THE RESPONSE OF RURAL SETTLEMENT TO A LOCAL HAZARD SYSTEM: A MODEL AND SIMULA'HON Thesis for the Degree of M. A. MTCHIGAN STATE UNIVERSITY MARK E NETTHERCUT 1977 3 1293 '1"62"6§"'644'5'3'" Michlgan State Umvemty ABSTRACT THE RESPONSE OF RURAL SETTLEMENT TO A LOCAL HAZARD SYSTEM: A MODEL AND SIMULATION BY Mark E. Neithercut A theoretical model of the interaction between the settlement process and a natural event system is con- structed, based on empirical and theoretical work in the areas of settlement geography and natural hazards research. Settlement geography lacks a strong theoretical framework, and tends to ignore settlement as a process. No basic generalizations have emerged concerning the interaction of settlement and the environment, and the role of hazard systems in historical processes has yet to be considered. The study focuses upon three important questions: (1) What is the general character of the process of settling a rural agricultural area? (2) How often and with what magnitude do natural events occur? and (3) How does the settler's perception of the hazard system influence the resulting settlement pattern? A verbal model is developed and trans- lated into an interactive computer simulation model. The stochastic simulation is written in the computer language BASIC. THE RESPONSE OF RURAL SETTLEMENT TO A LOCAL HAZARD SYSTEM: A MODEL AND SIMULATION BY 3 Mark E: Neithercut A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Geography 1977 Dedicated To LARRY NEITHERCUT ii ACKNOWLEDGMENTS I am grateful to Dr. David Stephenson for his initial encouragement and support of this project, and for the patient counsel he has given me during its evolution. I am indebted to Dr. Robert Wittick who has viewed my entire Master's Program with a constructive and critical eye. Dr. Daniel Jacobson has shown himself to be an eSpecially warm and helpful member of my Advisory Committee, and Dr. Harold Winters has been an invaluable source of aid and guidance during my stay at Michigan State Univer- sity. Thanks are also due to Dr. John Gullahorn for his insight into the use of computer simulation. Pat Terry and Peter Pirozzo of the Behavioral Science Instructional Laboratory need to be recognized for their friendly assist- ance during my use of their facilities. I am grateful to the Department of Geography for its support in the form of a Graduate Teaching Assistantship. I am particularly proud to have been associated with the Atlas of Michigan Project, and I thank Robert W. McKay, Chief Cartographer of the Project, for an Atlas Research Assistantship. Good friends and family are quite important to projects such as iii this and I am most appreciative of their not uncritical support; Joan Glass deserves special thanks. iv LIST OF LIST OF CHAPTER I. TABLE OF CONTENTS TABLES O O O O O O O O O O O O O O O O O FIGURES O O O O O O O O O O O O O C O O A THEORETICAL MODEL OF HAZARD-SETTLEMENT INTERACTION O O O O O O O O O O O O 0 II. SETTLEMENT GEOGRAPHY AND THEORIES OF THE SETTLEMENT PROCESS . . . . . . . . . . III. NATURAL HAZARD RESEARCH AND MODELS OF MAN-HAZARD INTERACTION . . . . . . . . IV. THE DEVELOPMENT OF THE MODEL . . . . . . V. A COMPUTER SIMULATION MODEL: SETSIM . . VI. CONCLUSION . . . . . . . . . . . . . . . APPENDICES A. A USER'S GUIDE TO SETSIM . . . . . . . . B. A SAMPLE SETSIM SESSION. . . . . . . . . C. A LISTING OF THE PROGRAM SETSIM. . . . . LIST OF REFERENCES . . . . . . . . . . . . . . . Page vi vii 15 40 59 82 110 113 116 127 138 LIST OF TABLES Table Page 1. Annual Floods on the Wabash River, at Lafayette, Indiana 1924-57 . . . . . . . . . . 94 2. Results of Polynomial Regression . . . . . . . . 99 vi Figure 1. 2. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. LIST OF FIGURES Bylund's Four Theoretical Models of Settlement Development . . . . . Bylund's Refined Model of Settlement Development . . . . . . . . . . The Magnitude Function, f(xl, x2, x3) Determines the Density. (2), of Settle- ment in Any Area, dB, of the Biotope . . . . A Rough Model of Decision . . . . Human Adjustment to Natural Hazards, of A General Systems Model . . . Human Adjustment to Natural Hazards Model of the Settlement Process . An Example of Gumbel Analysis . . Model of the Natural Events System Model of the Adjustment Process . A General Model of Rural Settlement-- Natural Hazard Relationships . . The Idealized Settlement Plane . . Outlines Mean Information Field of Equal Settlement in All Areas . . . . . . . . . . Mean Information Field of Four Settlers in Center Area . . . . . . . . . Cumulative Mean Information Field of Four Settlers in Central Area . . . . Flowchart of SETTLE Routine . . . vii Page 26 27 30 49 53 54 66 70 71 79 80 86 89 89 90 93 Figure Page 17. Annual Flood Series at Lafayette, Indiana on Logarithmic Probability Paper . . . . . . . 96 18. Histogram of 1000 Lognormal Variates . . . . . . 97 19. Annual Flood Series at Lafayette, Indiana . . . 98 20. Flowchart of EVENT Routine . . . . . . . . . . . 102 21. Hazard Perception Transformation Curve . . . . . 104 22. Flowchart of ADJUST Routine . . . . . . . . . . 106 23. Flowchart of SETSIM Program . . . . . . . . . . 108 viii CHAPTER I A THEORETICAL MODEL OF HAZARD- SETTLEMENT INTERACTION Introduction Two specific themes have dominated human geography since its founding: man's spread across the face of the earth making it his home and man's perception of and interaction with his environment. From Aristotle's expla- nation of the suitability of settling a region as a function of the distance to the equator, to the most sophisticated utilization of LANDSAT imagery, these themes have con- tinually been at the forefront of geographic investigation. The tOpics of human settlement and man's relation— ship with his environment are so basic to the human con- dition that their timelessness is not particularly sur» prising. These themes are sufficiently broad to attract the attention of members of other disciplines, yet con- sistently they have been a major focus of human geographers. Within geography, the interpretation of these themes has varied. Indeed, much of the history of the discipline and the history of geographical ideas in America is embodied in the rise and fall of differing orientations to these themes. This study is an investigation into an area at the core of these two themes: the relationships between human settlement and the environment. It is an attempt to create a general conceptual model which will explain the basic relationships between the process of man settling the land and the surrounding environment. The final goal of this study is the translation of the model into a computer simulation so that it might be eventually compared with real world settlement processes. The development of a model nearly always requires the focus to be narrowed in the difficult struggle to delineate the workings of a complex system. The translation of the model into a mechanical form, the simulation, understandably requires another degree of specificity. It is, then, the goal of this work to develop and then simulate a conceptual model of the relationships between rural settlement and the most dramatic elements of the environment, natural hazards. Settlement Geography While the study of settlement has been a common tOpic of investigation in geography, the maturity of the pursuit remains at a low level. An indication of this is the relatively recent attempt to establish initial specific definitions in settlement geography and the disagreement which followed (Stone, 1965; Jordan, 1966; Stone, 1966; Mitchell, 1966). Until recently, much of the work in settlement geography was simply historical narratives of geographical topics. Gentilcore's "Vincennes and French Settlement in the Old Northwest" (1957) provides a good example of this type of work. Attempts at analysis or explanation during this period were usually mechanistic and/or deterministic. An example is McDermott's attempt to explain an advance and retreat of settlement in Northern Ontario solely as a result of changing provincial policies (1961). The emergence of a sound analytic approach in historical geography, primarily through the influence of Andrew H. Clark (Jakle, 1971, pp. 1090-1), brought new depth to the study of settlement. These later studies usually endeavored to reconstruct, through the use of original survey and land office records, the detailed settlement pattern at a number of points in time. This time slice approach allowed interpolation between the reconstructed geographies of the past to show different periods of geographic change. This type of work is best exemplified by Harris' notable The Seigneurial System in Early Canada: A Geographical Study (1966). The weakness of such studies is that the data used to construct the cross sections may not effectively delimit the periods of devel- opment or changes within a settlement; perhaps more importantly the idea of process is simply ignored. There have been few research works devoted to the study of settlement as a process. Bylund constructed a deterministic model of settlement in Northern Sweden which allowed him to simulate settlement by assigning attractive weights to a church, a road and three parent settlements (Bylund, 1960). Hudson, borrowing a theoretical framework from plant ecology has attempted to construct a general location theory for rural settlements (Hudson, 1967; Hudson, 1969). Norton, most recently, has simulated the settlement of Southern Ontario stochastically, utilizing a number of indicators of township character, such as land quality and distance to a regional entry point (1976). These studies are major contributions to the initial understanding of the process of settlement, yet the major thrust of all of the models has been to predict rather than explain the process. Norton, for example, is not as interested in understanding and being able to explain the workings of the settlement process in Ontario, as he is in being able to develop a predictive tool that could be used to duplicate the process in areas where no land records exist. Indeed, even Hudson's attempt to construct an explanatory theory has been criticized by some who question its universality as well as the appli- cability of ecological theory (Grossman, 1971). Hudson's theory is based simply on the locational aspects of the settlers, that is their frequency and distance apart. In this way Hudson avoids an explanation of settlement as a human process. Geographers have long recognized the importance of differing environmental influences on the settlement process, and yet rarely have they attempted to generalize into a conceptual framework the character of these relation- ships. Griffith Taylor's monograph Canada: A Study of Cool Continental Environments and Their Effect on British and French Settlement (1947) was a landmark study in describing the physical and climatological stage on which Canada was settled, yet it remained a descriptive narrative. Perry explained that one of Taylor's ". . . major interests was in the effect of climatic conditions on settlement in Australia . . ." (Perry, 1966, p. 138), that is, he was interested in the end result, and not in the process which caused the result. Perry's article, encouragingly titled ”Climate and Settlement in Australia 1700—1930: Some Theo- retical Considerations" is designed to show the ". . . origin of their optimism . . ." and ”. . . the general trend of thinking about climate and settlement in Australia . . ." but does not deal with the theoretical consideration of the rela- tionship between environmental hazards and the settlement pro- cess (Perry, 1966, p. 139). Indeed, Heathcote's magnifi- cent study Back of Bourke: A Study of Land Appraisal and Settlement in Semi-Arid Australia (1965) is an inspired geosophy of Australian settlement, but again Heathcote does not attempt to elaborate on the general relationships between settlers and a hazardous environment. There have been a number of other studies concerning early settlement and environmental perception. Peters (1969), for example, has shown how early Kalamazoo County, Michigan, was settled in successive stages owing to the settlers' perception of the attractiveness of the prairies. Yet all of these studies suffer from a weakness that Taafe has criticized an earlier period of geography for: "Its failure to lead to cumulative generalizations. What one geographer found out about the effect of environmental features was seldom referred to in the next geographer's study" (Taafe, 1974, p. 5). Clearly then, while much of the recent research in the area of settlement and settlement's relationship to the environment has been quality empirical investigation and landmark attempts at predictive models, there is a distinct lack of attempts to develop explanatory models to describe the actual process in general terms. Geography and the Man-Land Tradition The study of man-environment relationships has long been a topic of investigation in American geography. As early as 1864, George Perkins Marsh wrote Man and Nature; or, Physical Geography as Modified by Human Action (1864). James explains that "in his wide reading, especially of the works of Humboldt, Ritter, Guyot, and Mary Somerville, [Marsh] recognized that a 'new geography' had appeared, focusing on the close interconnections between man and his natural surroundings" (James, 1972, p. 195). Marsh was particularly concerned with man's destructive effect upon the environment and in many ways was America's first "conservationist." The man-environment theme continued to be popular in American geography. It was embodied in the concept that human development and behavior is the result of direct responses to the natural environment. This concept, commonly referred to as environmental determinism or simply environmentalism was taken up by American geography as a guiding philosophy in its early years as a professional discipline. Environmentalism sprang to the forefront of geographic thought as geographers began to search for a professional identity, in part because of the common training in geology of most of the early American geog- raphers. During the first quarter of the century, geog- raphers clutched at the concept of environmentalism to provide them with an identity as well as a philosophic rationale. The aim of the environmentalists was clear: they strove to explain the relationship between the natural environment and human development and behavior in causal terms. The philosophical background of environmentalism encompassed the geologist's emphasis on the physical side of relationships, the popularity of social Darwinism during the period, and the immature position of the social and behavioral sciences. The retreat from the mechanistic interpretation of the man-land tradition began in the twenties. This is evident in Harlan H. Barrows' call to redefine geography as "human ecology," or ". . . the mutual relations between man and his natural environment" (Barrows, 1923, p. 5). Barrows went on to argue that "Geographers will, I think, be wise to view this problem in general from the standpoint of man's adjustment to the environment, rather than that of environmental influence'(Barrows, 1923, p. 3). Barrows advocated making the study of man—land relationship not simply a focal point of geography, but its entire defi- nition, giving up other parts of geography to other dis- ciplines. As to historical geography Barrows stated: ". . . it is the special task of the historical geographer to describe and so far as possible to explain this evolution of man's environmental relations" (1923, p. 11). The dominance of environmentalism and the extreme reactions to it are well known and need not be commented upon in this brief attempt to give a background of the study of man's environmental relationships in American geography. Indeed, as Taafe has pointed out, "regardless of position on an essentially philosophical continuum from determinism to free will, the subject matter emphasis remained the same. Both Barrows and Semple, for example, would study relations between man and his physical envi— ronment. Only the verbs they would use in describing their findings would differ" (Taafe, 1974, p. 3). For the most part, the reaction to environmentalism was to ignore the relationships between man and his envi- ronment. The division between the physical and human sides of geography grew. 'Barrows' attempt to change the focus of geography was ignored, astonishing as it is to us today living in an ecology-era. The rise of the Spatial analysis vieWpoint supplanted the man-environment theme in geog- raphy during the 19603. More recently, however, with the rise of an environmentally conscious sector of society, the study of man-environmental relationships has seen a revival. The strength of this comeback is based not only on a new awareness on the part of much of the population of environmental problems but also the need for practical applied geographic research in this area. Gilbert White and his students at the University of Chicago brought new life to the study of man-environment relations through their research on human occupance of flood plains. From these studies others naturally emerged dealing with envi- ronmental perception, that is, how man perceives his environment. These studies differ greatly from the earlier man-land work in that the emphasis was not on a ". . . fixed external environment but on ways in which men 10 structure their environment in their own minds" (Taafe, 1974, p. 12). The proposed focus of this study stands firmly on the shoulders of one hundred years of what Pattison has called "the man-land tradition" (Pattison, 1964). "The man-land tradition dwells on relationships; . . ." explained Pattison (p. 215), and while the orientation to these relationships has fluctuated, it has been a major theme in American geography. This study, which seeks to develop a conceptual model of the relationships of settlement pat- terns and natural hazards is seen as an extension of this historic tradition. The Problem Settlement geography, or ". . . the description and analysis of the distribution of buildings by which people attach themselves to the land" (Stone, 1965, p. 347), has only recently experienced initial attempts to develop a theoretical framework. Most of the work on settlement theory has been developed for predictive rather than explanatory purposes. Although the importance of the influence of the environment on the settlement process has long been recognized rarely have these studies progressed beyond basic description, failing to lead to generali- zations. The recent awakening of an environmentally conscious "Ecology" movement has helped to focus new interest on the 11 relationships of man and his environment. With the devel- opment of natural hazard research much more is now known about the occupation of hazard zones and the decision making process which affects this occupation. However, natural hazard research has tended to ignore the role of hazards in larger historical processes such as settlement. Settlement geography lacks a strong theoretical framework, and tends to ignore settlement as a process. No basic generalizations have emerged concerning the interaction of settlement and the environment, and the role of hazard systems in historical processes has yet to be considered. The development of a theoretical model of the relations between the process of settlement and hazard systems is sorely needed. Questions The initial decision of this study to investigate the relationship between settlement and the natural envi- ronment presented an intriguing if unwieldly problem. What are the influences between the process of man settling the land and the natural environment; what sort of inter- action occurs? There are, of course, a myriad of influ- ences. The image of the 19303 Great Plains drought and the widespread effect of this natural hazard on plains settle- ment patterns was impressive. The depth of recent research in the study of natural hazards and the fact that natural hazards are perhaps the most dramatic and distinctive 12 element of the natural environment encouraged consideration of the relationship between natural hazards and the settle- ment process. What are the relationships between settlers, settlement patterns and natural hazards? How does the natural environment through the agent of natural hazards influence or at least create constraints on the decision making of a settler? How do settlers react to varying levels of flooding, for example? These questions and more all seemed to lead to three specific primary questions: How does the settlement of an area occur? How often and with what intensity do natural hazards occur? and How does the settler's perception of the hazard events influ- ence the resulting settlement pattern? It is these three questions that this study is designed to investigate. In what manner is an area initially settled? There are, of course, a myriad of considerations in choosing a place to settle such as soil quality, distance from trans— portation lines, availability of services, etc. A number of studies have engaged in attempts to classify patterns of settlement and have stated that initial settlement patterns are usually of a uniform nature. However, specific empirical studies frequently point to the clustering of early settlements (Enequist, 1960) for obvious reasons such as equal perception of land attractivity, kinship ties and economic ties. It was hypothesized that generally as man 13 settles the land one settler will tend to settle near another although these small clusters of settlements might be randomly spaced over an area. How often and with what intensity do natural hazards occur? In line with the common concept of "loo-year" and "1,000-year" floods, it was hypothesized that hazard fre— quency and intensity is basically a linear relationship between the average occurrence of a specific event and its intensity. That is, the more infrequent a hazard, the more extreme its intensity. Thus the frequency of an event with a specific magnitude in any hazard system may be roughly determined by its relationship with the magnitude of the system's most frequent event. The most frequent event of any hazard system acts as a base index for that system. What is the relationship between the settlement process and natural hazards? A settler's perception of the hazard is the key to answering this question. The perception of a hazard was hypothesized to be dependent upon three basic assumptions: (1) Infrequent intense hazards of cataclismic proportion are primarily dismissed as flukes, or occur so rarely in a person's lifetime, if ever at all, that the hazard does not produce a significant level of behavioral adjustment on the part of the settler, either in reaction to the hazard itself or to the prospect of future hazards of the same type. That is, in response to a "freak" forest fire, a homeowner will rebuild his 14 house rather than move. (2) Frequent hazard events of low intensity such as an August dry spell, or below zero weather in the north, are perceived to be simply a part of a broader existence. Such events are unconsciously accepted, may be implicitly planned for, and often help define the cultural niche in which they occur. (3) Moderately frequent hazard events of moderate intensity contribute to the greatest perception of a hazardous situation. Moderately frequent hazards were thought to contribute to the accumu- lated perception of hazard more so than quite frequent or quite infrequent events and because of their intensity are most likely to initiate an adjustment in response to the hazard, thus affecting the process of settlement. The next chapter of this study provides a general background on settlement geography and traces the develop- ment of theory and definitions of settlement geography. Chapter Three charts the development of natural hazard research and focuses on a general systems model of hazard zone occupance. Chapter Four puts forth the model of settlement/hazard interaction, and Chapter Five documents the translation of the model into a computer simulation. The concluding chapter discusses the model, and the importance of theory in the development of simulation models. A user's guide to the simulation appears as Appendix A. CHAPTER II SETTLEMENT GEOGRAPHY AND THEORIES OF THE SETTLEMENT PROCESS The Scope of Settlement Geography While the study of settlement is an ancient pursuit, settlement geography as a specific focus within professional geography is of a more recent origin. German and French geographers of the Nineteenth Century were responsible for the first modern work in this area, influenced by Carl Ritter's early studies of human geography. Stone (1965) has explained that much of the early research in settle- ment geography was done by Germans in two specific areas, "house type (including distribution, architecture, and building materials) and urban centers" (Stone, 1965, p. 349).* As early as 1891, F. Von Richthofen lectured in Berlin on settlement geography; A. Meitzen published his four volume summary of settlement research in 1895 which classified all of the German settlements, thereby estab- lishing villages as a key focus of settlement geography *Much of the following discussion is taken from Stone (1965) and Kohn (1954). 15 16 (Meitzen, 1895). Meitzen also stressed form as the important element of settlement classification. 0. Schlfiter, often looked to as the founder of the field of settlement geography, rose to a leadership role near the turn of the century. His definition of the field was a broad one. Stone paraphrased: "To location, size, and growth of settlements and their relationships to nature he added the study of internal structure, external form and appearance, and areal arrangement as well as his— torical, economic and cultural conditions (including arbitrary choices by people)" (Stone, 1965, pp. 349-50). Schlfiter made a distinction between form and process, and Stone suggests that he was probably the first to explain that the major concern of settlement geography is with the resulting phenomena of people's activities and not the people themselves. The German definition of settlement geography con- tinued to be quite broad. A good deal of confusion has arisen from an emphasis on form and process at all scales. As recently as 1961, a German settlement text defined the study of rural settlement to include ". . . the study of man's use of plants and animals to procure food and cloth- ing, of an economic area larger than the dwelling, and of direct relationships between the economic area and the dwelling” (Stone, 1965, p. 351). 17 The development of French settlement geography was similar to the German experience, yet notably different in its much narrower definition of the field. Vidal de la Blache, the founder of modern human geography in France, had a strong influence on early settlement work, and it was one of his students, Albert Demangeon who dominated the develOpment of settlement geography in France for the first third of this century. Demangeon's major work on this topic described the major distributional characteristics of settlement throughout the world (1927). Early French settlement work was focused princi- pally on settlement form. This narrower scope contrasts with the broader German interpretation. Stone notes that (more recent work in France has tended to rely on a broader definition, yet with continued emphasis on the dwelling. Elsewhere in the world, definitions and themes in the study of settlement vary widely. In Belgium, M. A. Lefévre, a student of Demangeon's, has stated that the focus of settlement geography was the definition, classi- fication, and explanation of house type distributions. She also emphasized the process of settling as a part of the analysis of form. In Scandinavia, settlement studies seemed to stress the process of settlement, and investi- gations into delimiting the succeeding stages of settle- ment. Early English settlement studies were dominated by l8 analyses of house types, morphology of villages and field patterns. The development of settlement geography in North America has been slow and restrained. Few definitions of the field have appeared and little attention has been given to the development of a theoretical framework. The first notable settlement work done in America was Isaiah Bowman's "pioneer belt" studies (Bowman, 1931; Bowman, 1932; Bowman, 1937). Bowman's science of settlement was designed to be an interdisciplinary investigation of areas for potential colonization, and was quite popular during the 19205 and 19305. Particular attention was given to different pro— cesses of settlement and areas which had recently been or continued to be settled. The semi-arid belt of the Great Plains and the cool margins of the northern forests were popular areas of investigation. G. T. Trewartha narrowed the focus somewhat explaining that settlement geography was the study of house types and ". . . the characteristic grouping and arrangement of these buildings into coloni—' zation or occupance units called settlements; . . ." (Finch and Trewartha, 1936, p. 620). Trewartha defined "house" loosely, including any building that people live or work in. A major assessment of settlement geography was included in the 1954 Association of American Geographers volume American Geography: Inventory and Prospect (James 19 and Jones, 1954). Kohn states in the second paragraph of this assessment that, "in general, settlement geography has to do with facilities men build in the process of occupying an area" (Kohn, 1954, p. 125). Kohn divides the study into two halves, the examination of the process of occupying pioneer areas and the examination of settlement form and features. He sidesteps the first approach mentioning Bowman's work, and explains that there is "a growing interest" in the latter approach. In Kohn's review of what he calls "studies of specific facilities and their grouping" he considers (1) studies of archi- tectural style; (2) studies of roads and properties; and (3) studies of settlement ensembles. Clearly then, he felt that geography was moving away from studies of process towards a narrower view of form and classification. And yet, Kohn is quick to point out: Geographers do not examine architectural styles, roads and properties, and settlement ensembles only for the sake of identifying new categories or develop- ing new classifications. To be sure, there was a time when geographers, newly aware of the possibilities of detailed field mapping, limited themselves to the study of shapes. But in recent decades, studies of the facilities which men build have been undertaken for one of two purposes. One is for the light they throw on historical sequence-~studies of origin; the other is for the light they throw on functional relations-— studies of functions (Kohn, 1954, p. 136). Yet Kohn encourages the classification work nonetheless, at the expense of the analysis of the development of the styles, and properties and ensembles. 20 As to "The Prospect For Settlement Geography" and "Special Problems to be Solved," Kohn suggests "the com- parative study of settlements in different cultural areas . . ." and ". . . compilation of world maps showing the areas of individual farmsteads and the areas of compact farm villages, and of various combinations of these basic types" (p. 138). Kohn begins his article despairing that ". . . no analytical framework has been developed for settlement geography comparable to the principles of location in industrial geography . . ." (p. 125). His suggestion for further research would certainly not aid in the develOpment of such a framework. Kirk H. Stone recognized the great confusion which had arisen from the multiplicity of definitions and the extreme breath of settlement geography, and attempted to begin an initial ordering of the field with "The Develop- ment of a Focus for the Geography of Settlement" (1965). Stone explains that there is not only disagreement within the discipline internationally, but also etymologically. To create order out of chaos, Stone suggests a number of Specific definitions: It is suggested that the geography of settlement be defined as the description and analysis of the distribution of buildings by which peOple attach them- selves to the land. Further, that the geography of settling designate the action of erecting buildings in order to occupy an area temporarily or permanently (p. 347). 21 Stone then continues the emphasis of settlement geography on the dwelling and includes other buildings which are a direct ". . . tangible expression of man-land relationships . . ." (p. 347), such as barns and equipment sheds. "But excluded as a central subject of settlement geography are elements of the landscape such as functions of people, fences, land use, and lines of circulation and communi- cation; these are considered whenever essential to the analysis of distributional patterns of buildings, but as central topics they are assigned to other divisions of geography" (Stone, 1965, p. 347). Stone recognizes the difference between urban land and rural settlement. "It is suggested that the geography of rural settlement be defined as the description and analysis of the distribution of buildings by which people attach themselves to the land for purposes of primary pro- duction" (p. 347). This definition implies that the principal activity in an area of rural settlement might be farming, mining, trapping or fishing, and that the people would live in small (less than 200 buildings) clusters of individual dwelling units. Stone defines urban settlement as the larger grouping of buildings dominated by secondary and tertiary activities. Terry Jordan responded to Stone's article in "On the Nature of Settlement Geography" (1966). He has two specific criticisms. Jordan complains that Stone's 22 emphasis on ". . . the description and analysis of the distribution of buildings by which people attach themselves to the land . . ." (Jordan's italics, p. 26) ignores building types, including their design and construction materials. He argues that traditionally this has been a part of settlement geography. Secondly Jordan objects to Stone's emphasis on buildings, an emphasis that would leave the study of fencelines to agricultural geography but give barns to settlement geography. This division, he suggests, is an attempt ". . . to draw too short a circumference around settlement geography, . . . to mark its borders too sharply" (p. 26). As an alternative definition Jordan has offered ". . . the study of the form of the cultural landscape, involving its orderly description and attempted expla— nation" (1966, p. 27). He elaborates on "the form of the cultural landscape": It is synonymous with settlement morphology and includes (a) vertical arrangements and dimensions (such as the number of stories in a house), (b) hori- zontal arrangements and dimensions (such as the dis- tribution of buildings, the floor plans of houses, or the pattern of fences and fields), and (c) the material composition (such as brick vs. wood in house con- struction or line hedges vs. wire fences) (1966, p. 27). Jordan argues that it is the emphasis on form which is the primary element of settlement geography. The confusion and lack of agreement on definition is due, he feels, to a lack of recognition of the ". . . fundamental, central role of form in settlement geography . . ." (p. 27). 23 Stone replies (1966) in an effort to clear up what he sees as a misunderstanding. Stone explains that his article was designed as a "development of a focus" while Jordan's was a comment on the "nature of settlement geography." Stone is trying to produce a lowest common denominator for the study area to forge an agreement on its core theme. Jordan, Stone explains, is concerned with the peripheral boundary definitions. Stone is quick to add that by ". . . the description and analysis of the distribution of buildings by which peOple attach themselves to the land . . ." (p. 208) he meant that description includes classification of house types, form and construc- tion materials. If fences are important to an explanation of the distributional pattern of buildings Stone assures Jordan that that is quite an acceptable topic. Again, Stone had endeavored to establish the basic core of investigation. He explains Jordan's definition as one which attempts to delineate the circumference of the field. More insightful perhaps is Robert D. Mitchell's letter to the editor of the Professional Geographer in the same issue as Stone's reply (1966). Mitchell wonders if the study of form is not a shallow basis for a defi- nition of rural settlement geography. "Function must be added to provide reasons for form and distribution. That is only part of the problem. What would Messrs. Jordan and Stone do with process, the 'settling' behind the 24 'settlements?'" (Mitchell, 1966, p. 198). Indeed, enough attention has been given to whether fence types are part of settlement geography, and little thought has been given, since Barrows, to the process behind this form. It would seem that an understanding of process is important before an analytical framework may be develOped for settlement geography "comparable to the principles of location in industrial geography . . ." (Kohn, 1954, p. 125). American settlement geography, then, has charac— teristically focused on form and description, a narrower view than that taken by geographers in other countries, the Germans for example. Settlement geography has become synonymous with rural settlement geography in light of the major developments of urban and economic geography based on central place theory. For the purposes of this study, Stone's original definitions, though unfortunately linked to descriptive studies of distribution and form, will be used to encourage a codification of terminology. There- fore, in this study ". . . settlement refers to one or more buildings at a place and settling designates the action of erecting buildings in the use of an area" (Stone, 1965, p. 348). If settlement geography is ". . . the description and analysis of the distribution of buildings by which people attach themselves to the land" (Stone, 1965, p. 347), the study of the process of settlement and 25 the human behavior characteristics of such a process must be included within such a definition. Development of a Theoretical Framework for Settlement Geography Settlement geography with its ancient heritage and modern variances has long lacked a theoretical framework. This has been an oft noted and despaired fact. Yet within the last two decades a few initial attempts have been made to develop a conceptual framework for the study of settle- ment. Most of these have been attempts to create models which might predict settlement form, but not explain the process itself. Erik Bylund's 1960 article "Theoretical Consider- ations Regarding the Distribution of Settlement in Inner Northern Sweden" (Bylund, 1960) was the first attempt at generalization in settlement study. His work in the central Lappland area of northern Sweden on historical settlement before 1807 led him to construct four simple models of the way settlement moved in "waves" in this area (Figure 1). Each of these models relies on the assumptions of equal physical conditions and that further areas will not be settled until those closest to the "mother settle- ments" have been occupied. In an endeavor to add a degree of reality to his deterministic model, Bylund assigned attractive weights to a road, a church, and the parent settlement. These models rely on a process Bylund calls 26 Colonization development - origin - I“ stage 2"‘stoqo m 3" stage Fig. l.--Bylund's Four Theoretical Models of Settlement Development (after Bylund, 1960). 27 clone colonization, whereby a few pioneers initially locate in a region and further settling is carried out by their sons and daughters, and the next stage by their sons, etc. Bylund's final model, as illustrated in Figure 2, makes an initial attempt at the identification of stages in the settling of an area. He suggests that wasteland, or land parcels too small to support a settler at an earlier stage will be occupied during a later stage when the demand for land rises, and technical innovation allows the settling of smaller pieces of land. Thus, Bylund's model shows the settlement of the interstices of earlier generations by later settlers who are produced from all settlements not simply the most recent. Bylund ends with: Fig. 2.--Bylund's Refined Model of Settlement Development (after Bylund, 1960). 28 In order to adapt the models still better to reality, attention of course, must also be paid to other facts which have not been discussed or considered here; amongst other things, the natural advantages of the settlers' lands, which vary as between each other, e.g., concerning the occurrence of good soils or of profitable fishing lakes. It is, however, obvious, that the very complicated pattern of the spread of settlement does not in every case admit of explanation by physico-geographical conditions alone, however important these may otherwise be (Bylund, 1960, p. 231). John Hudson is the major innovator in the area of theoretical settlement geography. His dissertation entitled Theoretical Settlement Geography was completed in 1967, the major points of which are embodied in his article which followed two years later "A Location Theory for Rural Settlement" (Hudson, 1969). Hudson presents a theory which attempts to explain the changes of rural settlement over time. In the construction of his theory, Hudson relies on concepts from central place theory, diffusion studies, ecological distribution laws and morphological laws. The model attempts to emulate the settlement process and he specifically addresses himself to the assumption of central place theory that farm distribution is uniform. Hudson's theory rests on his postulation of three Specific phases of settlement development which are based in plant ecology. These phases are: l. Colonization, where a species penetrates a new region, and extends its habitat outside the limits of its old region. 2. Biological renewal, where a species regenerates through an increase in numbers, encouraging short distance dispersal filling the interstices of previous areas. 29 3. Competition, where weak individuals are pushed out due to the limitations of the environment, the pattern stabilizes as density increases. It is the first stage which is of the greatest interest here. Colonization is primarily associated with the spread of Settlement into a yet unsettled area, or new environ- ment. The characteristics of this new environment may be best interpreted, Hudson feels, by utilizing a number of concepts from ecological modeling. He explains that the density of human settlement at any one point can be thought of as a function of m environmental parameters, from which there may be derived a group of n statistically independent variables. If we think of each of these variables as vectors in n-dimensional space (n-space), each vector is by definition orthogonal to all others. These n variables may take on widely varying values reflecting different environments and the differing levels of the m environmental parameters. Given the values of these variables, we may imagine a form or volume delimited in this n—space by the intersection of n-planes defined by the value of each vector. This hyper-volume is defined as Niche space, or N. Hudson calls it the "fundamental niche of the population" (Hudson, 1969, p. 367). "Each vector.in this space has h components, defining a certain combination of values of the environmental variables. It is a familiar fact that there exist vectors (xil, x12, xi3, . . . x. ) in that are too extreme to permit human settlement" (p. 367). 30 These vectors represent hazard environments which might be too cold, too dry or too hot for settlement. An important concept is that niche space is not a tangible mappable space. Ecologists speak of biotope space (B), which is a mappable physical representation of the varying levels of the variable of niche space (see Figure 3). A function may be defined that translates niche space to biotope space: 5 (x1, x2, x3, . . . xn). It is not necessarily a single valued function; though permissible levels of n variables may exist, there is no assurance that these variables actually have physical manifestation in an area. Crudely, it might be a nice place to settle, but perhaps no one has chosen to do so. 1Y3 l T i g E I -—--<'.---~-—-1-- . I/ Fmdmmm : AflflMdeP I ‘1 N10 . ' , ) .4,,, I FUmflhyLfl _ : ///| // [b ”1'" —' L“ , ,/"/ '//" :/ )C-z. , 5hr] " ’52 r 7 ‘ 24rw' ‘*-~/’ /" /’ ” “7’“T't‘7{fs"‘ /. )f/ _ x, "’1 ’ /,. - ,IMCHE / ,1 SPACE Fig. 3.--The Magnitude Function, f(xl, x2, x3) Determines the Density, (2), of Settlement in Any Area, dB, of the Biotope (after Hudson, 1969). 31 To illustrate the difference between niche and biotope space, and their boundaries Hudson explains: Limits in the niche are counterparts of dis- tributional boundaries (regions) in the biotope. An example of a limiting value on a variable in the niche space is minimum farm size—-the size beneath which agricultural operations are economically unfeasible. Tree line, on the other hand, is a boundary in the biotope space, corresponding to some phytophysiological limit in the niche space (Hudson, 1969, pp. 367-8). It is important to note that these vectors do not have an equal influence on settlement. Indeed, gradients are quite important. "In general the function 6 defines a mapping that determines the density of settlement in the biotope space and is called the magnitude function" (p. 368). The different components of the function exert varying weights on its outcome. This function provides the density of settlement for an area, but not its pattern of settlement. Spread, Hudson's second phase, is similar to Bylund's recognition of another new stage where numbers increase and a filling of the gaps occurs. This process produces an increased density, that is, the magnitude function allows the biotope to grow vertically. Citing Bylund and other studies of larvae spread, Hudson accepts the assumption that successive generations show a limited spread away from the parental settlement. Yet Hudson is quick to point out that spread is not the only source of growth, as Bylund suggests, but also includes continued 32 in-migration from outside the area. "More important geographically, there is no reason to expect this new immigration to cluster around the settlements of the pioneers" (p. 370). This all suggests to Hudson a "greater regularity in the spacing of farmsteads, rather than clustering" (p. 370). Hudson feels that it is spread which encourages a cluster in settlement pattern and colonization which encourages regularity. Competition occurs when the biotope is completely settled. Weak individuals are forced out and the pattern begins to stabilize. This process of competition tends to produce greater regularity in the pattern and in turn pro- duces one condition for the development of a uniform network of central places. Hudson compared his model with actual settlement patterns from six Iowa counties. He concluded that in the early stages of settlement when the density of settlements is low and unsettled areas are common, the location of settlements is essentially independent of one another. As density increases competition increases and the settle— ment pattern changes from the clustered to a more regular pattern. Hudson therefore sets a constraint upon the assumption of central place theory that rural farm patterns are regularly spaced, by maintaining that a certain degree of competition is necessary. 33 A fascinating recent study by Norton (1976) is one of the first efforts to develop a stochastic simulation model of settlement. Norton endeavored to simulate popu— lation levels by township for southern Ontario utilizing the variables availability, distance to the nearest entry point, land quality and potential of each township. Norton's stated aims include an attempt ". . . to isolate the principal variables involved in the process of settle- ment, to construct settlement patterns for periods for which data are limited, and to produce patterns which might have developed given particular processes" (Norton, 1976, p. 270). That is, Norton is not only interested in the influence of the variables on the settlement process, but also in creating a predictive tool to simulate the settle- ment of areas and periods where data do not exist. Norton is also interested in being able to ask "What if . . ." questions. Norton's location process involves the calculation of an index of attractiveness (Ai) for each township utilizing the four variables of township character. The probability of a township receiving a settler may be then calculated from the index of attractiveness. No attempt was made to differentiate between locations within a township. Norton's function is based on four variables: 34 where: Ai = the attractiveness of the ith township Ii = the index of availability of the ith township Si = the distance between the ith township and the nearest entry point Qi = a measure of the land quality of the ith township Vi = a measure of the potential of the ith township Norton's attractiveness function is quite similar, though he does not point this out, to Hudson's magnitude function. Norton's variables can be seen as representing a four- dimensional niche space and the attractiveness function translates this into the probability of density in biotope space. Recognizing the potential differing influences of these variables, Norton simulated the settlement process many times changing the input of each variable. The general form for the calculation of attractiveness values that Norton used was: where the constant h is assumed to be equal to one, the attraction being scaled in a similar manner for all town— ships. The letters a, b, c, and d represent exponential constants, and it was these constants that Norton varied to change the influence of each variable. Arbitrarily, Norton decided that the exponents could take the values of 0.0, 0.5, 1.0, or 1.5. Given four variables with four possible exponential weights there were a possible 256 35 different simulations that Norton ran in an effort to determine the levels of influence of each of these vari- ables. This is a specific recognition of the importance of gradients in n-space as pointed out by Hudson. The probability of the ith township receiving a settler (Pi) was determined from: where m is the number of townships available for settle- ment. An assumption of Norton's model concerning the calculation of the index of availability is of particular interest here. In determining the availability of a township Norton decided that there was a maximum density of one settler per 100 acres. He calculated the number of available 100 acre locations at the beginning of each time period. To find the availability index he used: where: Ii = Index of availability for the ith township T. = Total number of 100 acre locations in the ith township 0. = Number of occupied 100 acre locations in the ith township Initially then, Oi equals zero; when maximum density occurs the value of Oi equals Ti' and Ii equals zero. Norton's major assumption then, is that "as the number of available lots in a township declines the attractiveness declines, 36 so that with no lots remaining the attractiveness is zero" (Norton, 1976, p. 274). This is identical to assumptions made by Hudson in the construction of his simulation: (1) The probability of a settlement occurring in an area, a, is a function only of the size of a, not its location in the study area; (2) the number of settlements in any small part of the study area is independent of the number falling in any other area; (3) the probability of more than one settlement occurring in a, approaches zero as a approaches zero, faster than a does (Hudson, 1969, p. 374). These assumptions, that are based on the central place theory premise of uniform dispersal of farms, are seemingly based on the specific North American experience and need to be examined further. Grossman (1971) objects to Hudson's use of bio— logical theory and argues that Iowan farmers with their high degree of individualism and their complex and diversified origin are unrepresentative of rural settlers. The implications of Hudson's and Norton's assumption are that an area will be completely settled at a broad scale, and the next stages of settlement will fill in the large gaps. Hudson suggests that ". . . it seems likely that new settlement would be somewhat repelled by the earlier settlement, under conditions of contiguous landholdings of approximately equal size typical of most homesteading in the United States" (Hudson, 1969, p. 370). This con- tradicts Bylund's observations of settling in Sweden where he has talked of clone colonization. That is, settling 37 that occurs in waves out from a focal point or focal region. Grossman's studies of clustered settlement in Nigeria have urged him to argue against the universality of Hudson's assumptions of dispersed farmsteads. The difficulty with the assumption of unclustered settlement is that time after time empirical studies have shown that for one reason or another early settlements are clustered (Bylund, 1960; Peters, 1969; Grossman, 1971). This might be explained by a common perception of the regions by settlers, by an extreme vector limit in niche space or a physical constraint of the region. It must be remembered that not always the best areas, even according to contemporary appraisal, were those first occupied. Settlers, in small numbers and with limited techniques at their disposal, are attracted to those areas which they are capable of managing and controlling. The physical characteristics of these areas are certainly closely relevant to this, but the concept of manageability is important especially where, for example, the control of water (power) supply, or the regulation of drainage, or the clearance of vege- tation is involved (Paget, 1960, p. 325). In Norton's model the simulated patterns which best fit reality are quite clustered. Not surprisingly there are clusters of settlement in the areas we now call Windsor, Niagara, Toronto, Kingston and near Montreal. How did these results occur given Norton's initial calcu— lation of an index of availability? The influence of the other variables was to counteract his assumption of availability. For example, one of the variables used, Distance to the nearest entry point (Si)’ was the 38 straightline distance between each township and the near— est entry point. The entry points used were the Detroit River, Niagara, York (Toronto), Kingston and the Quebec border (near Montreal). Indeed, Norton found this vari- able to be the most influential in the sense that of the 256 simulation runs, the patterns which most closely matched reality had used an exponent value of 1.5 for the distance variable, while the other variables generally had exponents of less influence. In his conclusion Norton recognizes that "Gentil- core noted that, during the early years of settlement, location was dependent upon the entry points and the availability of surveyed land" (Norton, 1976, p. 286). Indeed, Norton grudgingly admits that "the lot availability variable, . . . is formulated assuming a linear relation- ship between the available lots and the township attrac- tiveness. This is possibly unrealistic as townships with a minimal number of lots remaining might prove very attractive as they represent available land within devel- oped areas" (p. 286). In fact, not only does initial settlement in south Ontario reflect the date that land was surveyed, that is, land was settled as it was surveyed, but also that "the pattern begins as a series of cores and subsequently develops throughout the area, emphasizing several of the early cores" (p. 286). This seems identical to Bylund's observations in Sweden and to the idea of clone 39 colonization, and thus directly contradicts the avail— ability assumption that both Hudson and Norton made. An important building block of the model which will be formulated in this study is the recognition that early settlement tends to be clustered. This clustering may be a result of the movement of a settlement frontier, of a common perception as to the suitability of settling a certain type of land, e.g., river valleys, of clone colonization, of similar cultural origin, or of the proximity to entry points, military outposts or capital cities. In any event the suggestion that the settling of an area is an inverse function of number of settlements already there displays unfortunate ignorance of the impor- tance of cultural ties, behavioral adjustment and environ- mental perception. CHAPTER III NATURAL HAZARD RESEARCH AND MODELS OF MAN- HAZARD INTERACTION The Development of Natural Hazard Research Natural hazard research is a relatively recent development within geography. Although its origins may be traced back to Barrows and his interest in human adjustment tx: the environment (1923), natural hazard studies did not come into full bloom until a few decades later. "Natural hazards are those elements in the physical environment, harmful to man and caused by forces extraneous to him" (Burton and Kates, 1964, p. 413). Natural hazard research focuses on the interaction of man and nature, and the governing human—use system and natural event system. Thus natural hazard studies do not restrict themselves simply to the characteristics of natural events, but search out the relationships of the events and the human occupation of the affected area. This research is concerned with the question "How does man cope and adjust to the risk and uncertainty evident in hazardous geophysical systems?" Early natural hazard research was limited to studies of flood hazards, and even today more is known 40 41 about man's relationships with floods than any other hazard. The earliest work done in the area of flood hazard was carried out in the 19203 and 19303 when Congress dele— gated the Corps of Engineers to investigate the manage- ability of the country's river basins for reasons of flood control as well as irrigation, hydroelectricity, and navi- gation. Many of these initial reports were quickly adopted. Because of their appearance during the Depression, the proposals were designated public work projects. Geographers were actively engaged in these projects. Gilbert White, the founding figure of natural hazard research, was spurred to survey the alternatives involved in attempts to reduce flood loss. His Human Adjustment to Floods (1942) was the first in a continuing series of hazard related monographs by White and his students to emerge from the University of Chicago Department of Geography in its Research Paper series. In 1936, following a series of damaging floods, Congress authored the Flood Control Act of 1936. This act declared Congress' intent to grant financial support to any flood control project whose financial benefits out- weighed its cost. Twenty years later, having spent five million dollars, a geographic investigation was organized to determine the changes in urban flood plain occupation that had resulted from these programs. The investigation focused on seven representative sites. The variety of 42 adjustments utilized at these sites in response to the flood hazard were classified as actions that were designed to (a) modify the cause of the hazard; (b) modify the losses caused by the hazard; or (c) distribute the losses. A number of disconcerting conclusions emerged from the investigation (see: White et a1., Changes in Urban Occu- pation of Flood Plains in the United States, 1958). While flood-control measures had increased, so too had the amount of flood damage. The general goal of reducing flood damage had not been realized and there had been increased human occupation of the flood plains. The study also showed that the federal government's emphasis on flood-control and upstream management excluded other possible adjustments. It was clear, then, that the federal programs had failed, and that there was a critical need for increased under- standing of human occupation of flood-prone areas (White, 1973). A number of studies were begun on other aspects of floods as a natural hazard. Agricultural use of flood plains was reviewed by Burton (1962), coastal flooding along the eastern seaboard by Burton, Rates and Snead (1969), efforts to improve flood damage estimation and the evaluation of flood control efforts by Rates (1965), and methods for handling flood losses by Burton (1965). These studies established natural hazard research as a viable sub-field within geography. Five principal areas of 43 investigation were developed as a result of these studies and continue as a basic description of natural hazard research. These areas are: (l) assessing the extent of human occupation of hazardous regions; (2) identifying the complete range of adjustments to the hazard; (3) the investigation of human perception and evaluation of hazards; (4) describing the hazard adjustment adoption process; and (5) estimating the optimal group of adjustments, and their social and environmental consequence (Burton, Kates and White, 1968). Much of this new wave of hazard research was characterized by a healthy amount of interdisciplinary research work. Engineers were involved in analyses of structures and their flood resistivity. They were fre— quently the most responsive group in applying the lessons of geographic investigations. The Army Corps of Engineers was often involved as well as hydrologists from the U.S. Geological Survey. Economists were needed in advanced cost-benefit analyses, and investigations into the economic effects of zoning controls. Psychologists were drawn into the foray to help explain risk behavior and the perception of uncertain environments. From the beginning, natural hazard research has been unique in geography for its practical, applied value--a characteristic becoming more and more prized--and the degree of input by geographers in a broad interdisciplinary research effort. 44 By 1969, the importance of natural hazard research was recognized in a number of major ways. The International Geographic Union Commission on Man and Environment committed itself to hazard research as one of their two major areas of development in the following three years. The Commission played an important role in encouraging and coordinating hazard research around the world. Equally indicative of hazard research's acceptance within the world was an international seminar on flood problems held in Georgia, USSR, for three weeks in the fall of 1969. White commented soon after the conference that it was of special interest to geographers for two reasons. "It was the first system- atic recognition by the United Nations community of the importance of dealing with water-resources management in a way that takes account of the full range of alternatives open to man. It also brought into the discussions the findings and methods of analysis of geographers in close association with hydrologists, engineers, and economists" (White, 1970, p. 440). The seminar was sponsored by the Soviet Ministry of Reclamation and Water Management with the Georgian Scientific Research Institute of Hydraulic Engineering and Reclamation, and the United Nations' Transport Division of the Economic and Social Affairs Department. Twenty-eight countries were represented with participants from UNESCO and WHO and an array of Soviet research institutes. Among other topics, attention focused 45 on the range of alternatives for managing flood damage, and there were proposals for continued international collabo- ration. Geographers from a number of countries held a central role, both as United Nations' consultants and as authors of papers chosen for discussion (White, 1970). Recent developments in natural hazard research are indicative of a broadened, more advanced research effort. Investigations have branched out into a myriad of hazards other than floods including droughts, hurricanes, tornadoes, avalanches, earthquakes and tsunamis. Besides the cross-section studies of one hazard, investigations of all the hazards of an area, that is a regional hazard ecology, emerged (e.g., Hewitt and Burton, 1971). Behav- ioral science methodology became a prominent tool in the investigation of hazard perception and behavioral responses to hazards. Studies such as Saarinen's use of the Thematic Apperception Test in his analysis of Great Plains farmers' perception of drought (1966), and Sims and Bauman's use of sentence completion tests to contrast tornado perception in Alabama and Illinois (1972) were landmark works. These studies borrowed standard techniques from psychology and utilized them in ground-breaking research in the areas of environmental perception and natural hazards. Perhaps the single most important development in hazard research is the recognition of the key role of environmental perception in the adjustment process. 46 The importance of perception became clear in the re-evaluation of the government's flood policy in the late 19503. Much of the failure of the programs was seemingly due to the obvious differences in hazard perception by different respondents, such as between resource managers, a home owner or shopkeeper for example, and flood research scientists or hydrologists. To better understand the process of human adjustment to flood hazards it was clear that more needed to be known about hazard perception. Early investigations of attitudes toward flood hazards and hazard perception were reviewed by Burton and Kates (1964). The key to natural hazard research then became twofold: it was, of course, particularly important to understand the behavioral aspects of human occupation of hazard zones, but an emphasis on hazard perception and its relationship with the behavioral adjustments was a prerequisite to understanding this behavior. Models of Human Adjustment to Natural Hazards In an attempt to understand the relationships between hazard perception and adjustment adoption, a number of conceptual models have emerged. These models are rudi- mentary attempts at developing an explanation of the basic relationships involved in the decision making process of hazard adjustment. The assumptions of the earliest flood control projects were based strictly on economic 47 optimization. The cost-benefit analyses were based on an optimizing model which assumed complete knowledge of the hazard, and optimal economic adjustments on the part of the resource managers. The limitations of this model are obvious. A modified model, which White has called a ". . . subjective utility model," relaxes the complete knowledge assumption (White, 1973, p. 199). This model recognizes the subjective role of environmental perception and the subjective evaluation of the possible results of any adjustment. However, this model retained the optimi- zation assumption such that if a resource manager's per- ception of a hazardous area might be ascertained, it would be feasible to predict his behavior. The model assumed that a person would undoubtedly optimize the benefits of his/her environment within the personal constraints imposed by the role of perception. Neither of these models, however, were particularly helpful in explaining the behavior actually observed. While it was obvious that people in hazardous areas recog- nized differing levels of hazard from one part of the area to another, this recognition was not necessarily trans— lated into a behavioral response. It was not uncommon for people to return to areas where they experienced high per- sonal damages and extreme financial loss rather than move elsewhere close-by where they recognized that the danger was less. The models failed because of their key dependence 48 on the Optimization assumption. The need to adequately explain the behavior of occupants of hazard zones, and predict the behavioral response of these occupants to changes in policy required the creation of a new model. White has explained that "the obvious direction in which to move was the model of bounded rationality . . ." (White, 1973, p. 200), as developed by H. S. Simon (Simon, 1957). Kates' work in Lafollette, Tennessee (Kates, 1962) helped develop a model of bounded rationality for hazard decision making. Kates investigated the behav- ior and expressed perceptions of residents of a Tennessee river valley. He endeavored to find out how people per- ceive hazards, how they perceive the range of adjustments open to them and what factors might explain the varying perceptions of the same environment. White presents "a rough model of decision" (1973, p. 201) that illustrates the early attempts to create a model of the decision making process. This model recognizes the inputs of the environ- mental system and the social system on the decision maker (see Figure 4). The most recent version of a model of natural hazard decision making and behavior is found in Kates' "Natural Hazard in Human Ecological Perspective: Hypotheses and Models" (1971). Kates explains that: . . . it is only now that we can begin to struc- ture a primitive general framework of human adjustment to natural hazard, in which we try to preserve its human ecological perspective. In this perspective, 49 Social System Occupance Decision situation Range of a . . . n adjustments Decision Maker Perc:§:::2n:22twelghlng 0f: Ezggnence 9 Technology > Choice - Economic effect Personality Social linkages Environmental System Magnitude Frequency Duration Temporal spacing Fig. 4.—-A Rough Model of Decision (after White, 1973). with its focus on man as the ecological dominant, the interactions between men and nature tend, over the short run, to be stable, homeostatic, and self- regulating and, over the long run, to be dynamic, adaptive, and evolutionary in the direction of increasing control over nature's resources and buffering from nature's hazards (Kates, 1971, p. 438). Kates' purpose is to present a basic systems model of the short—run process of adjustment. He lists a number of basic hypotheses which sum up the present state of knowledge of hazard research and help provide a basis on which the model may be built. These hypotheses help explain the nature of natural hazards, adjustments to the hazards and the individual choices made by those involved. These hypotheses are listed below under Kates' sub-headings. A. Man-nature interaction. A natural hazard is a result of the interaction between a human-use system and 50 a natural system. It is a result of the basic process of man pursuing that which is beneficial and avoiding that which is hazardous. Clearly there are conflicts in per— ceived benefits, and frequently people accept hazardous situations knowingly due to their perception of net per- sonal benefit. B. Techno-social stages. Adjustments to hazardous situations may be categorized into three techno-social stages. Each of these stages has a preferred group of adjustments, and a distinctive choice process, primarily determined by cultural characteristics. Folk or pre-industrial adjustments are character- ized by mystical or unrational behavior or have been built into the culture as a stress relieving system (e.g., see Hsu, "The Cultural Ecology of the Locust Cult in Tra- ditional China," 1969). These adjustments require alter- ation of human behavior rather than control of the natural system. They require little capital, and may be imple- mented by small groups or individuals. Modern technological or industrial adjustments are characterized by a very limited number of actions relying on technological efforts to control the workings of the natural system. They are capital intensive and require community-wide action. Comprehensive or post-industrial adjustments are characterized by features of both of the previous stages 51 allowing a greater flexibility of action. They allow a variety of capital inputs and individual involvement and organization. C. Hazard Differences. Given the above stages of hazard response, the character of the hazards themselves necessarily varies. That is, given different levels of cultural and technological development, the hazardousness of different natural systems varies. Four specific attributes of hazards are key to these differences. It is the variation of these attributes which gives rise to different adjustments. Three of these attributes are characteristics of the event itself: frequency of occur— rence, magnitude of event, and suddenness of the onslaught of the phenomenon. The fourth attribute is whether the hazard is intrinsic to the purpose of the human occupation or is not related to this activity. D. Decision maker differences. The choice of adjustment process may vary given a specific hazard, between decision makers. The choices may be individual or col- lective ones. While the "management unit" may differ, from a house to a city, ". . . the ways in which choice of adjustment is made does not fundamentally differ" (Kates, 1971, p. 440). E. Individual differences. All individuals who choose an adjustment to a hazard perceive the hazard and are aware of a range of adjustments. These adjustments are 52 evaluated with reference to their economic benefit, feas— ibility and social suitability. And yet, while the decision process is similar, each perceives the hazard differently, is aware of a different range of adjustments and uses different criteria in the evaluation of the adjust- ments. Perception of the hazard may vary according to: the characteristics of the hazard, personal experiences with the hazard, and individual personality factors. The general outline of Kates' model appears in Figure 5. The model is ". . . only a small slice of the global system for which the above hypotheses represent the first step towards a theoretical formulation" (Kates, 1971, p. 443). The model is of a system at a single cross- section of space and time. That is, at a specific place for a specific moment, man and nature interact through their governing systems to produce a natural hazard. This hazard produces a specific set of hazard effects. The adjustment process governs the choice of adjustments that modify the natural event system, modify the human use system or modify the hazard effects through emergency adjustments. A more detailed representation of the model appears in Figure 6. To provide a greater understanding of the model as a whole each element will be discussed in terms of its relationship with the other elements. Important characteristics of the human use system include descriptive data concerning human occupation of .xasma .mmumx umummv Howe: mEopmmm amumsow a mo mmcflauso .moumNom Hondumz on unmeumsnpd smasmIi.m .me _' I I I I I I I I I I I I I I I I I I I | I I I l I I I I l I I I I I I | I l I I l J 53 F"—"'_-"'_-f'_'_"_-'-'—"_"_' L mhszkwnoQ< mkummmw Qmw 4m whszPmDon—A‘ m...zm>w 4uzwomm2m mULPwmehuw I. mm: 242:: do wu_hm~mmhum s<¢sa“ 555:2 _ i 3522: .10 8:255:25 l"'l _ pnwfimmmmla Ipzwzpmssea $52.8 j _ am<~o¢pmua ‘quza_=am mmos mpuuaau ne<~w . zuhm>m flflw¥ . tuhm>m mhzu>w ._<¢:b»_mzmn saga any muuuoza Azuzsmssn< TIE uuua ...wo mm. ma.mmm mm.nve mo.m¢ma m cowmmmummm 0» use soaumwum> mo mousom Hmflaochaom mmumoo m “Om mocmwum> mo mflm>amc¢ «Hummmmmmom. mOImoaoommN. mOImmmHmmmH.I mucmauflummou cofimmmuomm Ho+mmwooomm. udmoumucH m mmummn mo :oflmmoummm HmflEOC>Hom om.mQMH mm Hmuoa vm. Hm.mm Hm cofimmoummm usond COHuMfl>mo mo.n>m mm.mom mn.a>o m¢.mvma m scammoumwm 0» man GOwucflum> mo mousom Hmflfioc>aom mmumwa N How mocmwun> mo mflmxamsd mOImomvomwm. mOIMNovth~.I mucmflofimumou cofimmwuumm Ho+mmvoooav. umwoumucH m cosmos mo cofimmmummm Hoesosmaom om.m©ma mm Hopes mm.~a nm.mov mm scammuummm usonc conumfi>mo mm.mmm vn.on mm.omm mm.omm H coammoummm 0u use coHumwum> mo mousom AMHEOC»Hom mmummo a mom mocmflum> mo mwmxanqm moummmammmm. mucmgonummoo conmmwnoom Ho+mnommamn.u umwoumucH H common mo Godmmmummm Hmfleocmaom muumsvm mo 85m mo 05Hm> mumsvm moucsvm Eonmoum manna ca useEm>oumaH m com: «0 85m no common .cowmmoumwm Hassocaaom mo muHSmmmII.m wanna 100 immediately noticeable fact is the considerable improve- ment in the sum of squares figure between the first degree polynomial, a standard regression line, and the second degree polynomial, a basic curve. As we would expect from a series that we consider lognormally distributed, the improvement in the sum of squares drops considerably with the addition of another term. That is, a curve with two inflexion points was hardly able to improve on the basic one inflexion point curve's ability to approximate the data. Thus the results of the 2nd degree polynominal may be used to construct the equation: RI = 4.16 - .00218X + .0000000034X2 where R1 is the recurrence interval and x is the level of discharge for any given year. The quite small regression coefficients are simply due to the large discharge values (X), the small RI values, and using x to predict RI. The basic assumption of any polynomial regression is that random sampling has occurred and that the dependent vari- able is normally distributed. Because we have considered the annual series in a probabilistic manner as described by the Theory of Extreme Values (Gumbel, 1958) and because chance has no memory, an annual series is considered to be a random sample. The recurrence intervals are not at all normally distributed with i of 4.242, skewness index of 101 3.6, kurtosis of 13.877, and variance of 41.49. The assumption of normality was necessarily relaxed. With the development of the above polynomial equation a value of RI may be predicted each year upon the generation of a flood level. This allows the simulation to notify the user of not only the level of discharge, but also its recurrence interval and the probability of annual occurrence. These are, of course, much more meaning- ful to the user than the number of cubic feet per second. The flood level (X) is a status variable produced from an operating characteristic based on a log-normal probability function. It is a stochastic variable. Recurrence interval (RI) is also a status variable, but its level is determined by an internal tautological iden- tity. A flow chart representing the natural event system is found in Figure 20. The Adjustment Process In the development of the simulation model the translation of the adjustment process subsystem was a particularly important task. While the natural event system provided the dynamic element of the model, and the focus of the model as a whole was on the process of settlement, the adjustment process was in many ways the backbone of the model. The treatment of the perception of the hazard event, and the role of accumulated perception were the key elements in the behavioral feedback to the Fig. 102 ( 3 .J E) GENERATE FLOOD WMMATE CALCULATE RECURRENCE INTERVAL CALCULATE PROBABHJTY OF ANNUAL OCCURRENCE RETURN 20.--Flowchart of EVENT Routine. 103 settlement process. How was the perception of the annual flood to be simulated? How could the accumulated per- ception of the hazard system be handled? A basic assumption of the model was that medium- frequent medium-intense events were perceived as more hazardous events than any others, and that these events commonly contributed more to actual behavioral responses to the hazard system. A convenient way to translate the generated event into a perceived event, was through the use of a normal curve equation. By centering the normal curve at that area of the event scale which has been deter- mined to be perceived as the most hazardous, each event might be put into the normal curve equation. This would consequently weight medium-frequent events more, and weight the very frequent, and the quite infrequent events less. It was necessary to utilize the computed recurrence intervals in this procedure. The use of the RI value of each generated flood discharge level was more convenient in that rather than weighting discharge in cubic feet per second, the average occurrence in years could be utilized. More importantly however was that the whole concept of the importance of middle range events was based on the idea of a standardized perceived event system. By using the RI values the discharge levels are standardized in relation- ship to the system's most common event. A graphic 104 representation of the perception procedure may be found in Figure 21. The hazard perception transformation curve represents the value of the perceived hazard (H) at any given recurrence interval (RI). The accumulated perception of the hazard system was simulated by summing the annual levels of the perceived hazard factor for each settler. To simulate the effects of man's natural ability to cope with stress, and the importance of the recency of the event, each year a settler's accumulated perception of the hazard system (APHS) is reduced by an arbitrarily chosen 10 percent before that year's perceived hazard index is 40-1 30‘ 201 Perception of Hazard Index (H) IO‘ 8i 64 4. 2‘ I o:é& é éIo 20 in do so Recurrence Interval of Flood (Years) Fig. 21.-—Hazard Perception Transformation Curve. 105 added in. At the end of each year after the accumulated perception of the hazard system has been adjusted for the event of that year, each settler's APHS is compared with the mover/stayer threshold (M). This threshold may be determined by the user at the beginning of each simulated run. The user's choice is limited to low, medium and high perception thresholds. The low threshold is equivalent to experiencing three successive years of floods with a RI of five years, and medium threshold is equivalent to experi- encing three years of floods with a RI of ten years. The high threshold is equivalent to three years of floods with a RI of fifteen years. It was felt that typical rural agricultural units would be able to absorb both mentally and agriculturally two years of relatively hazardous events, but that the third successive year would be enough to make them move elsewhere. At the end of every simulated year each settler's APHS is compared with the user determined threshold (M). Should the settler's APHS exceed the threshold, that settler is assumed to pick up and leave the area, and is removed from the settlement plane. Both the perceived hazard factor (B) and the accumulated perception of the hazard system (APHS) are status variables, internally determined on the basis of tautological identities. The mover/stayer threshold (M) is of course an exogenous variable, determined by the user. A flow chart of the adjust process appears as Figure 22. 1106 ADJUST F‘ Q t CALCULATE HAZARDINDEX(H) FROM RECURRENCE INTERVAL is RELAX EACH SETTLER'S APHS av '096 t Aoo H'TO EACH SETTLER'S APHS YES AM6>M? \ NO SETTLER VACATES All RETURN Fig. 22.--Flowchart of ADJUST Routine. 107 The Simulation: Setsim The individual elements of the simulation as iden- tified above fit together to compose a user-oriented interactive stochastic simulation called SETSIM. The simulation has been written in BASIC, a FORTRAN—like language designed for interactive use, and is operable on the Computer Institute for Social Science Research's Hewlett-Packard 2000 mini-computer located in the Behavioral Science Instructional Laboratory, at Michigan State Uni- versity. The interactive user-oriented aspect of SETSIM suggests its use as a Computer-Assisted Instructional device (CAI). A flow chart of the simulation as a whole is pre- sented in Figure 23. A user, having begun running the simulation, is first asked if he/she needs an introduction to the simulation and an explanation of how it runs. This function is fulfilled by a routine called EXPLAN which orients the user to the simulation. It is felt that EXPLAN need only be called during a user's initial experience with the simulation. The next step involves the user determining the values of the input variables: number of years for the simulation to run (Y1); initial settlement pattern (P) and the mover/stayer threshold (M). Then the user is asked to specify the number of settlers that will arrive that year (N). In succession the following routines are called: SETTLE which locates the settlers (N) on the 108 SETSIM EXPLAN INPUT=YEARstvm PATTERN(PL THRESHOLD(M) IF—j INPUT= NUMBER OF SETTLERS (N) I SETTLE 4L EVENT \L ADJUST PRINT: PATTERN MORE YE ARS TO GO ? RUN SETSIM AGAIN ? YES STOP Fig. 23.--Flowchart of SETSIM Program. 109 plane, EVENT which produces the annual discharge level (X), and ADJUST which translates the event into a perceived event and updates the accumulated hazard perception and finds if any settler has exceeded the APHS threshold. At the end of the year the user has the Option to see the resultant pattern. The clock (T) is updated, and if the simulation is to continue for another year, the flow of the simulation is transferred back to the point where the user may determine the number of incoming settlers expected during the year about to commence. At the end of the simulation the user is given the option to begin a complete new simulation. A user‘s guide to the simulation, a sample SETSIM session, and a listing of the simulation program appear as Appendices. CHAPTER VI CONCLUSION In this study a conceptual model and computer simulation of the interaction between a natural hazard system and the process of settlement is developed. The model is put forth as an addition to the small but growing literature of settlement theory. It is designed as an explanatory model of the actual process of man settling the land and his interaction with a local hazard system. The model is also an attempt to portray the role of a hazard system in an historical process. Three questions were initially posed in the devel— opment of the model. These questions are: (1) In what manner is an area initially settled? (2) How often and with what intensity do natural hazards occur? and (3) What is the relationship between the settlement process and natural hazards? It was hypothesized that early settlements are usually clustered, that the frequency and intensity of hazard events have a linear inverse relationship, and that the behavioral relationship between settlement and hazards is dependent upon the settlers' hazard perception which tends to perceive events of moderate intensity and medium 110 111 frequency as requiring more serious adjustment than other events. It was shown that the process of rural settlement in its early stages does usually develop a clustered pattern. The frequency and intensity of natural hazard events, rather than exhibiting a true linearly inverse pattern, were shown to frequently approximate a log-normal distribution. Thus the inverse relationship holds, but logarithmic-probability paper is needed to straighten out the curve to approximate the hypothesized relationship. The behavioral relationship between settlement and hazard events is dependent upon the settlers' perception of the hazard system. The role of perception was to rationalize extreme events of infrequent occurrence, and accept events of common occurrence, yet interpret medium intense events of relatively frequent occurrence as creating a hazardous situation which required the most significant behavioral adjustment. The adjustment in this model was assumed to be the act of vacating the settlement location. Each of the answers to these questions is established as being grounded in sound empirical as well as theoretical research, and thus, become the basic theoretical assumptions of the model. This study, as an exercise in theoretical model- building, illustrates the key importance of clearly estab- lishing each assumption in existing theory. The utility 112 of any model is limited by the legitimacy of each of its assumptions and each assumption of the model presented in this study is shown to stand on sound theoretical under- pinnings. In the last analysis, the utility of the model presented depends upon its validation through comparison with actual real-world situations. Validation of any model or simulation is a difficult and time-consuming task, and is often impossible, or at least inconclusive. The vali- dation of this model is outside the scope of this study, and will be the focus of another project. However, the sound theoretical basis of the model allows it to be con— sidered as a small but significant addition to the litera- ture of settlement and hazard theory. Also the embodiment of the model as a simulation suggests its use as a computer- assisted instructional device. APPENDICES APPENDIX A A USER'S GUIDE TO SETSIM APPENDIX A A USER'S GUIDE TO SETSIM SETSIM is a stochastic simulation designed to model the response of rural agricultural settlement to a local hazard system. It is an interactive simulation, and has been designed as a Computer Assisted Instructional device (CAI). SETSIM is written in BASIC, and is operable on the Computer Institute for Social Science Research's Hewlett- Packard 2000 mini-computer located in the Behavioral Science Instructional Laboratory (BSIL), at Michigan State Univer- sity. SETSIM is designed to illustrate three specific relationships: (1) Dispersed rural agricultural settlement commonly exhibits clustered settlement patterns in its early stages of develoPment (Hudson, 1969), (2) The fre- quency and magnitude of hazard events often approximate a log-normal distribution (Wolman and Miller, 1960; Hewitt, 1970), and (3) Behavioral adjustment to a hazard system is dependent upon a person's hazard perception which tends to interpret moderately intense events of medium frequency as those which require the most significant response (White and Haas, 1977; Mitchell, 1974). 113 114 The simulation takes place on an idealized settle- ment plane similar to a game board. The plane is made up of 25 settlement "areas" which form a 5 * 5 matrix. Each area contains 16 settlement "cells" in a 4 * 4 matrix. One and only one settler may locate in any given cell. Each prospective settler is chosen a settlement area randomly. However, the probability of an area being chosen for settle— ment is greater the more settlers there are already settled in that area. A cell within the area is then chosen randomly. The simulation guards against the overloading of specific areas or the entire plane. After all the settlers have been located, a natural hazard event is generated. The hazard system was arbitrarily determined to be that of annual flooding and the hazard generator produces random flood variates based on the flood experience of Lafayette, Indiana on the Wabash River (see Dury, 1971). The flood discharge level is transformed into a perceived hazard event utilizing a normal curve equation. The equation weights medium frequent hazard events more, and weights the very frequent and quite infrequent events less. The simulation assumes that each settler is limited to two responses to the hazard system. A settler may absorb the hazard's damage or he/she may decide to move. This decision is based upon the settler's accumulated per- ception of the hazard system. When a settler's accumulated 115 perception reaches a threshold level, then the settler vacates his settlement location. The simulation functions on even time increments of one year. The user is asked to input the number of years the simulation is to run, the mover/stayer threshold, and the number of settlers that will locate on the plane each year. The user is also required to choose the initial settlement pattern to begin the simulation. The settle- ment pattern is displayed at the end of the simulation run. The user may choose to display the pattern at the end of each year, and has the option to see it more often. SETSIM, has two special facilities which may be initiated when the routine asks the user for his/her name at the beginning of each SETSIM session. By responding with the word FAST, SETSIM eliminates a great deal of output and allows the simulation to run faster. This facility is designed for the user who is interested in comparing final patterns and who is not interested in the evolution of the pattern from year to year. By responding with the word DEBUG, a switch system is initialized which causes a great deal of extra output to be generated at each step of the simulation. This is designed to aid in any revision or debugging procedures. A sample SETSIM session appears in another Appendix, as does a listing of the program. APPENDIX B A SAMPLE SETSIM SESSION APPENDIX B A SAMPLE SETSIM SESSION SETSIM WELCOME TO SETS IN TO PERSONALIZE THIS EXPERIENCE A BIT MORE PLEASE TYPE YOUR NAMF= ?MARY THANKS, MARY, DO YOU NEED AN INTRODUCTORY EXPLANATION ON HOW SETSIM OPERATES? (Y OR N)?Y SETSIM IS AN INTERACTIVE ROUTINE DESIGNED TO SIMULATE THE SETTLEMENT OF AN IDEALIZED SETTLEMENT PLANE. THE PRINCIPAL DYNAMIC FEATURE OF SETSIM IS ITS INCORPORATION OF A PROCESS WHICH IS DESIGNED TO GENERATE THE FREQUENCY AND INTENSITY OF A NATURAL HAZARD (e.g. A DROUGHT OR A FLOOD) AND SIMULATE THE RESPONSE OF THE SETTLERS TO THIS HAZARD. THE DEVELOPMENT OF THIS ROUTINE HAS BEEN PRIMARILY A CONCEPTUAL EXERCISE AND IS NOT BASED ON ANY EMPIRICAL STUDY. 116 117 (WHEN YOU ARE DONE READING A SET OF INSTRUCTIONS PRESS THE RETURN KEY TO CONTINUE.) THE CELL FRAME WORK LOOKS LIKE THIS= X X X X . . . . X X X X . . . . X X X X X X X X . . . . X X X X . . . . X X X X X X X X . . . . X X X X . . . . X X X X X X X X . . . . X X X X . . . . X X X X . . . . X X X X . . . . X X X X . . . . . . . . X X X X . . . . X X X X . . . . . . . . X X X X . . . . X X X X . . . . . . . . X X X X . . . . X X X X . . . . X X X X . . . . X X X X . . . . X X X X X X X X . . . . X X X X . . . . X X X X X X X X . . . . X X X X . . . . X X X X X X X X . . . . X X X X . . . . X X X X . . . . X X X X . . . . X X X X . . . . . . . . X X X X . . . . X X X X . . . . . . . . X X X X . . . . X X X X . . . . . . . . X X X X . . . . X X X X . . . . X X X X . . . . X X X X . . . . X X X X X X X X . . . . X X X X . . . . X X X X X X X X . . . . X X X X . . . . X X X X X X X X . . . . X X X X . . . . X X X X A SETTLEMENT AREA IS DEFINED AS A BLOCK OF 16 CELLS WHICH MAKE UP A FOUR BY FOUR SQUARE. THE PLANE IS MADE UP OF 25 SETTLEMENT AREAS, OR A 5 AREA BY 5 AREA MATRIX. THESE 'AREAS' ARE EMPHASIZED ABOVE. THE SIMULATION FUNCTIONS ROUGHLY LIKE THIS: 1) EACH PROSPECTIVE SETTLER IS CHOSEN A SETTLEMENT AREA. THE PROBABILITY OF AN AREA BEING CHOSEN IS INCREASED THE MORE THE AREA IS ALREADY SETTLED. A CELL WITHIN THE CHOSEN AREA IS PICKED RANDOMLY. EACH CELL REPRESENTS A POSSIBLE LOCATION FOR ONE (1) SINGLE-FAMILY FARM. 2) AFTER ALL THE SETTLERS HAVE FOUND A HOME, A NATURAL HAZARD IS GENERATED, THE LEVEL OF WHICH IS GOVERNED BY CHANCE. 3) EACH NATURAL HAZARD IS PERCEIVED BY EACH SETTLER WHO ADDS IT TO HIS/HER ACCUMULATING SENSE OF HAZARD AWARENESS. THE LEVEL OF HAZARD PERCEPTION FROM EACH GENERATED NATURAL HAZARD IS WEIGHTED IN A MANNER WHICH TENDS TO GIVE MODERATELY INTENSE EVENTS MORE INFLUENCE UPON THE PERCEPTION OF THE SETTLER, THAN THE VERY FREQUENT OR EXTREMELY INFREQUENT HAZARDOUS EVENTS. 4) AT THE END OF EVERY 'YEAR' EACH SETTLER IS EVALUATED TO SEE IF HIS/HER ACCUMULATED SENSE OF HAZARD HAS REACHED A THRESHOLD, WHICH YOU WILL SPECIFY. IF THE THRESHOLD IS REACHED THE SETTLER IS ASSUMED TO VACATE, AS THE AREA OF HIS SETTLEMENT IS TOO HAZARDOUS FOR THAT PARTICULAR SETTLER'S LIKING. 118 FOR AN INITIAL SETTLEMENT PATTERN YOU HAVE FOUR CHOICES= 1). THE CENTER OF EACH AREA IS SETTLED. 2). THE EAST OF THE STUDY PLANE IS SETTLED. 3). THE CENTER OF THE ENTIRE GRID IS SETTLFD. 4). THE PLANE IS LEFT UNSETTLED. WITH WHICH PATTERN WOULD YOU LIKE TO BEGIN THE SIMULATION? ?3 CHOSEN INITIAL SETTLEMENT PATTERN = OOOXXOOOOOOI oooXXoooo HOW MANY YEARS WOULD YOU LIKE THIS RUN TO PROGRESS?6 YOU HAVE THREE CHOICES FOR AN ACCUMULATED PERCEPTION THRESHOLD: l) 0 LOW, 2) e MEDIUM' 0R 3) 0 HIGH. 'THE LOW THRESHOLD REPRESENTS THE EQUIVALENT OF EXPERIENCING 'FHREE SUCCESIVE ANNUAL FLOODS WITH AN AVERAGE OCCURRENCE OF 5 YEARS, THAT IS THREE '5-YEAR FLOODS'. 'PHE MEDIUM THRESHOLD REPRESENTS THE EQUIVALENT OF EXPERIENCING ‘THREE SUCCESIVE YEARS OF A 'IO-YEAR FLOOD'. 'THE HIGH THRESHOLD REPRESENTS THE EQUIVALENT OF THREE SUCCESSIVE YEARS OF A '15-YEAR FLOOD'. 119 WHAT THRESHOLD DO YOU WISH TO USE DURING THIS RUN? (MARY--PLEASE TYPE 1,2,0R 3) ?1 HOW MANY SETTLERS WILL BE MOVING INTO THE SETTLEMENT REGION THIS YEAR?40 STANDBY-'8ETTLERS BEING LOCATED ALL SETTLERS HAVE BEEN LOCATED. TO SEE THE RESULTING PATTERN TYPE 'Y', TO CONTINUE TYPE 'N'. ?Y RESULTS OF SETTLING ALL SETTLERS = x O O x O O O O O O O O O O I O O O O O . X . . . . . . . . . . . . . . . . . . . . . X . X . . . . . . . . . . . . . . . . . . . . . . . X . . . . . . . . . . . . . . X . . . X . . . . . . . . . . . . . . . . . . X X X X . . . . . . . . . . . . . . . . X X X X . . . . . . . . . . . . . . . . X X X X . . . . . . . . . . . . . X . . X X X X X . . . . . . . . X . . . . . X . X X . . . . . . . . . . X X . . . . . X . X X . . . . . . . . . . . X . X X X . . X . . . . . . . . . . . . . . . X . . . . . . . . . . . . . . X . . . . . . . . . . . . . . . . . . . . . . . . . X . . . . . . . . . . . . . . . . . . . . . . X . . . . . . . . . TO CONTINUE HIT RETURN FLOOD THIS YEAR PEAKED AT 49497 CUBIC FEET PER SECOND EVENT HAS AN AVERAGE OCCURRENCE OF 1.90YEARS WITH A PROBABILITY OF OCCURRING IN ANY GIVEN YEAR OF 0.52 120 PERCEIVED HAZARD INDEX = 16.9 TO CONTINUE HIT RETURN YEAR 1 HAS CONCLUDED DO YOU WISH TO SEE THE SETTLEMENT PATTERN? ?N A NEW YEAR IS ABOUT TO BEGIN HOW MANY SETTLER'S DO YOU EXPECT TO ARRIVE THIS YEAR?30 STANDBY--SETTLERS BEING LOCATED ALL SETTLERS HAVE BEEN LOCATED. TO SEE THE RESULTING PATTERN TYPE 'Y', TO CONTINUE TYPE 'N'. ?Y RESULTS OF SETTLING ALL SETTLERS = x O O x O O I O O O O I O O O O O O O O . X . . X . . . . . . . . . . . . . . . . . X X . . X . . . . . . . . . . . . . X . . . X X X . . . . . . . . . . . . . 0 O O O O O x O O O O O O O O O O O O O O O O O x O O O x O O O O O O O O O O . . . . . . X . X X X X . X . . . . . . . . . . . X . . X X X X . . . . . . . . . . . . X . . . X X X X . . . . . . . . . . . . . X X X X X X X X . . . . . . . . X X . . X . X . X X . . . . . . . . . . X X . . . X . X X X X . . . . . . . . . . X X . X X X X X X X . . . . . . . . X . . . X . X . X . . . . . . . . . . . . X . . . . . . . . . . . . . . . . . . O x O x O O O O O O O O O O O O O O O O . . . . . . . X . . . . . . X . . . . X . . . . . . . . . . X . X X . . . . . . TO CONTINUE HIT RETURN FLOOD THIS YEAR PEAKED AT 39622 CUBIC FEET PER SECOND EVENT HAS AN AVERAGE OCCURRENCE OF 1.00YEARS WITH A PROBABILITY OF OCCURRING IN ANY GIVEN YEAR OF 1.00 121 PERCEIVED HAZARD INDEX 3 15.0 TO CONTINUE HIT RETURN YEAR 2 HAS CONCLUDED DO YOU WISH TO SEE THE SETTLEMENT PATTERN? ?N A NEW YEAR IS ABOUT TO BEGIN HOW MANY SETTLER'S DO YOU EXPECT TO ARRIVE THIS YEAR?25 STANDBY-“SETTLERS BEING LOCATED ALL SETTLERS HAVE BEEN LOCATED. TO SEE THE RESULTING PATTERN TYPE 'Y', TO CONTINUE TYPE 'N'. ?Y RESULTS OF SETTLING ALL SETTLERS = X . . X X . . X . . . . . . . . . . . . . X . . X X X X . . . . . . . . . . . . . X X X . X X X . . . . . . . . . . . . X . X . X X X . . . . . . . . . . . . . . . . . . . X . . . . . . . . . . . . . O O O O O O O O O O x O O O O O O O O O O O O O O O x O O O O O O O O C O O 0 O O O O O O X D O O x O O O O O O O 0 O O . . . . . . X . X X X X . X . . . . . . . . . . X X . X X X X X . . . . . . . . . . . . X . . . X X X X . . . . . X . . . . . . X X X X X X X X X . . . . . . . . X X . . X . X X X X . . . . . . . . . X X X X . X X . X X X X . . . . . . . . . . X X . X X X X X X X . . . . . . . . X . X . X X X . X X . . . . . . . . . . O x O O O O O O C C O O O x O O O x O O . X . X . . . . . . . . . X . . . . . . O O O O O O O x O O O O O O x O O O O x . . . . . . . X . . X . X X . . . . . . TO CONTINUE HIT RETURN FLOOD THIS YEAR PEAKED AT 40382 CUBIC FEET PER SECOND EVENT HAS AN AVERAGE OCCURRENCE OF 1.03YEARS WITH A PROBABILITY OF OCCURRING IN ANY GIVEN YEAR OF 0.97 TO SEE THE RESULTING 122 15.0 ING ALL SETTLERS ' HAS CONCLUDED HOW MANY SETTLER'S DO YOU EXPECT TO ARRIVE THIS YEAR?30 STANDBY--SETTLERS BEING LOCATED ALL SETTLERS HAVE BEEN LOCATED. PATTERN TYPE 'Y', TO CONTINUE TYPE 'N'. ?N SETTLER UNABLE TO FIND A SUITABLE LOCATION TO SETTLE IN AREA 17 ?Y THIS YEAR AND MOVED AWAY FROM-THE SETTLEMENT PLANE. DO YOU WISH TO SEE THE SETTLEMENT PATTERN? A NEW YEAR IS ABOUT TO BEGIN PERCEIVED HAZARD INDEX = TO CONTINUE HIT RETURN YEAR 3 . . . . . . . . . . .X . . . .X . . . . . . . . . . . .X . . . . .X . . . . . .X . . . . . . . . . . . . .. . . . . . . .X . . . . . . . . . . .XX .X . . . . . . .XX .X . . . .X . . . . . . .X .xxxx.xxx.... .XXXXXXX . . . ova“ xxxxxxxxx . o . ...U oxxxxxxxx o . . on . .XXXXXXX . .XXR ox oxxxxxx . . . .m X .xxxxxxx . . .XXXXXXX... . . . ..XXXX .X. . . . ..XXXX ... . . . ..XXXXXX. . . . ..XXXX... TO CONTINUE H 123 FLOOD THIS YEAR PEAKED AT 16635 CUBIC FEET PER SECOND EVENT HAS AN AVERAGE OCCURRENCE OF 1.49YEARS WITH A PROBABILITY OF OCCURRING IN ANY GIVEN YEAR OF 0.67 PERCEIVED HAZARD INDEX = 16.0 TO CONTINUE HIT RETURN YEAR 4 HAS CONCLUDED DO YOU WISH TO SEE THE SETTLEMENT PATTERN? N A NEW YEAR IS ABOUT TO BEGIN HOW MANY SETTLER'S DO YOU EXPECT TO ARRIVE THIS YEAR?10 STANDBY--SETTLERS BEING LOCATED ALL SETTLERS HAVE BEEN LOCATED. TO SEE THE RESULTING PATTERN TYPE 'Y', TO CONTINUE TYPE 'N'. '?Y RESULTS OF SETTLING ALL SETTLERS 3 X X X X X X X X . . . . . . . . . . . . X X . X X X X X . . . . . . . . . . . . X X X X X X X X . . . . . . . . . . . . X . X . X X X X . . . . . . . . . . . . O O O O O O x O O O O O O O O O O O O O O O O O O O O O O O x O O O O O O O O O O O O O O O x x O O O O O O O O O O O O O O O O O x O O O x O O O O O O O O O O . . . . . . X . X X X X . X . X . X X . O O O O x x x x x x x x O O O O O O O O . . . . X X X X X X X X . . . . X X X . . . . . X X X X X X X X X . . . . . . . X X X X X X X X X X X X . . . . . . . . X X X X X X X X X X X X . . . . . . . . X X X X X X X X X X X X . . . . . . . . X X X X X X X X X X X X . . . . . . . . O x O O O O O x O O O O O x O x O x O O . X . X . . . . . . . . . X . . . . . . . . . . . . . X . . . . . . X . . . . X . . . . . X . X . . X . X X X X . . . . TO CONTINUE HIT RETURN 124 FLOOD THIS YEAR PEAKED AT 44223 CUBIC FEET PER SECOND EVENT HAS AN AVERAGE OCCURRENCE OF 1.33YEARS WITH A PROBABILITY OF OCCURRING IN ANY GIVEN YEAR OF 0.75 PERCEIVED HAZARD INDEX = 15.7 TO CONTINUE HIT RETURN YEAR 5 HAS CONCLUDED DO YOU WISH TO SEE THE SETTLEMENT ?N PATTERN? A NEW YEAR IS ABOUT TO BEGIN HOW NANY SETTLER'S DO YOU EXPECT TO ARRIVE THIS YFAR?5 STANDBY-'8ETTLERS BEING LOCATED ALL SETTLERS HAVE BEEN LOCATED. TO SEE THE RESULTING PATTERN TYPE 'Y', TO CONTINUE TYPE 'N'. ?N FLOOD THIS YEAR PEAKED AT 57055 CUBIC FEET PER SECOND EVENT HAS AN AVERAGE OCCURRENCE OF 3.06YEARS WITH A PROBABILITY OF OCCURRING IN ANY GIVEN YEAR OF 0.32 PERCEIVED HAZARD INDEX = 19.5 TO CONTINUE HIT RETURN SETTLER 'HAZARDED OUT' FROM AREA 1 CELL 1 SETTLER 'HAZARDED OUT' FROM AREA 1 CELL 4 SETTLER 'HAZARDED OUT' FROM AREA 1 CELL 6 SETTLER 'HAZARDED OUT' FROM AREA 1 CELL 11 SETTLER 'HAZARDED OUT' FROM AREA 1 CELL 12 SETTLER 'HAZARDED OUT' FROM AREA 1 CELL 13 SETTLER 'HAZARDED OUT' FROM AREA 2 CELL 5 SETTLER 'HAZARDED OUT' FROM AREA 2 CELL 11 SETTLER 'HAZARDED OUT' FROM AREA 2 CELL 13 SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER SETTLER 'HAZARDED 'HAZARDED 'HAZARDFD 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDLD 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED |HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDFD 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED 'HAZARDED OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' OUT' FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM FROM 125 AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA AREA ARFA AREA AREA AREA AREA AREA AREA AREA AREA AREA CELL CELL CELL CELL CELL CELL CFLL CEIJ. CELL CEIJ. CELL CELL CEIIIJ CELL CELL CELL CELL CELL (fiilal. (3E?I.I. CELL CELL (IEQIQI. CELL CFIJ. CELL CELL (IEII.I. CELL CELL CELL CELL CELL CEIJ. (ZIEI.I. CEIJIJ (IEEIJI. CELL CELL CEIIIJ CEIJIJ CELL CELL CELL CEIJ. CELL CELL CELL CELL CELL CEIJ. (TIEI.I. CEI‘IJ H‘DGQO‘U‘IWN 126 SETTLER 'HAZARDED OUT' FROM AREA 18 CELL 11 SETTLER 'HAZARDED OUT' FROM AREA 18 CELL 12 SETTLER 'HAZARDED OUT' FROM AREA 18 CELL 13 SETTLER 'HAZARDED OUT' FROM AREA 21 CELL 2 SETTLER 'HAZARDED OUT' FROM AREA 21 CELL 6 SETTLER 'HAZARDED OUT' FROM AREA 21 CELL 8 SETTLER 'HAZARDED OUT' FROM AREA 22 CELL 12 SETTLER 'HAZARDED OUT' FROM AREA 23 CELL 15 SETTLER 'HAZARDED OUT' FROM AREA 24 CELL 11 SETTLER 'HAZARDED OUT' FROM AREA 24 CELL 13 SETTLER 'HAZARDED OUT' FROM AREA 24 CELL 14 SETTLER 'HAZARDED OUT' FROM AREA 25 CELL 12 WELL, MARY, THAT'S IT FOR THIS RUN AFTER 6 YEARS THE FINAL SETTLEMENT PATTERN LOOKS LIKE THIS= . X X . X X X X . . . . . . . . . . . . X . X X . X X X . . . . . . . . . . . . X X . . X X . X . . . . . . . . . . . . . . X . . . . X . . . . . . . . . . . . . . . X X . . X . . . . . . . . . . . . . . . . . . . . . . . X X . X X . . . . . X . X X . . . . . . . . . . . X . . . . . X X X . . . . . . . . X X X . . . . . X . . . . . . . . . . . . . . . X . . X X . X . X . . X . . . . . . . . X . . X X X . X . . . . . . . . . . . . X X . . X . . . . . . . . . . . . . . . X X X . X . X . X X X . . . . . . . . . . . . . X . . . . . X . X . X . . . . . . . . . . . . . . X . . . . . . . . . . . . . . . . . . . X . . . . . . . . . . . X . X . . . . . X X . . . . WOULD YOU LIKE TO RUN THE SIMULATION AGAIN?(TYPE Y OR N)?N TYPE BYE TO LOGOUT DONE APPENDIX C A LISTING OF THE PROGRAM SETSIM File: 100 110 120 130 140 150 160 170 180 190 200 210 220 230 240 250 260 270 280 290 300 310 320 330 340 350 360 370 380 390 400 410 420 430 440 450 460 470 480 490 500 510 520 530 540 APPENDIX C A LISTING OF THE PROGRAM SETSIM OUT Program Name: SETSIM Aug-O4-77 Thursday Page: 1 REM‘tiiitti*ttifiittttititiiiIfifitfitiiSETSInitifil‘iittii*il’iitiitiifikiiiiiiitiii RBM REM A SIMULATION OR THE RESPONSE OP RURAL SETTLEMENT REM TO A LOCAL HAZARD SYSTEM REM REM REM MAIN PROGRAM REM Y1= NUMBER OP YEARS SIMULATION Is To RUN REM M= MOVBR/STAYRR THRESHOLD REM N= NUMBER OP SETTLERS TO BE LOCATED IN REM A GIVEN YEAR REM Te CLOCK, YEARS SIMULATION HAS RUN REM Tl' ‘OTAL SETTLERS ON PLANE Ran A(25,16)= MASTER ARRAY- 25 ARRAs WITH 16 CELLS REM REM COM leBO],NSlBO],DSIBO],B$I80],M$[80] DIM Al25,18] T=0 PRINT “WELCOME TO SETSIM” PRINT LIN(3) PRINT 'TO PBRSONALIZE THIS EXPERIENCE A BIT MORE PLEASE TYPE YOUR NAME=' INPUT zs REM SWITCH INITIALIZATION RBM J1 . DEBUG SWITCH REM J2 - FAST SWITCH IF zs="PAST' THEN 393 J2=o GOTO 420 J2=1 J1=1 GOTO 460 IF Z$='DEBUG" THEN 450 J1=1 GOTO 460 J1=o 1? J2 THEN 610 PRINT LIN(2) PRINT "THANKS. ';zs;', DO YOU NEED AN INTRODUCTORY EXPLANATION ON HOW ' PRINT “SETSIM OPERATES? (Y OR N)”; INPUT Ns IP Ns=“y" THEN 570 IP N$='n' THEN 610 IP N$-'Y' TMBN 570 IF Ns-'N' THEN 610 1J2? 128 File: OUT Program Name: SETSIM Aug-04-77 Thursday Page: 2 550 PRINT 'PLEASE TYPE Y OR N' 560 GOTO 500 S70 REH(((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((()(- 580 REM CALL EXPLAN SUB 590 60808 4440 600 REM))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))) 610 T1=0 620 NAT A-ZER 530 RBH(((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((( 640 REM CALL INITAL SUB 650 60808 2800 660 REM))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))) 670 PRINT LIN(1) 680 REM USER DEPINES EXOGENOUS VARIABLES 690 REM 700 PRINT “HOW MANY YEARS NOULD YOU LIKE THIS RUN TO PROGRESS': 710 INPUT Y1 720 1? Y1 >- 1 AND Y1 <- 30 THEN 750 730 PRINT 'PLEASE CHOOSE A NUMBER BETWEEN 1 AND 30' 740 GOTO 710 750 PRINT LIN(2) 760 PRINT 'YOU HAVE THREE CHOICES FOR AN ACCUNULATED PERCEPTION THRESHOLD:' 770 PRINT '1). LON, 2). MEDIUM, OR 3). HIGH.‘ 780 PRINT LIN(1) 790 PRINT 'THE LON THRESHOLD REPRESENTS THE EQUIVALENT OP EXPERIENCING“ 800 PRINT 'THREE SUCCESIVE ANNUAL PLOODS NITH AN AVERAGE OCCURRENCE“ 810 PRINT '0? 5 YEARS, THAT IS THREE 'S-YEAR PLOODS'.‘ 820 PRINT LIN(1) 830 PRINT “THE MEDIUM THRESHOLD REPRESENTS THE EQUIVALENT OP EXPERIENCING“ 840 PRINT 'THREE SUCCESIVE YEARS OF A 'IO-YEAR PInOOD'.‘I 850 PRINT LIN(1) 860 PRINT 'THE HIGH THRESHOLD REPRESENTS THE EQUIVALENT OP THREE SUCCESSIVE' 870 PRINT "YEARS OF A '15-YEAR PLOOD'.‘ 880 PRINT LIN(1) 890 PRINT 'NHAT THRESHOLD DO YOU WISH TO USE DURING THIS RUN?“ 900 PRINT '(' 32$: '°-PLEASE TYPE 1, 2 ,OR 3)‘ 910 INPUT M 920 IE N-l THEN 970 930 IE M-2 THEN 990 940 1? M-3 THEN 1010 950 PRINT "TYPE 1, 2, OR 3, PLEASE' 960 GOTO 910 970 M-65.S7 980 GOTO 1030 990 M-95.41 File: 1000 1010 1020 1030 1040 1050 1060 1070 1080 1090 1100 1110 1120 1130 1140 1150 1160 1170 1180 1190 1200 1210 1220 1230 1240 1250 1260 1270 1280 1290 1300 1310 1320 1330 1340 1350 1360 1370 1380 1390 1400 1410 1420 1430 1440 129 OUT Program Name: SETSIM Aug-04-77 Thursday Page: 3 GOTO 1030 M=108.11 GOTO 1030 PRINT LIN(2) IE J1 THEN 1060 PRINT TAB(40):'CHOSEN THRESHOLD=':M REM PRINT LIN(2) , PRINT ”HOW MANY SETTLERS WILL BE MOVING INTO THE SETTLEMENT REGION THIS“ PRINT "YEAR“: INPUT N I? N >= 0 AND N <- 100 THEN 1140 PRINT ”PLEASE CHOOSE A NUMBER BETWEEN 0 AND 100' GOTO 1100 IE Nco THEN 1300 Tz-T1+N IE T2 <= 400 THEN 1250 PRINT LIN(1) PRINT 25;", THERE ARE ALREADY '3T1:' LOCATED SETTLERS.“ PRINT 'AN ADDITIONAL ":Ns' SETTLERS WOULD OVERLOAD THE SETTLEMENT PLANE.‘ PRINT ' THERE ARE ONLY 400 POSSIBLE SETTLEMENT LOCATIONS.” T2-400-T1 PRINT LIN(1) PRINT "PLEASE CHOOSE A NEW NUMBER BETWEEN 0 AND ';T2 GOTO 1100 REM REM((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((( REM CALL SETTLE SUB GOSUB 1880 REM)))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))) REM REM(((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((( REM CALL EVENT SUB 60808 3660 :EN))))))))))))))))))))))))))))))))))))))))))))))))1))))))))))))))))))))) REM((((((((((((((((((((((((((((l((((((((((((((((((((((((((((((((((((((((( REN LADJUST SU - 60803 3970 35:11))))))))))))))))))))))))))))))11)))))))))))))))))))))))))))))))))) IP T=Y1 THEN 1630 IF J2 THEN 1570 PRINT LIN(1) PRINT "YEAR '3T3' HAS CONCLUDED'I File: 1450 1460 1470 1480 1490 1500 1510 1520 1530 1540 1550 1560 1570 1580 1590 1600 1610 1620 1630 1640 1650 1660 1670 1680 1690 1700 1710 1720 1730 1740 1750 1760 1770 1780 1790 1800 1810 1820 1830 1840 1850 1860 1870 1880 1890 130 OUT Program Name: SETSIM Aug-04-77 Thursday Page: 4 PRINT "DO YOU WISH TO SEE THE SETTLEMENT PATTERN?" INPUT BS IF B$="Y" THEN 1530 IR Bs-'y' THEN 1530 IF 35¢"N" THEN 1570 IE Bs-'n- THEN 1570 PRINT "PLEASE TYPE Y OR N" GOTO 1460 PRINT "SETTLEMENT PATTERN AT THE END 0? YEAR NUMBER":T:"=" REN((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((( Rm CALL PRINT SUB GOSUB 3310 REM))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))) PRINT LIN(2) PRINT "A NEW YEAR IS ABOUT To BEGIN" PRINT "HOW MANY SETTLER'S DO YOU EXPECT To ARRIVE THIS YEAR": GOTO 1100 REM PRINT "WELL, “:252'. THAT'S IT EOR THIS RUN" PRINT LIN(2) PRINT "AFTER": Y1:"YEARS THE EINAL SETTLEMENT PATTERN LOOKS LIKE THIs-' REM((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((( RE CALL PRINT SUB GOSUB 3310 REM)))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))) PRINT LIN(2) PRINT "WOULD YOU LIKE TO RUN THE SIMULATION AGAIN?(TYPE Y OR N)": INPUT Bs IE B$-'y' THEN 1790 IE Bs-'Y' THEN 1790 IE B$-'N' THEN 1810 IP 88¢"n" THEN 1810 PRINT "PLEASE TYPE Y OR N" GOTO 1720 REM GOTO 470 REM PROGRAMMING HISTORY: DEVELOPED BY MARK NEITHERCUT REM DEPT OP GEOGRAPHY REM MICHIGAN STATE UNIVERSITY REM JUNE, 1977 PRINT LIN(2) PRINT "TYPE BYE TO LOGOUT" STOP Rfiuttttttittiittitttitttttttttttiitthtiitittttttttflitttittiititttttt Rguittiifl...iiitiiiiItCittfittiitt*fifiittttiItiifltitfitiiiiiiii*iiiiitfi File: 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100 2110 2120 2130 2140 2150 2160 2170 2180 2190 2200 2210 2220 2230 2240 2250 2260 2270 2280 2290 2300 2310 2320 2330 2340 131 OUT Program Name: SETSIM Aug-04-77 Thursday Page: 5 REM SETTLE SUBROUTINE Rgnittt*tittit...ttttitittiititttttttttttittikthitttfififittitttttiittfl. REM Sl=TOTALS NUMBER OF SETTLERS ON PLAN14 REM 25 E 26' ARGUMENTS POR PROBABILITY DENSITY REM FUNCTIONS REM A(I.18)= MIF UPPER BOUNDARY FOR AREA I PRINT "STANDBY-~SETTLERS BEING LOCATED" REM 7 REM CHOOSES WHICH SETTLEMENT AREA TO BE SETTED DEF PNA(ZS)=1+(1000-1)*RND(1) DEP PNB(Z6)=1+(16-1)*RND(1) 51-0 EOR I-l TO 25 Sl-(A[1.17]*2)+Sl+1 IE J1 THEN 2060 PRINT TAB(40);A[I,17];" SETTLERS IN AREA I “:1 NEXT I REM REM ALLOCATE MIE PROBABILITIES ON BASIS OE I SETTLED REM IN EACH AREA REM Taco EOR I-1 TO 25 A[1,18]-(((A[1,17]*2)+1)/Sl)*1000+T8 TB-A[1,18] I? 01 THEN 2170 PRINT TAB(35);'TOP MIE LIMIT EOR AREA 9 ":I:"IS":A[I,18] NEXT I REM REM CHOOSE SETTLEMENT AREA REM Sz-ENA(ZS) IE J1 THEN 2240 PRINT TAB(40);'AREA VARIATE - ';52 EOR I-l TO 25 IE sz>A[I,18] THEN 2280 IE A[I,17] >= 16 THEN 2210 GOTO 2310 NEXT I PRINT "ERROR IN SETTLE SUB. EAULTY MIE ALLOCATION“ STOP IE J1 THEN 2330 PRINT TAB(40);"AREA CHOSEN e '31 $81 REM 132 File: OUT Program Name: SETSIM Aug-04-77 Thursday Page: 6 2350 REM 2360 REM NON CHOOSE CELL RANDOMLY WITHIN PICKED AREA 2370 N580 2380 C'FNB(Z6) 2390 IF J1 THEN 2410 2400 PRINT TAB(40):'CELL VARIATE IS ';C 2410 IF A[S,C]=0 THEN 2490 2420 IP N5=15 THEN 2450 2430 N5=N5+1 2440 GOTO 2380 2450 PRINT LIN(2) 2460 PRINT "SETTLER UNABLE TO FIND A SUITABLE LOCATION TO SETTLE IN AREA":S 2470 PRINT "THIS YEAR AND MOVED AWAY FROM THE SETTLEMENT PLANE." 2480 GOTO 2540 2490 AIS,C]=1 2500 AIS,17]'A[S,17]+1 2510 T1°T1+1 2520 IF J1 THEN 2540 2530 PRINT TAB(40);"CELL 8 ":C;"AREA i "35;"SETTLED " 2540 REM 2550 REM CONTINUE TO SETTLE THE REST OF THE SETTLERS IN THE QUEUE 2560 N=N-1 2570 IF N>0 THEN 2010 2580 1? J2 THEN 2760 2590 PRINT LIN(2) 2600 PRINT “ ALL SETTLERS HAVE BEEN LOCATED. To SEE THE RESULTING“ 2610 PRINT “PATTERN TYPE 'Y', To CONTINUE TYPE 'N'.“ 2620 INPUT DS 2630 IE D$=“y“ THEN 2690 2640 I? D$-“Y“ THEN 2690 2650 IE Ds-“n“ THEN 2760 2660 IF D$=“N“ THEN 2760 2670 PRINT “TYPE 'Y' OR 'N' PLEASE “ 2680 GOTO 2620 2690 PRINT “RESULTS OE SETTLING ALL SETTLERS -“ 2700 Rgflccecc====eeccecaecu-ccccuuuuacnluucsaccent-cucuzluncunce==¢na==ac 2710 REM CALL PRINT SUBROUTINE 2720 GOSUB 3310 2730 Ran-n:c=====¢==eec==nccauccccccc=ecInca:sunsets-caaucccnncceac: 2740 PRINT "TO CONTINUE HIT RETURN " 2750 LINPUT MS 2760 REM 2770 REM 2780 RETURN 2790 STOP File: 2800 2810 2820 2830 2840 2850 2860 2870 2880 2890 2900 2910 2920 2930 2940 2950 2960 2970 2980 2990 3000 3010 3020 3030 3040 3050 3060 3070 3080 3090 3100 3110 3120 3130 3140 3150 3160 3170 3180 3190 3200 3210 3220 3230 3240 133 OUT Program Name: SETSIM Aug-04-77 Thursday Page: 7 Raniittiiitt*tittfiiititittiiiitiiittiititttiflitttitt*ittitttittfiiiiti Rsnflfliititttiitiflttititttflflfltfliiititit*tiitttfittiiifiifliiifififlflfitflifitfifi REM INITIALIZATION SUBROUTINE RB"...tittifittttitfittititttititttttttit*tittttttt*tttittttitittttitttt REM PRINT LIN(2) PRINT "FOR AN INITIAL SETTLEMENT PATTERN YOU HAVE FOUR CHOICES=" PRINT 1). THE CENTER 0? EACH AREA IS SETTLED. " PRINT " 2). THE EAST OF THE STUDY PLANE IS SETTLED." PRINT " 3). THE CENTER OF THE ENTIRE GRID IS SETTLED." PRINT " 4). THE PLANE IS LEFT UNSETTLED." PRINT "WITH WHICH PATTERN WOULD YOU LIKE TO BEGIN THE SIMULATION?" INPUT P I? P'l THEN 3010 I? P°2 THEN 3080 I? P33 THEN 3150 I? P" THEN 3250 PRINT IILLEGAL PATTERN CHOICE, PLEASE RETYPE' GOTO 2920 RE" RE" 1's ARE USED TO DESIGNATE CELLS THAT ARE SETTLED FOR I'l T0 25 AII,7]¢1 Allplol‘l AIIr171=2 NEXT I T1'50 GOTO 3220 FOR I35 TO 25 STEP 5 Ally‘lgl AII'12]=1 AIIo17l=2 NEXT I T1-10 GOTO 3220 A[13,6]=1 A[13,7]=1 A[13,10]=1 A[13,11]=1 All3p17l=4 T1'4 GOTO 3220 PRINT LIN(2) I? J2 THEN 3270 PRINT .CHOSEN INITIAL SETTLEMENT PATTERN a“ 134 File: OUT Program Name: SETSIM Aug-O4-77 Thursday Page: 8 3250 GOSUB 3310 3260 REM 3270 RETURN 3280 STOP 3290 REM 3300 REM 3310 REH***it*ifiiiiiii*****.*****i****itttiitttitiiit****fiflfifli*******t*** 3320 REHitiitiiiifititit!t**.i****iti*i********itiifii.************fittitflfii 3330 REM print subroutine 3340 REHttttitititiiiitit*tiittittflt*ititititiifliitifitttfititttittiitttiii 3328 REM PRINT SUBROUTINE T0 OUTPUT CURRENT PATTERN REM 3370 IF J1 THEN 3390 3380 PRINT TAB(40);"ENTER PRINT SUB" 3390 Y=0 3400 FOR Lcl TO 5 3410 X=0 3420 IF J1 THEN 3440 3430 PRINT TAB(40);"NEW BLOCK" 3440 FOR I=1 T0 4 3450 IF Jl THEN 3470 3460 PRINT TAB(40);"NEN LINE" 3470 FOR J=1 TO 5 3480 S=J+Y 3490 FOR K=1 TO 4 3500 C=K+X 3510 IE AIS,C]>0 THEN 3550 3520 PRINT USING 3530 3530 IMAGE 4," .' 3540 GOTo 3570 3550 PRINT USING 3560 3560 IMAGE t," x“ 3570 NEXT K 3580 NEXT J 3590 x=x+4 3600 PRINT 3610 NEXT I 3620 Y=Y+S 3630 NEXT L 3640 RETURN 3650 STOP 3660 REuti*flttii*fl.****i*ittii****fiitiiiifiit*flttiiitifiittttifiitittiflittifl 3670 Rani.*tfifitittttiflitititttfliiiiiititiifiifliitifit.*****i****i*i*iii*** 3680 REM EVENT GENERATOR 3590 Ranittittit*ttttttit*ttttttttititittittttitittttttttitttttiittttttfii File: 3700 3710 3720 3730 3740 3750 3760 3770 3780 3790 3800 3810 3820 3830 3840 3850 3860 3870 3880 3890 3900 3910 3920 3930 3940 3950 3960 3970 3980 3990 4000 4010 4020 4030 4040 4050 4060 4070 4080 4090 4100 4110 4120 4130 4140 135 OUT Program Name: SETSIM Aug-04-77 Thursday Page: 9 REM REM GENERATES LOG-NORMAL VARIATES E2310.7594 S9=0 SZ'.4453 FOR I'l TO 12 R'RND(1) S9389+R NEXT I REM XBEXP(E2+(SZ*(S9-6))) X'INT(X) R1'4.16005+(-2.1784E-04*X)+(3.4805E-09*(X ** 2)) IF R1 >¢ 1 THEN 3850 R1=1 PRINT LIN(2) PRINT USING 38703X IMAGE "FLOOD THIS YEAR PEAKED AT ",DDDDDDD," CUBIC FEET PER SECOND" R1=INT(R1*100)/100 PRINT USING 39003R1 IMAGE "EVENT HAS AN AVERAGE OCCURRENCE OF ",DDD.DD,"YEARS" Y9=1/Rl Y9=INT(Y9*100)/100 PRINT USING 39403Y9 IMAGE "WITH A PROBABILITY OP OCCURRING IN ANY GIVEN YEAR OF ",D.DD RETURN STOP REHCOQOitiiititifliiiflittfltiiiflfiititiitii*tiiiiflifiiiiflflflititiitiiifiiifi Ranitiiititttittittit*ttitttttitttitttttttttttttttittttttttiittittttt REM ADJUSTMENT PROCESS SUBROUTINE Reutitititttttittiitittihttititttttttfittitttttttifittitttatttttittttttti REM REM TRANSLATES FLOOD EVENT INTO A PERCEIVED EVENT UTILISING REM A NORMAL EQUATION x1=15 8'10 P183.14159 E32.71828 A=(R1-X1) ** 2 B=2*(S ** 2) C=A/B Del/(E ** C) GBZ‘PI IF G >= 0 THEN 4160 PRINT "NEGATIVE ARGUMENT IN SQUARE ROOT FUNCTION IN ADJUSTMENT PROCESS" File: 4150 4160 4170 4180 4190 4200 4210 4220 4230 4240 4250 4260 4270 4280 4290 4300 4310 4320 4330 4340 4350 4360 4370 4380 4390 4400 4410 4420 4430 4440 4450 4460 4470 4480 4490 4500 4510 4520 4530 4540 4550 4560 4570 4580 4590 1J36 OUT Program Name: SETSIM Aug-04-77 Thursday Page: 10 STOP F=1/(S*SQR(G)) H=D*F H‘H‘1000 H81NT(H*100)/100 PRINT LIN(l) PRINT LIN(l) PRINT USING 42303H IMAGE "PERCEIVED HAZARD INDEX = ",DDD.D PRINT "TO CONTINUE HIT RETURN" LINPUT MS REM UPDATES ACCUMULATED PERCEPTION 0F HAZARD REM SYSTEM AND CHECKS TO SEE IF MOVER/STAYER REM THRESHOLD HAS BEEN REACHED EOR I=1 T0 25 EOR J=1 TO 16 IE A[I,J]=0 THEN 4390 AII.J1=((AII.Jl-1)“.9)+1 AII.J1=AII.J1+E IE (A[1,Jl-l)