CONDI T I O N - P O L . 2 * 6 6 0 2 IF W<22) = 2 . 3 T H E N 6 6 1 0 6 6 0 3 A ( 8 , 2 6 ) = 1 6 6 0 4 I F (W(33) > - 1 .2 ) AND (W( 33) < = 2 . 2 ) T H E N A ( 8 , 2 6 ) = A ( 8 , 2 6 ) - . 2 : GOTO 6 6 0 7 6 6 0 5 I F W(33) = 2 . 2 T H E N A ( 8 , 2 6 ) =• A ( 8 , 2 6 ) - . 3 s GOTO 6 6 0 7 6 6 0 6 I F W(33) > 2 . 2 T H E N A ( 8 , 2 6 ) = A ( 8 , 2 6 ) - . 4 6 6 0 7 I F W(35) = 3 T H E N A ( 8 , 2 6 ) = A ( 8 , 2 6 ) - . 3 s GOTO 66 . lO 6 6 0 8 I F W(35) = 5 T H E N A ( 8 , 2 6 ) = A ( 8 . 2 6 ) -Discourage forestry in areas of downstream impact. This policy begins by establishing a value of 1 and then subtracting value in the case of downstream environments potentially sensitive to forest impacts (agriculture, urban, item 33). Measured on an ordinal scale using qualitative data. 162 6 6 1 0 REM • » # * * 6 6 2 0 REM •*•* E X C L U S I O N P O L I C I E S * * 6 6 3 0 REM «-**•« Exclude uiban development from the A . L R . 6 6 4 0 REM * A . L . R . * "] 6 6 5 0 I F W(48) 1 THF rN 6681 ^ u l b a n raon8S are set to zero for planning units 6 6 6 0 FOR K = 15 TO 17 contained entirely within the A . L R _ 6 6 7 0 A ( 5 , K > = 0 6 6 8 0 N E X T K 6681 REM * C R I T I C A L LAND # 6 6 8 5 I F <W(43> = 3 . 2 ) OR (W(43> = , , , . 3 . 1) T H F N 6 6 9 7 Exclude rural development from critical agricultural land. b b 9 l \ \* ! ^ , 2 ! " 3 ' 2 ) ° R ( W ( 4 2 ) = 2 A l l rural development ratings are set to 0 for planning .-..1) I HEN 6 6 ? / * - ^ ^ j m 3 agricultural capability (item 16) 6 6 9 2 I F W < 16) > 3 T H E N 6 6 9 7 ^ ^ development (items 42 and 43). 6 6 9 3 I F <W<42> = 2 . 4 ) OR <W(43) = 2 . 4 ) T H E N 6 6 9 7 6 6 9 4 FOR K = 8 TO 1.1 6 6 9 5 A ( 3 , K > = 0 6 6 9 6 N E X T K 6 6 9 7 REM » F A R M I N G ON SI. O P E S * 6 6 9 8 I F W(8) < 6 T H E N 6 710 6 6 9 9 FOR K - 1. TO 5 6 7 0 0 A (1 , K > -- O 6 7 0 5 N E X T K 6 7 1 0 REM *# END OF P O L I C I E S Exclude farming from steeply sloping areas. This exclusion policy was designed to prevent unreasonable results in the form of farming in the mountain environments where cultivation is not possible. Fanning is rated at 0 for any areas with a slope greater than 30% APPENDIX FOUR This appendix consists of the output from the base scenario with all policy weights being equal. As can be seen, the suitability scores are all low. This is due to the percentage vote being spread evenly among all 26 policies, thereby reducing the potential score of any one land use. The LANDPLAN function can be expressed as: Maximize St, = £ v • Z • V* where I = 1, 2, . . . n = planning parcels or mapping units, / m 1,2, ... tn  m feasible land-use options, k = 1, 2, . . . p B preference policies whose satisfaction is to be maxi-mized, £(£ = 0 o r l ) = exclusionary policy proscribing given uses from some parcels, R (0 < R < 1) • policy satisfaction rating or degree to which a given use on a given site satisfied a particular policy, V = policy weight or "vote" for a policy by a given individual, S • aggregate policy satisfaction for a given use on a given parcel of land. (McDonald and Brown, 1984, p.130) OUTPUT YOUR WEIGHTS ARE: POLICY 1 2 3 4 5 6 7 8 WEIGHT 5 5 5 5 5 5 5 5 PERCENTAGE 3. B5 3. 85 3.85 3.85 3. 85 3. 85 3. 85 3.85 POLICY 9 10 1 1 12 13 14 15 16 WEIGHT 5 5 5 5 5 5 5 5 PERCENTAGE 3.85 3. 85 3.85 3.85 3.85 3. B5 3.85 3. 85 POLICY 17 18 19 20 21 22 23 24 WEIGHT 5 5 5 5 5 5 5 5 PERCENTAGE 3.85 3 • 85 3. 85 3.85 3.85 3.85 3. 85 3. 85 POLICY 25 26 WEIGHT 5 5 PERCENTAGE 3.85 3.85 163 Y O U R PLAN I S l POLYGON AREA PREFERRED LAND USE RANK SUITABILITY SCORE 1 1.62 1 2 7 B 6 .16 .08 .02 O 0 5 4 3 0 0 O 2 1.55 1 2 7 B 6. .16 .07 O 0 0 5 4 3- 0 0 0 3 1.75 7 1 2 8 6 .12 .08 .07 0 O 5 4 3 0 0 0 4 1.49 1 2 6 B 7 .13 .07 .05 .04 1 5 4 3 O 0 0 5 5.25 7 3 B 2 6 .14 .08 .05 .04 0 5 4 1 0 0 0 6 9.5 .1 3 2 B 7 .14 .08 .07 .05 1 6 5 4 O 0 O. 7 1.42 1 2 8 7 6 .14 .07 .05 O 0 .5 4 3 - O O O B 2.8 ' 3 7 1 2 8 .11 .1 .09 .04 3 6 5 4 O O O 9 1.2 1 2 7 8 6 .15 .07 0 O O 5 4 3 O O O 10 12 1 3 2 8 6 .15 .08 .08 .05 5 7 5 4 .01 0 O 11 1.55 1 3 2 7 8 .15 -.1 .07 -0 0 6 5 4 O 0 0 12 15.25 1 3 2 8 7 .13 .09 .07 .02 2 6 5 4 O O O 2 2 O 1 3 9 . 7 4 1 3 2 B 7 . 1 3 . 1 .OB . 0 2 6 5 4 . 0 0 0 1 4 4 . 4 6 1 3 2 6 7 . 1 5 . 1 . 0 8 . 0 5 8 5 4 0 0 0 1 5 2 1 6 3 2 7 . 1 3 . 1 . 0 9 .OB 8 5 4 (j 0 0 1 6 4 . 1 6 3 1 8 6 2 . 1 1 . 0 9 . 0 5 . 0 5 5 7 5 4 . 0 4 0 O 17 18 19 21 22 23 24 25 26 2.0Z 1.97 2.46 2.83 2. 38-1.9 , 75 3. 12 2. 25 1. 07 4.4! 1 8 8 3 8 1 7 3 8 3 7 3 8 7 1 5 3 2 3 5 1 *j 1 5 3 6 1 6 1 6 2 6 1 8 5 8 2 4 7 5 1 1 4 5 4 5 4 1 4 8 3 1 4 . 14 0 . 1 1 .02 . 14 0 . 1 . 03 .09 O . 14 O . 1 6 .02 . 1 1 0 . 14 .02 .09 O . 14 0 . 1 0 . 1 0 .09 0 . 09 0 . 07 0 . 09 O . 13 O . 07 O . 1 0 .08 0 , 09 0 , 09 0 . 08 0 .08 0 .09 0 .06 0 .08 0 . 1 0 . 05 O . 09 0 165 ,03 . 09 ,05 ,02 .07 . 06 . 05 . 08 .07 , O S 28 29 30 31 .32 33 2.06 2.73 2. 79 ,92 .83 9.4B 1 6 3 7 1 6 8 5 2 S 1 s 3 5 1 5 3 5 3 4 1 4 7 4 4 6 4 7 4 7 1 7 3 1 1 0 , 13 ,02 . 12 .05 .07 0 .07 0 . 16 0 . 1 o . 1 o . 1 o .05 O .06 O . 14 O .08 0 .03 0 .08 0 .05 O .01 O .07 O ,03 .03 , OS ,04 34 35 37 39 40 41 43 44 45 46 47 4B~ 49 50 51 3.24 15 4. 08 1.37 3 . 3 8 4. 16 9.9 1 . 97 3. 26 6.27 2.45 4. 24 31 4.: 1 5 7 5 2 5 1 6 1 .95 1. 18 1 5 1 5 I 6 1 8 1 8 1 7 1 8 1 8 1 7 1 8 7 5 3 6 4 1 4 7 4 4 1 4 7 4 5 3 5 7 5 3 8 3 5 3 5 5 6 7; 5 1 4 1 3 7 3 -> 3 1 3 7 4 7 3 7 7 4 2 4 3 4 5 4 2 4 2 4 3 4 2 4 2 3 8 . 14 0 , 16 0 , 07 0 . 13_ 0 . 15 0 .07 0 . 18 0 . 19 . 0 . 12 0 . 15 0 . 17 0 . 15 .02 . 15 0 . 15 0 . 1 .01 . 16 0 . IS 0 . 13 0 .07 0 . 14 0 .07 0 . 09 0 . 07 0 . 06 0 . 13 0 .07 O .09 o .09 0 . 1 o .09 0 .09 0 .09 0 . 1 0 .09 0 .08 0 .09 O .01 0 .07 0 .06 0 . 08 O . 01 o 0 o .07 0 .01 0 . 08 o .07 0 .09 O . 06 0 .07 0 .07 O .09 0 .07 0 .07 0 .09 0 166 o .07 , 03 ,05 ,08 ,05 ,05 , 05 ,08 , 05 .05 167 o 52 5 3 54 55 56 57 58 39 6 0 61 62 63 64 65 66 67 68 69 1 . 75 34 4.41 1. 37 ,41 2. 15 2. 27 9.3 8. 18 19. 36 1 O -> 38 4. 47 2. 56 6. 38 1 . 36 3.24 1 6 1 5 1 8 1 5 3 7 1 2 1 6 1 5 1 5 1 5 1 5 1 5 7 6 7 8 4 6 4 5 4 5 4 5 4 6 4 3 5 2 4 1 6 3 5 7 4 7 4 7 4 7 4 2 4 1 5 3 4 8 3 7 3 8 3 8 3 4 7 3 6 4 3 4 2 4 2 3 2 3 3 7 3 4 5 1 7 1 8 1 7 1 7 1 .09 0 . 17 0 . 14 0 . 12 0 . 17 . 03 .09 .04 . 16 0 . 18 0 . 18 0 . 17 0 . 17 0 . 17 0 . 22 0 .09 .03 . 08 0 .08 0 . 0 5 0 .07 0 .09 0 .06 0 .09 0 .07 0 .07 0 .08 0 .07 0 .08 0 . 13 0 . 12 0 .07 0 .07 O . 12 0 .09 0 . 05 0 . 06 0 . 05 0 .05 0 ' . 09 0 . 05 0 .07 0 .01 0 .07 0 .08 0 .07 0 .07 0 .07 0 .07 0 .07 0 .02 0 .05 0 .08 0 . 03 0 .05 0 . 04 0 .05 0 .01 .05 ,05 .07 ,07 ,05 ,06 . 02 .02 .02 ,02 .0 70 71 72 73 74 73 76 77 78 79 80 81 5. 39 1 . 33 7. 56 3.64 1.33 11.57 1.47 1.86 9. 48 6. 34 9.3 7 5 7 5 8 5 8 5 8 5 8 5 8 5 4 5 7 5 8 5 8 5 8 5 4 3 4 3 7 3 7 3 4 3 7 3 4 3 8 3 4 3 7 3 7 3 7 3 1 6 1 4 1 4 1 7 1 4 1 2 1 7 1 8 1 4 1 4 1 4 1 . 1 1 0 . 14 0 . 1 o .08 0 . 1 0 .09 0 . 1 0 . 08 0 . 11 0 . 1 0 .09 o .08 o . 1 0 .08 0 .09 0 . 06 0 . 1 0 .05 0 . 1 0 .07 0 .05 0 .03 0 .04 0 .06 0 . 06 0 . 05 0 .08 0 .02 0 .03 0 .03 0 .02 0 .06 0 .05 0 .03 0 .03 0 .02 0 168 . 05 .03 .02 .02 .02 . 02 .02 .02 , 02 .02 , 02 ,02 EXERCISE SUMMARY « PREFERED LAND USE LAND USE 1 2 3 4 5 6 7 a TOTAL SOLUTION AREA 234.18 9.07 31.58 15.4 0 O 37. 77 44. 65 372.65 AREA C/.) 62.84 2.43 8. 47 4. 13 0 0 10. 14 1 1. 98 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items