AN ABSTRACT OF THE THESIS OF Richard Roger Hagestedt for the degree of Master of Science in Forest Management presented on Title: Spatial Allocation of Land Uses in Land Use Planning Abstract approved: Signature redacted for privacy. Pame1J. Lase The purpose of land use planning, as conducted by the Forest Service, is to allocate land uses. The techniques employed in the current planning process fail to take location of the land allocations into account in any systematic manner. The resulting solutions may be inconsistent with planning goals; the land use patterns produced may not provide the maximum value of goods and services possible while protecting long-term biological prodicti.vity of the Forest. This study examines the impact of location on land allo- cation decisions, developing a strategy and set of techniques for in- corporating spatial factors into the allocation process. Three spatial factors affect land allocion decisions: 1) the size of a land unit required to make management of a use practical, conflicts caused by the adjacent location of specific uses, and the need to organize ises across the landscape to take advantage
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
AN ABSTRACT OF THE THESIS OF
Richard Roger Hagestedt for the degree of Master of Science
in Forest Management presented on
Title: Spatial Allocation of Land Uses in Land Use Planning
Abstract approved: Signature redacted for privacy.Pame1J. Lase
The purpose of land use planning, as conducted by the
Forest Service, is to allocate land uses. The techniques
employed in the current planning process fail to take location of
the land allocations into account in any systematic manner. The
resulting solutions may be inconsistent with planning goals; the
land use patterns produced may not provide the maximum value of
goods and services possible while protecting long-term biological
prodicti.vity of the Forest.
This study examines the impact of location on land allo-
cation decisions, developing a strategy and set of techniques for in-
corporating spatial factors into the allocation process. Three
spatial factors affect land allocion decisions: 1) the size of a
land unit required to make management of a use practical,
conflicts caused by the adjacent location of specific uses, and
the need to organize ises across the landscape to take advantage
2
of certain characteristics of the planning unit.
Three promising strategies are investigated: 1) an opti-
mizing algorithm, 2) an efficient solution algorithm, and 3) an
assignment algorithm. The optimizing algorithm replaces the
linear program currently employed in the planning process with an
integer program able to consider location of land units in the allo-
cation process. The efficient solution algorithm uses an integer
program to create a land use pattern from the linear program
acreage allocations. Computer core size limitations and the size
and complexity of the planning problem prevent application of these
strategies. The assignment approach overcomes these difficulties
with a heuristic algorithm designed to locate linear program
allocations on the platining unit.
The computer programs required to support the spatial
allocation strategy include: 1) a computer mapping program,
2) a detail reductioti program, 3) an adjacency program,
4) the heuristic program, and 5) a conflict detection program.
The mapping program creates land units and keeps track of
their location. The detail reduction program eliminates some
complexity from the land base data. The adjacency program
identifies adjacent land units. The conflict detection program
detects conflicts caused by uses located adjacent to each other
and violations of minimum land unit size.
The spatial allocation strategy and its associated tools
are tested on the Clackamas Planning Unit of the Mt. Hood
National Forest. Results of this case study indicate that the
approach is workable with minor modifications.
Spatial Allocation of Land Usesin Land Use Planning
by
Richard Roger Hagestedt
A THESIS
submitted to
Oregon State University
in partial fulfillment ofthe requirements for
the degree of
Master of Science
Commencement June 1980
APPROVED:
Signature redacted for privacy.
Professor of Fopst Management in Charge of Major
J
Signature redacted for privacy.
H?'adof Department of Forest Management
Dean of Graduate School
Date thesis is presented
Typed by S. Hagestedt for Richard Roger Hagestedt
ACKNOWLEDGMENT
I would like to express my appreciation and gratitude to
the following people. It was through their assistance, support,
and understanding that this thesis was possible. My thanks to:
Pan-i Case,
Bill Ferrell,
Dave Butler
Jeff Arthur,
Doug Brodie,
John White and the Mt. Hood National Forest Planning
Staff,
my fellow graduate students,
my parents, sisters, brother, and my dear wile, Sarah.
TABLE OF CONTENTS
I. Introduction 1
The Spatial Problem 2
The Present Study 4
U. Location and Land Allocation 6
Linear Program Models of the PlanningProblem 7
Current Forms of Spatial Allocation 13
Types of Spatial Factors 15
Promising Algorithms 20Limiting Factors in the Search for an
The Spatial Allocation Strategy 35Tools Needed for the Procedure 35
The Computer Mapping Routine 35The Detail Reduction Program 38The Adjacency Program 41The Spatial Assignment Process 41The Conflict Detection Program 41
The Procedure 44Inputs to the Spatial Allocation Procedure 44Development of Management Units . . . 46Development of Spillover Matrix 48Activity Assignments 49Detrimental Spillove r s and Minimum
Management Unit Size 52Readjustment Procedure 54
V. Spatial Allocation of the Clackamas PlanningUnit 56
The Clackamas Planning Unit 57The Spatial Allocation Procedure 58
Inputs to the Spatial Allocation Pro-cedure 60
Construction of Management Units . . . 64Construction of the Spillover Matrix . . 66Assignment of the Activities 69Detrimental Spillovers and Minimum
Management Unit Size 70Discussion of Results 75
VI. Future Research Steps 85Difficulties in Application 86Tool Improvements and Subjects for Further
Investigation 88
Bibliography 94
AppendicesAppendix A 97Appendix B 101Appendix C 137Appendix D 142Appendix E 146Appendix F 149
LIST OF ILLUSTRATIONS
TABLE
I The Linear Program Model of theAllocation Problem
II Formulation for the Linear ProgramAlgorithm
PAGE
10
11
III Linear Program Solution to the Allocation 11
Problem
IV Mapping Routine Functions 39
V The Spatial Allocation Procedure 45
VI Activity Information for Clackamas 62Planning Unit
VII Acreage Allocations 65
VIII Allocation Achievement Table 71
IX Allocation Achievement Table 72
X Allocation Achievement Table 73
XI Acreage Breakdown Between Activities 83
XII Expected Values for Map 1 84
XIII Expected Values for Map 3 84
LIST OF ILLUSTRATIONS
Figure Page
A Typical Land Type Map 12
Spillover Matrix 49
A Digitized Map 61
The Original Expanded Legend 67
Spillover Matrix for Case Study 68
Detrimental Spillovers 74
Detrimental Spillovers 76
Violations of Minimum Management 77Unit Size
SPATIAL ALLOCATION OF LAND USESIN LAND USE PLANNING
I. INTRODUCTION
The U. S. Forest Service currently is engaged in compre-
hensive planning for use and management of National Forest lands.
The plans produced for each Forest within the National Forest
System are changing the pattern of land use throughout the System.
Changes in land use will change the mixture of goods and services
flowing to the public from these lands for the foreseeable future.
The significance of the changes in goods and services which
might be made as a result of the planning process has led the Forest
Service to design its planning system as carefully as possible. The
system emerging from early trials is an assembly of highly technical
components, each for analysis of some aspect of land capability:
ecological effects of alternative management regimes, social con-
sequences of shifts in land use patterns, or the achievement of
budgetary and operational criteria under alternative plans.
The basic task of the planning process is to allocate land
to uses. The plan produced must achieve two important objectives:
it must supply a mixture of goods and services to the public and it
must maintain the long-term biological productivity of forest Land.
The mixture of public goods and services is determined by maxi-
mizing the total value of the consumer products and recreational
opportunities which the Forest could provide. Protection of long-
term productivity is achieved by: a) identifying the kind of environ-
mental impacts produced if a land use were to be imposed on a given
area of land, b) identifying acceptable levels of these impacts and
then, c) constraining the allocation process so that excessive
impacts are avoided.
The Spatial Problem
The location of land uses on the planning unit can cause
undesirable effects. For example, land uses, such as timber pro-
duction or developed recreation, which are located on land types
consistent with their physical and vegetative requirements, may
sometimes have adverse effects on some other resource, such as
when they intrude upon the movement of wildlife from summer to
winter range. Location also affects the supply of forest products
realized by the land use allocations. Timber production, for
example, could be assigned to one area of the planning unit on the
basis of compatibility with land type, but its location within the
planning unit, in terms of road access or size of the unit, could
make the assignment impractical or uneconomic. In a related
example, land uses, such as timber production and developed
recreation, which are consistent with the ecological requirements
of the land types to which they are assigned, become incompatible
if located next to each other.
The size of a land unit required to make management of a
use practical, the problems of conflicts between adjacent land uses,
and the need to organize uses across the landscape to take advantage
of roads or protect streams are all examples of the importance of
spatial factors in land allocation decisions.
Spatial factors are excluded from the most critical phase
of the land allocation process in the current system of land use
planning. Land allocation is performed in a single phase, in this
system, with use of either a linear or goal program. Land use
assignments, using this technique, are made on the basis of:
1) causing minimal disturbance to the land type to which they could
be assigned and, 2) achieving the maximum product value possible.
Spatial factors are treated only as a second step when the linear
program output is transferred onto a map of the planning unit.
These factors are treated only in cursory fashion: they are con-
sidered only if the individual doing the mapping recognizes that the
location of a land use will cause some type of resource management
problem.
Resolution of any recognized spatial problem usually is left
to the professional discretion of the individual doing the mapping.
In dealing with such problems, land use planners often are forced
to modify the initial land allocation produced by the linear program.
Modification of these results undercuts their Toptimalityhl in ways
that probably are not clearly understood and which probably vary
from planner to planner.
The Present Study
4
The general purpose of this study is to develop a strategy
and set of techniques for incorporating spatial factors into the
allocative phase of the planning process.
The general problem of location and land allocation is
discussed in Chapter II. Three primary spatial factors are isolated
and identified during this discussion. The chapter concludes with
a description of the effects of these factors on the land allocation
process.
The third chapter contains a review of the most promising
strategies and techniques which might be used for spatial allocation
of land uses. Criteria for a useful technique are described and
each tool is evaluated with respect to these criteria. One technique,
a heuristic search algorithm originally developed by J. P. Ignizio
(1978), satisfies most of the listed criteria.
The fourth chapter describes adaptation of this algorithm
to the spatial allocation problem. The problem itself is broken
down into discrete components and its overall structure is described.
The tools and procedures necessary to apply the heuristic algorithm
are also described.
The fifth chapter describes application of these tools and
procedures to a real problem. The Clackamas Planning Unit of
the Mt. Hood National Forest possesses many of the features of a
typical problem and is used to illustrate the application. This
chapter concludes with a description and discussion of the practical
difficulties encountered in the application of this strategy.
The final chapter summarizes the work completed in this
thesis and concludes with a suggestion for the next research steps
needed in managing the spatial allocation problem.
5
II. LOCATION AND LAND ALLOCATION
Only recently have spatial factors been excluded from the
land allocation process while planning for use of the National
Forests. Originally, site capability was the primary determinant
in the location of a land use. Land use decisions were made by pro-
fessional foresters who matched use to site. The decisions were
based on professional understanding of resource management
requirements and rules describing the Thighestit or best? use of
the land.
Three social changes altered traditional decisionmaking.
First, the absolute number of land use demands has increased
dramatically since the end of World War II (Public Land Law
Review Commission, 1970). Since the land base remains the same,
increased demands have resulted in more intense competition for
the limited resources of the forest.
Second, the scale of planning and decisionmaking expanded
to include larger areas and longer time periods. Land use
decisions in one area more and more frequently began to be
affected by the decisions made on neighboring areas (Hirsch, 1970;
Hufschmidt, 1969). These mutual effects increase the need for
coordinated land use decisions between areas and the consideration
of larger areas in the planning process. Similarly, longer planning
6
horizons result from the realization that environmental impacts of
many land uses extend far into the future.
The third change, probably an inevitable consequence of
increased competition and expanded scale, was in the decision
process itself. The first two changes require an individual to
consider more information when making land use decisions. Pro-
ces sing this information in any systematic way exceeds human
capabilities. Synoptic tools, such as mathematical programming
algorithms, and small-scale analytical models, such as computer
simulations, are more and more frequently replacing the in-
dividualts professional judgment (House, 1976).
The present land use decision is no longer made as a
single choice, a result of the professional's judgment about the
correspondence between land type and land use activity. Rather,
land use decisions are aggregates of choices, many made with
synoptic tools and models. The decision process itself is shaped
more by the strengths and weaknesses of the available tools than
by conscious structuring of choices and design of tools to assist
with particular problems.
Linear Program Models of the Planning Problem
The allocative phase of current Forest Service planning
7
processes is structured to a considerable extent by the use of linear
8
programming. The purpose of the land allocation process is to
create a land use pattern which maximizes the value of the consumer
products and recreational opportunities provided by the land while
minimizing the associated environmental impacts. A linear pro-
gramming algorithm is used by most Forest planning teams to
create this pattern.
Linear programming and the problems it addresses are
best summarized by Hillier and Lieberman (1974).
Briefly, linear programming typically dealswith the problem of allocating limited resourcesamong competing activities in the best possible(1. e., optimal) way. This problem of allocationcan arise whenever one must select the level ofcertain activities that compete for certain scarceresources necessary to perform those activities.
In the case of land use planning, the "limited resource is
the productive capability of the forest land. The productive capability
of a forest depends upon its physical and vegetative characteristics.
Since the forest's physical and vegetative characteristics vary a
great deal from place to place, forest land is sub-divided into land
type units, each with unique physical and biological characteristics
(Bell, 1976). These land type units become, in effect, the limited
resources of the forest.
The Itcompeting activities" in land use planning are manage-
ment practices which will occur when the land is assigned to a
certain use. Each land use will have a unique set of management
9
practices associated with it. Since an area of land often can be used
for more than one purpose at a time, the 'competing activitie&T are
often sets of management practices designed to produce certain
land use combinations.
Management activities of any sort produce consumer pro-
ducts, rec reational opportunities, and env-ironmental effects.
The allocation problem is regarded as one of assigning the limited
acreage in each land type to the competing activities so that the
maximum amount of consumer products and opportunities are pro-
duced while environmental disturbance is minimized.
An example of the land allocation problem formulated for
the linear programming algorithm and the type of solution produced
are shown in Tables I through III. As Tables I and II show, all
the acreage of a given land type is grouped together to form a
sum total of acreage for that type. In actuality, however, the
acreage for each land type is scattered in different sized parcels
throughout the planning unit (Figure 1). Therefore, the linear
programming model is insensitive to location.
TABLE I. THE LINEAR PROGRAM MODEL OF THE ALLOCATION PROBLEM
LT1 LTZ LT3Management Activities
Al AZ A3 AZ A3 Al A3
Variable (1) (2) (3) (4) (5) (6) (7)
units per acre
Goal 1 10 ZO 15 40 maximum
GoalZ 5 10 = 5,000
Goal 3 15 20 10 20 = 12, 000
Goal4 1.0 0.5 1.0 2.0 = 2,000
LT 1 (acres 1 1 1 = 1,000
LT 2 (acres) 1 1 = 500
LT 3 (acres) 1 1 = 800
Output Land Types Goals andConstraints
TABLE II
FORMULATION FOR THE LINEAR PROGRAM ALGORITHM
MaximizeZ = lOX2 + 25X3 + 20X4 ± lsX5 + 40X7
Subject to
5X1 + lOX2 + lOX6 5,000
15X1 + 20X2 + lOX4 + 20x6 - 12,000
TABLE III
LINEAR PROGRAM SOLUTION TO THE ALLOCATION PROBLEM
11
Acreage Allocations Goal Levels
X1 = 400 acres Goal 1 44, 000 units
X2 = 0 acres Goal 2 = 5, 000 units
= 600 acres Goal 3 = 15, 000 units
X4 = 300 acres Goal 4 1,900 units
X5 = 200 acres
X6 = 300 acresX7 = 500 acres
l.0xl + 0.5x3 + l.0x5 + 2. OX7 = 2,000
xl + X2 + X3 = 1,000
x4 + x5 = 500
X6 + X7 = 800
xl , . . . , x7 = 0
LT2
LT3
LT1
LT2
LT3
Figure 1. A Typical Land Type Map
LT1
12
Current Forms of Spatial Allocation
Current forms of spatial analysis in the land use planning
process typically begin with the linear program output and a land
type map of the planning unit. The land allocations produced by
the linear programming algorithm must be transferred to the planning
unit, creating a land use pattern. Since the linear program output in-
cludes the acreage allocations without reference to the location of
this acreage, special procedures must be taken to locate the
activities on the planning unit.
The process of mapping the activities onto the planning unit
is done entirely by humans. A plaimer must locate acreage of a
particular land type on the map and assign one of the allocated
activities to it. He must do this for the entire planning unit,
assigning every allocation while taking into consideration the acre-
age and location of each assignment. As the pattern of land uses
develops, the planner must consider each new assignment in the
context of those made previously. This consideration is crucial if
the final pattern is expected to achieve the value of forest products
and maintain the level of environmental impacts indicated by the
linear program solution.
The basic problem with the procedure is that humans cannot
optimize in problems of this complexity. They can't create an
13
14
optimal pattern for three reasons. First, all of the needed infor-
mnation can't be considered by a single individual at once. Second,
the individual often is selectively sensitive to the needs of land
uses or restrictions on resources which are most familiar to him
because of professional training. That is, selective perception
often causes an individual to overlook some spatial conflicts among
certain resources. Third, the individual has inadequate time to
search every possible combination of activities and compare every
The overall result of planners manually locating the land use
assignments is to undercut the optimality attained by the linear
programnrning algorithm. The assignments made by the planner
may meet the acreage allocations specified by the linear program
solution but, in many cases, will not produce the effects predicted
by this solution.
The scope of the problem becomes most apparent after the
pattern of activities is completed for the first time. A careful
examination of the pattern will turn up areas where conflicts between
activities will prevent achievement of the linear programming goal.
This inspection compounds the problem for the planner; he must
rearrange the assigned activities to eliminate this condition, not
create additional location problems, and still correctly assign the
acreage allocated by the linear program. In many cases, the
planner may not be able to anticipate all the consequences of any
possible rearrangement, forcing him to settle on a pattern which
fails to achieve the optimal condition.
Types of Spatial Factors
To discover what kinds of spatial factors intrude upon land
allocation decisions, a number of Forest Service planning documents
were examined and the land use planning staff of several National
Forests in Region Six were interviewed. Most of the staff members
had never considered the spatial allocation problem in any depth and
therefore had limited perception or knowledge of the spatial factors
crucial to a land use allocation process. This fact eliminated any
possibility of using a survey of planners to develop a description of
the crucial spatial factors.
However, these discussions, combined with my study of
the Environmental Impact Statements (the land use plans), did
produce a number of examples of spatial problems in land allocation.
Examination of their common characteristics indicate that they could
be grouped into three basic types about which generalizations could
be made.
The three underlying spatial variables affecting land
allocation decisions are:
15
16
management unit size,
adjacent-use spillovers,
collocation patterns.
In the land allocation process, one of the critical spatial
factors is the size of an individual management unit. Certain
activities require a minimum number of acres grouped together
before they can be imposed on the land.
Examples of activities which require a minimum manage-
ment unit size include some wildlife habitats, timber harvest units,
and wilderness areas. The wildlife habitat for some species, such
as pileated woodpeckers, must have a minimal acreage of a par-
ticular land type to maintain the species (biological requirement).
The timber harvest process requires management units of sufficient
size to offset the fixed costs associated with access and management
of the timber (administrative requirement). Originally, by
Congressional designation, wilderness areas had to contain a
minimum of 5, 000 acres (administrative requirement).
The management unit must contain sufficient acreage so
that the activity assigned to the site can meet the applicable bio-
logical, physical, or administrative requirements. Since the reason
for the requirements varies from activity to activity, the minimum
management imit size varies for each activity. In addition, the
appropriate minimum unit size also depends on the land types
17
invoLved in the allocations. Some land types are able to satisfy the
requirements for a particular activity with fewer number of acres.
Therefore, the minimum mangement unit size depends on the
activity under consideration, the applicable requirements, and the
land types involved.
Another critical spatial factor is the 'spilloverT caused
by locating some activities adjacent to one another. Adjacent land
uses may produce both positive and negative spillovers, depending
on the activities involved. A planner must decide whether the
aggregate impact of these spillovers produces a desirable or an
undesirable result. If the total effect is desirable, then deliberate
steps should be taken in the allocation process to locate these
activities adjacent to each other. Undesirable spillover effects
require relocation of the activities to avoid their adjacency.
An example of a desirable grouping arises in the establish-
ment of wildlife habitat. Some wildlife species, such as elk, require
a certain combination of vegetative habitats for forage, shelter,
and reproductive functions. An example of an undesirable grouping
is the placement of timber management next to developed camp-
grounds. Timber harvesting activities often create noise, hazards,
and undesirable scenery. The result of such placement would
produce a land use pattern which would fail to provide the recre-
ational opportunities dictated by the linear programming solution.
18
The third prominent spatial factor in the allocation process
is collocation. "CollocationT refers to the process of deliberately
locating activities on the planning unit so that they form a particular
pattern across the face of it. The pattern to be formed depends
upon the special requirements of some activity or upon the spatial
organization of the planning unit as a whole. For example, two
prominent physiographic features frequently involved in the location
of activities are road systems and streams. Developed campgrounds
usually are assigned to the land next to the road system. In order
to minimize the environmental effects created by automotive camping
and protect the aesthetic attractions of the campgrounds, it is
desirable to space the campgrounds some distance from each other.
However, it is also important to consider travel distance and access
in designing developed recreational opportunities for an urban
public
Collocation and adjacent-use conflicts obviously have much
in common and may not actually be mutually exclusive variables.
It may be that adjacent-use conflicts simply depend more on the
nature of the activity than the land type while the reverse is true
for collocation problems. Or, the two factors may vary only
according to the scale of the spatial factors involved in a specific
planning problem. For the purposes of this study, however, the
magnitude of differences between these factors and the fact that
they must be treated in different ways, seems to warrant treating
them as two distinct spatial variables.
19
III. PROMISING ALGORITHMS
There are two major flaws in the current methods of trans-
lating linear program allocations onto maps of planning units.
First, perception of location conflicts and judgment about their
resolution vary from individual to individual. Second, individual
planners are unable to carry in their minds all the information
needed to make synoptic location decisions. The first flaw can
be overcome if planners know what spatial factors to take into
account, if they have a systematic procedure for making activity
assignments and, if they have a uniform set of rules for resolving
conflicts in the location of activities. Yet, the amount of detailed
information to process in the spatial analysis problem still would
prevent planners from finding the most efficient land use pattern.
The spatial allocation problem, even though extremely
complex, is one of decisionmaking under the condition of certainty.
That is, the effects of assigning any activity to any location can be
known. The feature of certainty in the problem implies that there
is an optimal or most efficient land use pattern for any set of
activity-location assignment values. The troublesome problem is
how to design a procedure or system of decision rules (i. e. , an
algorithm) for finding the most efficient land use pattern.
A computerized algorithm capable of searching for the
20
21
optimal pattern would improve the present situation in several ways.
Its use would decrease the selective perception and variability
between people, since the technique would standardize treatment
of the spatial factors in the assignment procedure. An algorithm
increases the efficiency of the user in two ways. It drastically
reduces the time required to solve highly detailed problems. More
importantly, it allows the user to focus his attention on the value-
judgment components of a problem, rather than on its computational
aspects. Finally, because the decision rules are publicly available
in the algorithm, they can be scrutinized over time, compared with
others, and amended as knowledge of spatial influences on land use
decisions grows.
Limiting Factors in the Search for an Algorithm
A great number of promising algorithms exist in the field
of operations research (Phillips, et al., 1976; Ackoff and Sasieni,
1968). These include the specialized procedures of mathematical
programming, which are designed for problems of resource
allocation under conditions of certainty (Hillier and Lieberman,
1974; Simmons, 1972; Taha, 1971; Wagner, 1970). Two elements
of spatial allocation problem greatly narrow the search for an
algorithm among the mathematical programming techniques.
The first restricting element is the choice of principal
22
criterion to be used in the algorithm. There are two broad options.
One is to strive for the optimal solution to the total problem. In
this case, the linear program would be replaced with some other
optimizing technique. The algorithm would find the location for
all activities which would maximize the value of consumer products
and opportunities developed in the planning unit while meeting
environmental constraints. The alternative is to retain the linear
programming technique and use the results it produces as the
criterion for location decisions. Some special algorithm could
be devised for making location assignmeits: these assignments
would have to sum so as to match the acreage allocations first
provided by the linear program solution. The second option
implies that the criterion of Ttefficiency would be substituted
for "optimalityt' and that ?Iefficiencyll would be judged in relation
to the linear program solution.
The second limiting factor in the search for an algorithm
is the size of the problem. Size is expressed by the number of
units of analysis and the number of variables in the problem.
Essentially, the greater the size of the problem, the more
numerous the decisions to be made in assigning a land use activity
to a land type.
Size of the problem determines whether it can be handled
by a computerized algorithm or not. The more numerous the
An Optimizing Algorithm
A strong argument can be made for adopting optimality as
the principal criterion for the proposed algorithm. The linear
programming model of the planning unit does not take the spatial
organization of land type units into account, yet spatial factors
will affect the products and opportunities which can be developed
in the unit. The linear program model may be unrealistic and,
therefore, the "optimal solution' it produces may also be un-
realistic. A genuinely optimal solution could be developed if the
linear programming technique were replaced by another optimizing
algorithm which coild take location into account.
An integer programming algorithm in place of the linear
program might overcome the difficulty associated with locating
23
decisions to be made, the less likely the problem will fit within
existing computer capabilities. The larger the problem, the less
likely that an optimizing algorithm can be used (since these require
simultaneous consideration of a great deal of information), and
the more likely the problem must be broken into sub-parts in
order to solve it. As indicated above, the larger the problem,
the more likely that an efficiency solution algorithm criterion
must be selected.
Promising Algorithms
24
acreage allocations on the planning unit. The problem formulation
for the integer program is similar to the linear programming
formulation except that the acreage in the planning unit would be
divided into "management units" rather than land type units. While
the linear program allocates acreage from each land type to the
various activities, the integer program allocates each management
unit to a single activity.
Acreage in the planning unit could be divided along land
type boundaries to create the management units. The larger blocks
might be split into two or more management units in order to
allow the algorithm to allocate this acreage to more than one
activity when optimal. Another possibility would be to create a
management unit out of the acreage of two or more land types.
This tactic would be useful when there is insufficient acreage of any
one type to support a specific allocation.
Since the allocation of a management unit to a particular
activity is essentially a yes-or-no decision, a zero-one integer
programming algorithm could be used. This type of algorithm
examines the contribution which the assignment of each activity to
a particular management unit would provide to the value of consumer
products and recreational opportunities, then selects the assignment
which meets all the constraints (environmental or spatial) and
contributes the most value (Geoffrion and Mar sten, 1972; Hillier
25
and Lieberman, 1974).
Since the algorithm assigns an activity to a management
unit and the location of that management unit is known, the operator
can now consider the relative location of activities in the allocation
process. That is, the operator can include additional constraints
in the problem formulation to control the allocation of management
units based on their location.
The most common use of this constraint would be to make
the assignment of an activity to a particular management unit
depend partially on the activities assigned to adjacent management
units. Constraints of this form either strictly prohibit adjacent
assignment of certain activities or place a penalty on the occurrence
of this condition.
The zero-one integer programming formulation of the land
allocation problem with location constraints is presented in
Appendix A. This information can be compared with that provided
for the linear programming solution technique in Tables I through
III.
The major advantage of the integer programming algorithm
is that it would consider both minimum management unit size and
adjacent-use conflicts in producing an optimal assignment pattern.
However, the algorithm in this form cannot consider the collocation
factor. Unfortunately, the algorithm also runs into a number of
26
practical disadvantages.
The problem formulation for most planning units would
probably exceed the computer core size required to solve the
problem. Any attempt to make the problem fit the computer core
size, by reducing the number of management units and increasing
their size, affects the precision of the solution and may make it
invalid. A number of runs usually are required in the allocation
procedure to determine solution bounds under various constraints
and also to produce alternative allocations from which to select.
The total cost of all these runs may be prohibitive. So, although
a zero-one integer programming algorithm may be a good
theoretical replacemeiit for a linear program, practical limitations
hinder its application.
An Efficient Solution Algorithm
The practicality of a pure optimization algorithm is limited
by the size and complexity of the spatial allocation problem. The
alternative is to seek an efficient, rather than an optimal solution.
The integer programming technique also might be used for this
purpose. Most of the problem formulation would remain as above,
but the integer program would contain only three types of con-
straints. These would be: a) the acreage assigned must be equal to
the linear program acreage allocations, b) only one activity could be
Z7
assigned to each management unit, and c) the assignment of one
activity would depend on the assignment of other activities. This
problem formulation is also presented in Appendix A.
Dividing the spatial allocation problem into two smaller
problems permits both the linear and integer programming to fit
within existing hardware capacity. However, a number of linear
program runs are required to bound the allocation problem and
create alternative solutions. Plus, the zero-one integer program
must be run a number of times for each alternative linear program
allocation. Each of these runs include a different composition of
spatial constraints. Some constraints are relaxed and others are
added until all important spatial factors have been considered in
the final assignment pattern. The problems are still so large that
the cost of making several runs of each algorithm is prohibitive.
Few National Forests would find the solution costs for a normal
problem to be a reasonable investment.
An Assignment Algorithm
A more promising possibility employing the efficiency
criterion is to substitute an assignment algorithm for an optimizing
algorithm. Ramalingam (1976) notes that assignment algorithms
are particularly compatible with the efficiency criterion in resource
allocation problems. They,
. . arise in the context of associating eachof the N requirements to each of the N avail-able means of satisfying them (requirements).
. . One of the important characteristics ofthe assignment problem is to allocate resourceson a one-to-one basis. . Another character-istic is that the total number of requirementsequals the total number of resources. Theeffectiveness coefficients indicate what mightresult from allocating each requirement toeach resource. . . . The objective is toestablish the assignment that minimizes the sumof effectiveness coefficients of the selected com-binations.
Assignment algorithms overcome the size problem, but at
the expense of finding an optimal solution. Assignment decisions
are made in a step-by-step fashion. At each step, after the
assignment is made, the size of the problem is reduced for all
subsequent steps. The algorithms consider the current assignment
and its affect on future options, but can't consider possible adjust-
ments to earlier assignments which could improve the final
solution. Essentially, assignment algorithms can overcome the
size problem because they do not consider all possible land use
patterns.
Ignizio (1978) has developed an assignment algorithm which
can be adapted to the spatial allocation problem. His algorithm
searches the array of options created by potential assignments
and selects that assignment (land use activity location) which
maintains the greatest number of future possibilities. Accordingly,
28
his algorithm, 'Tthe heuristic search algorithm, can deal very
well with the adjacent-use spatial factor. Although the algorithm
cannot handle the collocation factor directly, the problem formu-
lation could easily be structured so that the desired collocation
pattern appears. The algorithm does not deal with the problem of
minimum management unit size, but the procedure in which the
algorithm is embedded also could be developed to take this factor
into account.
Problem Formulation for the Heuristic Algorithm
Il the heuristic technique were to be used, the formulation
of the spatial allocation problem would have the following
characteristics:
There must be a finite number of management units in
the planning unit.
Each management unit can be assigned to only one of a
finite number of activities.
Because of the characteristics of adjacent activities,
assignments will be promoted where a favorable condition arises
or avoided where an adverse situation appears.
The acreage of the management units assigned to each
activity should equal the acreage allocation determined by the
linear programming algorithm.
29
30
(5) The primary objective will be to assign activities to
management units in such a way that the impact of each adjacency
condition is considered and the acreate allocations are met.
The heuristic approach requires the computer to make a
number of conditional checks; each of which narrows the selection
process substantially and allows the activity assignments to be made.
The procedure followed by this approach is briefly outlined below.
The planner must arrange the management units in some
order. Since the computer algorithm will assign the management
units in that order, the planners may wish to place certain units
near the beginning or order them by decreasing acreage.
For all adjacent management units, the planners must
indicate the occurrence of any favorable or unfavorable conditions
which will result if certain activities are assigned. With this
information, the computer algorithm starts the assignment process
by considering the first management unit. The computer checks
its land type, acreage, and all linear program acreage allocations
of that land type. The acreage of that land type may be allocated to
any number of the activities. The computer program compares
the management unit acreage to the acreage allocations to determine
which activities are eligible for assignment to this management
unit. If the management unit acreage is the smaller acreage for
a particular activity, then that activity is considered for possible
31
as signment.
The computer repeats this comparison for each activity
until it has identified all eligible activities. The computer program
then checks the impact each possible assignment has on future
assignments to adjacent management units. The favorable and
unfavorable conditions created by each possible assignment are
compared and the activity which creates the most desirable situation
is selected. When the assignment of two or more activities are
equally desirable, the activity with the largest remaining acreage
allocation is selected. The acreage of this management unit is
deducted from the appropriate acreage allocation.
The updated acreage allocations represent a running tally
of the remaining acreage of each land type to be assigned to each
activity. The algorithm must consult this tally to determine the
eligible activities for each management unit and to assure that the
algorithm is attempting to meet the linear program acreage
allocations.
The same assignment procedure is repeated for each
management unit, except now, the algorithm also must consider
previous assignments made to adjacent management units. The
favorable and unfavorable conditions created by these activity
assignments will further restrict the selection process. The
algorithm continues through the list of management units, attempting
32
to assign an activity to each unit.
However, in some cases, no activity will be eligible for
assignment. This occurs because of two constraints which any
activity must satisfy before it can be assigned to a management
unit. The assigned activity must not: 1) create a conflict with any
previously assigned adjacent activity and 2) cause the acreage
assigned to exceed the linear program acreage allocation by more
than a specified tolerance. After assigning all management units
possible with these constraints in effect, the algorithm stores the
assignments and a matrix indicating how close to the acreage
allocations the assignment process came. Then, the algorithm
allows the user to terminate the assignment process or continue it
with one or more of the constraints relaxed.
Subsequent runs assign the remaining management units
based on the activities previously assigned. The constraints are
relaxed one after another in the following order: 1) strict prevention
of adjacency conflicts with previously assigned activities is
eliminated and activities are selected which simply minimize all
possible (current or anticipated) adjacency conflicts, 2) the tolerance
level for achieving the acreage allocation is eliminated bit an activity
still cannot be assigned if that allocation was previously satisfied,
and 3) all attempts to meet the acreage allocations are dropped
and the activity is selected which minimizes all possible adjacency
conflicts. Since the results of each run are stored, the user can
examine the assignments at each step to get a feel for the location
of conflicts or select one of the intermediate runs and assign the
remaining management units manually.
A more detailed description of how the computer algorithm
carries out the above procedure and an example of its operation
is presented in the following chapters.
Summary
Inability to treat spatial influences in land allocation
decisions in any systematic way has caused them to be ignored in
the allocation process itself. A computer algorithm which could
take these factors into account would be a useful tool for the land
use planner and would help reduce the present variability in
allocation decisions among National Forests. Unfortunately, the
size of the spatial allocation problem prevents us from making use
of an optimizing algorithm, such as integer programming. This is
so even when a new unit of analysis, the management unit, is
created to replace the land type unit and reduce the size of the
problem.
The best alternative seems to be to sub-divide the problem
into smaller components. Sub-division allows us to retain the
linear programming tool and devise a specialized algorithm for
33
34
making location assignments. Sub-division also implies that the
objective of the spatial allocation problem is to find the most
efficient approximation to the linear programming solution.
The collection of assignment algorithms is compatible with
this objective. Ignizio's heuristic technique, in particular, provides
a simple and effective way of taking the two most difficult spatial
factors into account when assigning activities to management units.
This algorithm has several advantages over the optimizing strategies
explored. It is relatively simple to understand and inexpensive to
use. Unlike the optimizing routines considered here, it can be
modified to take the collocation factor into account. The major
difficulties with the algorithm are that it does not produce optimal
solutions and it may not produce consistently efficient solutions
over a wide range of spatial allocation problems. Optimal solutions
cannot be developed because the algorithm cannot search all possible
improvements to the first assignment. The consistency of the
algorithm is unknown at this time and will have to be tested before
the algorithm could be widely used.
IV. THE SPATIAL ALLOCATION STRATEGY
A number of computer programs were needed to perform
certain tasks in the spatial allocation process. The tools developed
include: a) a computer mapping routine, ) a detail reduction
program, c) an adjacency program, d) a spatial assignment pro-
cedure, and e) a conflict detection program. The mapping routine
creates management units and keeps track of location. The detail
reduction program eliminates some of the complexity in a land type
map. The adjacency program detects adjacent management units.
The assignment procedure utilized in the spatial allocation process
is the heuristic approach presented in the previous chapter. The
conflict detection program is a computer routine designed to detect
activity assignments which produce detrimental spillover effects
when they are located next to each other. It also assists the planner
in identifying assignments which violate the minimum mangement
unit size rule. This chapter describes the tools in greater detail.
It focuses on their functions in the overall process and the linkages
between them.
Tools Needed for the Procedure
The Computer Mapping Routine
The spatial arrangement of activities on a planning unit
35
36
affects achievement of the goals described by the linear program
acreage allocations. Some device is needed for keeping track of
the locations of the management units, and later, of the activity
assignments.
A number of computer mapping routines are available which
are able to keep track of location. In most cases, they are unable
to provide all of the mapping functions needed or handle the type of
data used in land use planning. The E-ZMAP (Child and Rollin,
1976), IMGRID (Sinton, 1976), and SYMAP (Dougenik and Sheehan,
1976) computer mapping systems each contain many of the required
mapping functions but they are unable to handle the amount of data
needed. That is, the number of symbols needed to display land
types, management unit5, or activities usually exceed the symbol
list available in each of these mapping systems.
Some mapping systems may satisfy both requirements but
were not available for use in the present research. The RAP mapping
system, for example, is recommended for adoption throughout the
Forest Service for use in land use planning, but is currently in-
compatible with Oregon State Universityts computer system. The
Mt. Hood National Forest, the only Forest in the Pacific Northwest
to make use of a computerized mapping system, is currently using
a modified version of R3MAP. This program contains functions
necessary to process and display planning unit data. The functions
37
include display of maps and their legends, an ability to combine
maps, and an ability to aggregate symbols. These functions are
used to develop maps of land type units by combining maps of in-
dividual forest resources. The modified R3MAP program uses
some computer system features not currently available on Oregon
State University's computer.
The computer mapping routines discussed above all use
cellular mapping techniques. The map information is displayed in
square cells of a fixed acreage. The symbol assigned to each cell
represents some characteristic of the acreage of that cell. All of
the acreage inside the cell boundaries is assumed to have similar
characteristics. Since the characteristics of the planning unit are
not laid out in neat squares of fixed acreage, each cell symbol
represents the dominant characteristic of the acreage in that cell.
Mapping routines, like R3MAP, have standard functions
which are used by planners to create land type maps. The current
land allocation process limits the mapping routine to this contribution
and possibly to a display of the final land use pattern. However, the
proposed spatial allocation procedure requires additional information
and data processing which only an expanded computer mapping
routine can provi.de. The activity assignment procedure requires
the planning unit to be broken down into management units and
information provided on their acreage, land type, and the adjacent
38
location of other management units. The conflict detection program,
which detects detrimental spillover effects and provides information
on management unit size, also requires additional information not
available from conventional mapping routines. This information
includes each management unitrs acreage, land type, and adjacent
management units, as well as the activities assigned to it and the
adjacent units.
In light of these requirements, a greatly expanded version
of the R3MAP routine was built to support the spatial allocation
procedure. The functions provided by this mapping routine are
listed in Table IV along with a short explanation of their purpose.
A program listing of the mapping routine is presented in Appendix B.
The Detail Reduction Program
The land type map of many planning areas may contain a
great deal of complexity. That is, in addition to a large number of
land types, each land type is broken into small parcels and these
are scattered across the planning unit. Each parcel of each land
type must be treated as an individual management unit when the
mapping routine creates the management unit map. A large number
of single-cell management units will result; pushing the total
namber of management units beyond that which the computer memory
can contain or which can be handled at reasonable cost.
SUBROUTINE PURPOSE
DIRECT
EXCESS
TABLE IV. MAPPING ROUTINE FUNCTIONS
39
The map directory. This mapping routine is aninteractive computer program which allows up totwenty maps to be available to the operator duringa single run. The directory lists and keeps trackof the available maps, aiding in the selection ofmaps for other mapping functions.
Handles the map overflow. When the number ofmaps input or created exceeds the twenty allowed,this subroutine locates one map on the recodeddata file.
COPY Transfers maps onto the recoded data file.
RECODE Recodes map symbols into a standardized numericcode and collects statistics on the occurrence ofeach symbol.
IN Inputs maps in either of two formats.
SELECT Selects and stores maps on the recoded data fileat the completion of a run.
OUT Displays maps at the terminal or outputs them ona line printer file.
COMBO Combines the characteristics on several mapsinto a set of symbols on a single map. Eachsymbol on the new map represents a uniquecombination of the original characteristics. Thelegends of the original maps are also combinedto create a legend which describes each uniquecombination.
AGGASS Aggregates or assigns cells of each symbol intogroups and documents those groups. Each groupconsists of any number of the original symbols.
Table IV. Mapping Routine Functions (Cont.)
SUBROUTINE PURPOSE
Reorders the symbol list and legend of a givenmap based on a selective order or by decreasingacreage size.
Searches a map for symbols which share one ormore common characteristics and lists thedocumentation associated with these symbols.
Detects and groups cells of the same symbolwhich are adjacent to one another.
Changes the titles and/or legend of any map.
Manipulates map cells: moves individual cellsfrom symbol to symbol, aggregates cells of onesymbol with another, and splits cells of a singlesymbol into groups of several symbols.
40
The number of management units must be reduced to an
acceptable level without reducing the accuracy of the final land use
pattern. The easiest and most practical strategy is to group the
isolated single cells of each land type with the adjacent group of
cells exhibiting the most similar characteristics. The detail
reduction program (REDUCE) is designed to perform this operation
on a land type map before it is broken into management units by
the mapping routine. A program listing of REDUCE is presented
in Appendix C.
REORDER
HILITE
PARCEL
UPDATE
MANIP
The Adjacency Program
The spatial assignment procedure and conflict detection
program both require a list of adjacent management units. The
assignment procedure uses the adjacency information to avoid
assigning activities next to each other which produce detrimental
spillovers when located on adjacent management units. The conflict
detection program uses information about adjacent management
units to identify any detrimental spillovers which appear in a land
use pattern.
The adjacency program (ADJ) examines a map of marage-
ment units, identifying the management units adjacent to each unit
and recording the number of times each adjacent pair occurs. The
program listing of ADJ is presented in Appendix D.
The Spatial Assignment Process
41
The heuristic approach described earlier in Chapter III is
used here. The program listing of the heuristic (HEURIST) is
presented in Appendix E.
The Conflict Detection Program
The conflict detection program is a computer program which
provides information about the management unit size of assigned
activities and detects detrimental spillovers created by the assign-
42
ment of activities to adjacent locations on the planning unit. This
program simply provides the planner with information which he can
use to make a quick check for violations of minimum management
unit size or the location of any adjacent activities producing detri-
mental spillovers.
The size of a management unit is checked after activity
assignments are made. A single activity may be assigned to
adjacent management units, not necessarily of the same land type,
to produce a larger management unit for that activity. This is done
by ignoring the old management unit boundaries and creating new
management units based on adjacent cells assigned to the same
activity. The mapping routine carries out this process. First, it
assigns the activities to the land type map, erasing the management
unit boundaries. Then, adjacent cells assigned to the same activity
are grouped into a parcel or !mnewhl management unit. The adjacency
program detects the activities assigned to locations next to these
management units.
This information is transferred to the conflict detection
program. The computer program examines each management anit,
checking its acreage against the minimum acreage figure provided
for management units assigned this activity. Lf the acreage is
smaller than the minimum required, the program lists the assigned
activity, the management unit acreage, and the activities assigned
43
to adjacent management units. This information, plus a map display
of the assignments, allows the planner to detect violations of
minimum management unit size.
Detection of detrimental spillovers can use either the fold?1
management units or the TtnewT ones developed in the above pro-
cedure. The Holdu management units will pinpoint the location of
those management units which should be reassigned by the spatial
assignment procedure in the next run. However, the tmnewfl manage-
ment units indicate the true acreage assigned to an activity. The
"true" acreage of a management unit is important since sufficient
size often overcomes detrimental spillovers by providing a buffer
for the activities. However, if a detrimental impact still exists
(as shown by the "new" management unit), the planner must refer
back to the uoldtI management unit map to locate those units needing
reassignment. Having both options available allows the planner to
select the information important in different circumstances.
The planner loads the conflict detection program with:
a) information on the activities assigned, b) the management units,
and c) a matrix indicating which activities will produce detrimental
spillovers and the severity of these spillovers. For each management
unit, the program checks the activities assigned to adjacent manage-
ment units. The spillover matrix tells the program if a detrimental
spillover exists between the activity assigned to the management unit
44
being considered and those adjacent to it. If a detrimental spillover
is found, the program lists the severity of the spillover, the
activities involved, and information about the management units
involved.
The conflict detection program provides the planner with
information about the land use pattern which is difficult to detect
with only a simple visual examination of an activity assignment map.
The information can be used by the planner to improve the spatial
assignments made in a subsequent run. A further explanation of
this process is described in the next section of this chapter. A
listing of the conflict detection program (DETECT) is located in
Appendix F.
The Procedure
The procedure can be described most easily by referring to
its major stages. These stages are shown in Table V.
Inputs to the Spatial Allocation Procedure
The spatial allocation procedure is only one step in the
entire land use planning process. Rather than reconstruct the
entire process here, the results of intermediate steps in the
planning process will be treated as inputs to the spatial allocation
procedure. The two inputs are: 1) a complete land type map with
TABLE V. THE SPATIAL ALLOCATION PROCEDURE
INPUTS TO THE SPATIAL ALLOCATION PROCEDURE
REDUCEPARCELMANIP
Land type map.Linear program acreage allocations.
DEVELOPMENT OF MANAGEMENT UNITS
Eliminate isolated single cells.Break into management units.
*Modify size or shape of management units.
*DEVELOPMENT OF A SPILLOVER MATRIX
ACTIVITY ASSIGNMENTS*Preassign activities to some management units.
REORDER *Place management units in order.ADJ Find adjacent management units.
Input linear program allocations.Input spillover matrix.
HEURIST Make activity assignments.AGGASS Transfer assignments to the map.OUT Display activity assignments.
Check allocation achievement table.
*These steps can be modified in a readjustment procedure.
45
DETRIMENTAL SPILLOVERS AND MINIMUM MANAGEMENT UNITSIZE
the management units. This algorithm allows the planner to pre-
assign activities to some management units and then assigns
activities to the remaining units. The praassignment feature
fixes activities to specific locations on the planning unit, forcing
the algorithm to fit the other acreage allocations around these
locations.
The order in which management units are assigned affects
the final land use pattern. The heuristic algorithm assigns an
activity to a management unit only if it: 1) helps to meet the
acreage allocations produced by the linear program, 2) does not
create an undesirable spillover condition with an activity which was
previously assigned to an adjacent management unit, and 3) pro-
vides the maximum number of future assignment options by mini-
mizing the potential spillovers with the adjacent unassigned manage-
ment units. Each assignment restricts the activities which can be
assigned to subsequent management units by: 1) meeting part of
the acreage allocation and 2) blocking future assignments which
produce undesirable spillovers on adjacent management units.
Subroutine REORDER in the mapping program permits
arrangement of the management units in any selected order, will
arrange the units by decreasing acreage, or will allow a select
group to be placed first with the remainder arranged by decreasing
acreage. Arrangement by decreasing acreage causes the larger,
50
51
more critical management units to be assigned first, leaving the
smaller units to adjust and fine-tune the final pattern.
The heuristic algorithm, as previously stated, is concerned
with activities assigned to adjacent management units. Therefore,
the computer program ADJ is used to detect all adjacent manage-
ment units and this information is input to the heuristic algorithm.
Other inputs to the heuristic algorithm include the linear
program acreage allocations and the spillover matrix. The
function of both in the assignment procedure was also discussed
earlier.
The heuristic program attempts to assign activities to
management units in such a way as to meet the three conditions
previously listed. The first two are constraints: 1) restricting
each assignment on the basis of previous activity assignments to
adjacent management units and 2) preventing the assignment of an
activity which will exceed the acreage allocations by more than a
specified tolerance. With these constraints in effect, the assign-
ment procedure usually is unable to assign every management unit.
Subsequent runs relax each of the constraints, one at a time, until
activities can be assigned to all of the remaining management
units.
The results produced by each run can be checked by the
planner. The heuristic program stores the assignments made up
to that point and updates their effect on the achievement of the
acreage allocations. This last piece of information is presented
in an allocation achievement table which the planner uses as one
measure of the success of this assignment run. The assignments
for each run also can be displayed on the planning unit map through
the use of the mapping program. The subroutine AGGASS transfers
the assignments to the management unit map and the subroutine OUT
displays the assignment map at the terminal or places it on a line
printer file.
Detrimental Spillovers and Minimum Management Unit Size
The conflict detection program, DETECT, provides two
important functions to the spatial allocation procedure. It checks
the map of activity assignments for undesirable spillovers and
violations of minimum management unit size requirements. This
program simply detects and lists the occurrences of these
situations; it does not include procedure to correct them. How-
ever, this does not diminish its importance. This program
detects those undesirable spillovers which could not be avoided
in the assignment procedure, listing them so that the planner can
take steps to eliminate the unacceptable spillovers in a later
readjustment. It also provides the only systematic list of those
assignments of insufficient acreage; the only other option being
53
an individual inspection of the assignment map.
There are two strategies for the detection of undesirable
spillovers. Both require the input of the spillover matrix, but
differ in the rest of the required inputs. Both strategies list certain
basic information: 1) the severity of the detrimental spillover,
2) the activities involved, and 3) location on the map. One
strategy focuses on management units of a single land type and
includes the land types of the management units involved in the
spillover. The other strategy focuses on management units of a
single activity and the acreages provided reflect the true I size of
the management units involved.
The first strategy, emphasizing land types, requires the
input of activity assignments, the list of adjacent management units,
and the documentation of the management unit map. The conflict
detection program uses this information, plus the spillover matrix,
to list every undesirable spillover on the planning unit.
The second strategy, emphasizing the htruel? management
unit size, requires a number of additional steps. The activity
assignment map is input and subroutine PARCEL groups the
assigned acreage into a new' set of management units based on
the activities. The adjacency program, ADJ, finds the adjacent
management units in this new map. The conflict detection program,
in this case, uses the spillover matrix, management unit map of
54
the assigned activities, and list of adjacent management units to
detect the undesirable spillovers.
The conflict detection program also uses the newly developed
management unit map of assigned activities, the list of adjacent
management units, plus a list of minimum acreages to identify
all management units which violate minimum management unit
size. The program screens each management unit against the
minimum acreages, listing the location, activity, acreage, and
adjacent activities of any management unit which is smaller than
the minimum.
Readjustment Procedure
Major readjustments of the land use pattern can be
carried out in additional runs of the spatial allocation procedure
by changing certain inputs. The spillover matrix can be altered
to promote or prevent a particular adjacent activity assignment
in subsequent runs. The order in which management units are
assigned can be changed or some of the management units can be
preassigned activities. In some cases, a further step back in
the process may be required. The management units developed
from the land type map can be changed, with additional aggregations
of smaller units or divisions of the larger units. All of these
readjustments should cause changes in the final activity assignments.
The selection of which adjustments to make has to be based on
experience with the assignment algorithm and an inspection
of the assignments produced with the current set of inputs.
55
V SPATIAL ALLOCATION OF THE CLACKAMASPLANNING UNIT
This chapter describes the application of the spatial
allocation procedure developed earlier to a representative planning
unit of the Mt. Hood National Forest. Illustration of the procedure
through a case study probably is not necessary to provide an
understanding of its underlying strategy. However, the general
complexity of the procedure seems to warrant the time and
expense involved in the development of a case study. In any newly
developed procedure with the complexity this one contains, there
may be hidden flaws or unforeseen details in the planning unit
which prevent the procedure from being fully utilized. It seemed
wise to see whether the spatial allocation strategy outlined could
survive the test of workability.
As anticipated, there were difficulties which required
modification of the original spatial allocation procedure. The
major complication was that the size of the problem, or the
complexity of the land types in the planning unit, exceeded most
expectations about them. This chapter, in addition to describing
the planning unit and illustrating use of the procedure, describes
the difficulties presented by such complexity in the planning unit.
56
The Clackamas Planning Unit
57
The Clackamas Planning Unit is located on the Mt. Hood
National Forest in Oregon. It consists of 353, 779 acres of forest
land oii the western slopes of the Cascade Range, south of Mt.
Hood. There are three vegetative zones on the planning unit as
designated by the dominant climax tree species: 1) western hemlock
(with Douglas-fir as sub-climax species), Z) Pacific silver fir,
and 3) mountain hemlock. Within these three zones are 51
distinct forest plant communities. The unit also contains 1, 500
miles of rivers and streams, plus 160 ponds, lakes, and reser-
voirs. The diverse habitats found within the vegetative zones
support a variety of wildlife including 73 species of mammals, 150
species of birds, and 25 species of reptiles and amphibians.
The current use pattern includes management for timber,
developed recreation, dispersed recreation, and wildlife. Timber
harvest is a dominant activity on this planning unit, providing 187
MMBF/year or fifty percent of the total output from the Mt. Hood
National Forest. The biological potential for the unit with pre-
commercial thinning, genetic stocking, and full stocking level con-
trol is 199 MMBF/year. The major commercial species include
Douglas-fir, western hemlock, Pacific silver fir, noble fir, and
grand fir. The Clackamas Planning Unit also contains Z4 camp-
grou.nds with 417 camp units and 106 picnic areas. There are two
roadless areas in the unit, Bull of the Woods (43, 735 acres) and
Olallie (8,673 acres). There is also a total of 180 miles of trail
throughout the planning unit.
There are two reasons the Clackamas Unit was selected to
illustrate the procedure. First, the Mt. Hood Planning Staff
volunteered both the problem and their assistance in developing the
case study. 1 Second, the planning unit is representative of both
the size and diversity of the Forest-wide planning units to be used2in the near future. Physical characteristics and current use
patterns are typical of the majority of National Forests in the
Pacific Northwest Region.
The Spatial Allocation Procedure
The six stages of the spatial allocation procedure presented
1The Planning Staff provided all the data needed for thisstudy and served as a sounding board whenever unforeseen diffi-culties arose.
2The Clackamas Planning Unit acreage is actually threetimes the size of the typical planning unit used during the last fiveyears. However, recent changes in land use planning regulationsinstruct the Forest Service to regard National Forests, ratherthan their sub-divisions, as the proper unit of analysis. MostNational Forests have begun the shift to Forest-wide planning.The Clackamas Unit is actually a composite of three earlier,unfinished units and is being treated by the staff as a prototype forthe entire Forest.
58
in the previous chapter are:
inputs to the spatial allocation procedure,
construction of management units within the planning unit,
construction of the adjacent-use spillover matrix,
assignment of activities,
detection of detrimental spillovers and violations of
minimum management unit size,
readjustment of assignments.
Two variations were made from the procedure shown in
Chapter IV. First, the information received from the Mt. Hood
Planning Staff was not in the correct form to be input into the
procedure. It contained some detail unnecessary to an illustration
while lacking other data important to the demonstration. The
information submitted was modified to an acceptable form. The
second variation has to do with the last stage of the procedure.
Readjustments to the land use pattern were not made in the case
study as they might be in a real planning problem. The case
study is designed to demonstrate the procedure; there is no
need to correct any unrealistic results.
59
3Additional information on each of these stages is providedin Table V and the descriptions contained in Chapter IV.
Inputs to the Spatial Allocation Procedure
The information received from the Mt. Hood Planning Staff
includes a digitized land type map of the Clackamas Planning Unit
and a list of activities for which the unit can be managed. A portion
of the digitized map is shown in Figure 3. The activity list
describes twenty-seven activities: six of these were chosen for
purposes of demonstration in this study. These representative
activities are:
reserved sites,
developed recreation,
dispersed recreation,
commercial timber,
visual timber,
wildlife.
Each activity contains statements of the impacts it will
have on the total Forest and products it will provide if assigned
to the Forest. The physical characteristics of any land type unit
which would limit or prohibit its assignment to that land type are
also described. An example of an activity description and site
identification is shown in Table VI.
A seventh activity, lakes and streams, was added to the
list. Thj addition was made so that the lakes and streams would
60
.10i303flJ33kWL ..000L1030 13 4 $ 9
4 36
..* ..;k 93 S
+
$ ..L3OUk39393 QL3b O313b Z
4 /33939339 393 UbD6 1
$ cIS
4 .39393U939313UL3OO39Ei b
4 3031Di0OA U93i39d98t9O V'
+ 606uLJ933kUi3b3% L
4 .bi0) 9
4
4 i 9393L3O
Z
4 31O1O1b313D oaa i'
o
$ 3639d 9339939393L3 b
+ W
4 / 4 9
+ S
4 iiI o OO&66bUOUOUO113 .
. .0 9O9399a9O633 &b131130(4O ic
4 Z
4 " 4 o
+ .3b3 L313930Cb 3963CQ 9u69agog09og09O9O9a9b9U b +3 31 99i 9O9U9O9O'Q99$939
4 ..9 3v1313 39o9O99O9393é3 31313 99S9O9O9O9O9O9O9O998 i
The spatial allocation problem involves three important
spatial factors: 1) management unit size, 2) adjacent-use conflicts,
and 3) collocation patterns. Construction of strategies which
optimize land allocation locations, when these factors are taken
into account, generally are prohibited by the size of the allocation
problem. Even when the problem is sub-divided and an efficiency
criterion substituted for optimization, optimizing algorithms
either exceed computer memory size or are prohibitively ex-
pensive. Assignment algorithms are more useful and are suitable
for the objective of finding the most efficient approximation of the
linear program solution.
Igniziots (1978) heuristic technique, a search and assign-
ment algorithm, can be modified so as to treat the three spatial
factors as variables in the assignment problem. Before the
technique can be used, the problem must be broken into several
components, then tools developed to handle each component. The
tools and procedures developed through this research are not
completely satisfactory, although they are workable and make it
possible to deal with spatial factors in a systematic manner.
85
Difficulties in Application
86
A number of difficulties are encountered when applying the
spatial allocation procedure to an actual planning unit. These
difficulties include: 1) the amount of detail, 2) computer core
limitations, 3) algorithm efficiency, 4) the amount of information,
5) the number of interactions, and 6) the lack of exact rules for
making tradeoffs. Each of the difficulties will be discussed in a
general manner.
The amount of detail refers to the complexity of the
planning unit. The diversity of land types plus their location
creates a complex and detailed pattern. This complexity forces
solution techniques away from mathematical programming algo-
rithms (which can only handle a limited number of variables)
toward algorithms with more flexible limits. It also strains
computerized data manipulation routines, such as the mapping
routine developing management units.
The computer core limitations refer to the amount of
information which the computer can load and process at any one
time. These limitations become important in problems where a
large number of factors must be considered at the same time. Any
mathematical programming technique attempting to include location
in its allocations encounters this limitation. Several operations
87
in the mapping program also face this problem. However, in the
mapping program, this limitation is overcome by sub-dividing
the problem. This strategy also can be applied to mathematical
programming problems, but formulation is difficult and multiple
rutis make a final solution quite expensive.
Algorithm efficiency is reflected in the cost of obtaining
a solution. The solution cost is affected by: 1) the amount of
information considered by the algorithm, 2) the required pre-
cision of the solution, and 3) the number of runs necessary to
calculate a solution. Mathematical programming algorithms
considering location and factors associated with location in their
allocations must sub-divide the problem into a number of smaller
problems, each of which must be solved. This, coupled with the
requirement of an Moptimalhl solution for each sub-problem, usually
creates a solution with high costs. Cost also may be prohibitive
for other types of algorithms depending on their design, their
solution requirements, and the amount of information they process.
The amount of information included in an allocation or
assignment algorithm must be examined in light of computer core
limitations and solution cost. These factors sometimes require
a simplification of one or more aspects of a particular problem.
In the spatial allocation problem, the linear program acreage
allocations and adjacent activity assignments were considered
88
directly in the assignment process while minimum management
unit size and collocation requirements were encouraged with the
spillover matrix but only checked after assignments were made.
The number of interactions between the information
examined when making assignments adds to the complexity of any
assignment algorithm. When a tradeoff exists between two factors
affecting the solution, the algorithm must consider this special
case (and every other special case) or must have rules to avoid
these complications. Consideration of every case will increase
the amount of information to be examined, possibly causing com-
puter core or cost problems.
The lack of exact rules occur in the case of interactions
and tradeoffs. The problem arises because of the complexity of
the spatial allocation process and the number of factors involved in
development of a land use pattern Attempts to anticipate every
tradeoff and develop a rule for each are virtually impossible and
create additional complexity in solution procedures. The best
approach may be a simplification of the problem with a standard
procedure which ignores tradeoffs.
Tool Improvements and Subjects for Further Investigation
This study has exposed the problem of spatial allocation in
forest land use planning. There is room for improvement in the
89
tools developed. The problem requires greater study in its own
right and the heuristic technique is in need of reliability and efficiency
te sting.
The adjacency program, for example, is limited to working
with only those management units sharing a common boundary. Ex-
amination of spillover impacts are restricted to those located along
a boundary. The adjacency program, as a result, cannot account
for the effects of a single activity assignment on the overall land
use pattern. A useful improvement to the procedure would be a
program which kept track of the relative location of all management
unit assignments. This program would allow the proximity of
other activities to be considered in each assignment and, thereby,
allow the collocation factor to be treated directly in the procedure.
The spillover matrix, or the values in it, require further
study. The values in the matrix represent the spillover effects
expected if two activities were to be located on adjacent manage-
ment units. The matrix is a useful device for making identification
of spillovers and the judgments about their importance explicit.
However, the values necessarily are a simplification of several fac-
tors. These include: 1) the type of effect expected from proximity
of two activities, 2) the impact of this effect on production of
consumer goods and services, 3) the sensitivity of each activity
to the type of spillover, and 4) the relative importance of the types
90
of spillovers identified for the matrix. Very little is known about
these spillovers and their effects on production of goods and services
from a planning unit. Even though explicit in the allocation process
developed here, the value judgments in the matrix remain sub-
jective.
A related task is to more thoroughly investigate the spatial
factors involved in land use allocation on National Forests. Three
factors were identified here. It seems unlikely that these are the
only spatial factors or that they are the most important ones. It is
not even certain that the three identified here are mutually
exclusive factors. Further research into the identity of spatial
factors and their influence on land use allocations seems warranted.
On another tack, planning problems vary in size and com-
plexity. It would be useful to know how size of the problem affects:
a) computational efficiency, b) the flexibility of assignment options
mentioned earlier, and c) the sensitivity of the solution given by
the heuristic technique. If the problem is very large, the planner
will make simplifications in variables and reduce the level of
detail in the units of analysis. Similarly, it will be important to
know how such reductions, when made, are likely to affect the
accuracy of the final solution.
The most troublesome shortcomings are in the heuristic
technique. Since the 11ocation algorithm (or strategy) seems
91
relatively independent of the kind of spatial factors involved in the
problem, high priority should be placed on further study of this
strategy.
The heuristic search algorithm looks to future assignment
options and considers previous assignments when selecting an
activity to assign to a management unit. However, the algorithm
is unable to reassign management units even if such an action were
determined to be the most beneficial to the final land use pattern.
The heuristic search algorithm could be modified so that
it is able to reassign management units. This modification should
permit an improvement in the land use patterns which is currently
prevented by the algorithm.
The heuristic algorithm assigns activities to as many
management units as possible without violating the values in the
spillover matrix. To assign the remaining management units,
the spillover constraint is subsequently relaxed and activities are
assigned which fulfill the acreage allocations and minimize the
detrimental spillovers. However, there is no assurance that a
land use pattern would not have been created with fewer detrimental
spillovers had this relaxation always been in effect.
Detrimental spillovers block assignment options in the
heuristic search algorithm. There is probably a range of detri-
mental spillovers which is acceptable in the heuristic algorithm.
92
Outside of this range the spillovers disrupt the algorithm, per-
mitting assignments to only a few isolated management units. The
acceptable range of these spillovers would be useful information
to a planner who may wish to select only the most severe conflicts
in order to create a land use pattern.
The tradeoff between computational efficiency and assign-
ment options caused by the number of management units should
be investigated. There is a need to establish a range in the number
of management units which fits the computer core limitations, can
be manipulated with some degree of efficiency, and provides
adequate flexibility in the assignment of activities to management
units. The range would assist planners in determining the amount
of detail the land type and management unit maps could contain,, if
the hearistic assignment algorithm was being used in the spatial
allocation procedure.
Finally, the reliability of the heuristic algorithm is un-
known. It seems likely that the ability of the technique to produce
the most efli.cient possible solution is affected by the number of
variables in the problem and by the range in these variables.
Tests should be conducted to determine how sensitive the algorithm
is to: a) a change from the use of positive, negative, and zero
values from the spillover matrix to the use of a full range of
values from +3 to -3, b) variation in the spillover matrix values,
and c) the order in which management units are scheduled for
assignment.
93
BIBLIOGRAPHY
Ackoff, R. L., and M. W. Sasieni. 1968. Fundamentals ofOperations Research. Wiley Sons, Inc. , New York.455 p.
Bell, Enoch F. 1976. Goal Programming for Land Use Planning.USDA Forest Service General Technical Report PNW-53,Portland, Oregon. 12 p.
Child, Dennis R., and Linda S. Rollin. 1976. E-ZMAP: Com-puterized Map Overlay Procedure. Range Science SeriesNo. 23. Colorado State University, Fort Collins,Colorado. 43 p.
Dougenik, James A., and David E. Sheehan. 1975. SYMAPUser's Reference Manual. Harvard University, Cambridge,Mass. 187 p.
Geoffrion, A. R., and R. E. Mar sten. 1972. Integer programmingalgorithms: a framework and state-of-the-art survey.Management Science 18:465-491
Hillier, Frederick S., and Gerald J. Lieberman. 1974. OperationsResearch. Holden-Day, Inc., San Francisco, Calif. 800 p.
Hirsch, W. Z. and S. Sonenblum. 1970. Selecting RegionalInformation for Government Planning and Decision - Making.Praeger Special Studies, N. Y. 198 p.
House, Peter W. 1976. The Quest for Completeness. D. C.Heath & Co., Lexington, Mass. 236 p.
Hufschrnidt, Maynard M., ed. 1969. Regional Planning: Challengeand Prospects. Praeger Special Studies, New York. 396 p.
Ignizio, James P. 1978. Solving large scheduling problems byminimizing conflict. Simulation. March, 1978. pp. 75-79.
Phillips, D. T., A. Ravindran, andJ. J. Solberg. 1976. Oper-ations Research: Principles and Practice. Wiley & Sons,Inc., NewYork. S85 p.
94
Public Land Law Review Commission. 1970. One Third of theNation's Land. Government Printing Office, Washington,D. C. 342 p.
Rarnalingam, P. 1976. Systems Analysis for ManagerialDecisions. Wiley &z Sons, Inc., New York. 607 p.
Simmons, Donald M. 1972. Linear Programming for OperationsResearch. Holden-Day, Inc., San Francisco, Calif. 288 p.
Sinton, David F. 1976. An introduction to I. M. G. R. I. D.: aninformation manipulation system for grid cell datastructures. Harvard Graduate School of Design, Dept.of Landscape Architecture, Cambridge, Mass. 23 p.
Taha, H. A. 1971. Operations Research: An Introduction.Macrnillian Press, New York. 626 p.
Wagner, H. M. 1970. Principles of Management Science.Prentice-Hall, Inc., Englewood Cliffs, New Jersey.562 p.
95
APPENDICES
96
APPENDIX A
97
98
USE OF INTEGER PROGRAMMING FOR MAKING ALLOCATIONS
Problem FormulationI = activity
1j = management unit
Objective Function
M.AXZ
-WT (d++...+d+ +d +d+ +...+d+d+)2 1 m m+1 m+1 n n
subject to2Allocation Constraints
a11X11+a12X +... +a X .b12 lj lj
a X +a X +...+a.X.b221 21 22 22 2j. 2j
a. X. + a. X. + ... + a. .X.. = b.11 11 i2 i2 ij ij
Adjacent Conflict Constraints3
X11 + X2 + d - 4 = 1
x +x +d d+ =1m m
Management Unit Size Constraints4
X11 - X1 + d+i - d++l 0
+X -x :+ddn = 0
5Multiple Choice Constraints
xli + x2l + ... + xii = 1
xl2+x22-... +xi2 =1xii + x2 - ... + xii = I
Zero-one Constraints
Xfg = 0 or 1 for every f = 1, . .. , i and g = 1, .., j
1 The objective function maximizes one goal or product while
minimizing the adjacency conflicts and management unit size
violations. The weights are used to indicate the relative im -
portance of these two objectives.
2 The allocation constraints are similar to those in the linear pro-
gram formulation except each management unit is assigned in its
entirety to a single activity.
3The conflict constraints prevent two adjacent management units
from being assigned to conflicting activities. One constraint is
required for each pair of management units where the conflicting
situation may arise. The possible situations which can arise are:
1 + 1 + 0 - 1 penalty
0 + 0 + 1 - 0
1+0+0 -00 + 1 + 0 - 0
Since only the d+ variable is penalized in the objective function,only the first situation (where the conflicting activities areassigned to both management units) produces a penalty and is,therefore, avoided.
4The management unit size constraints require two adjacent manage-
ment units to be assigned to the same activity. The possible
0 - 1 + 1 - 0 penaltySince the same activity is to be assigned to both management unitsor neither, a penalty is attached to the situations where the activity
is assigned to only one of the management units. This situation
arises in the last two cases when either the d or d+variable takes
99
on a value other than zero.
5The multiple choice constraints assure that each management
unit will be assigned to one and only one activity.
6The zero-one constraints prevent the variables from taking on
any values other than zero or one.
USE OF INTEGER PROGRAMMING TO ASSIGN ALLOCATIONS
Objective Function'
MIN Z = WT(d +... + d + d + d +... + d + d)1 m m+1 m+1 n n
Allocation Constraints
a. X. + a. X. + ... + a. .X.. = b.ii ii iZ 12 1J 1J 1
Conflict Constraint3
Management Unit Size Constraints3
Multiple Choice Constraints3
Zero-one Constraints3
'In this use of the objective function, the algorithm must simply
minimize the spatial assignment penalties.
2The allocation constraints require the management units assigned
to each activity to meet the acreage allocations. These constraints
can also be formulated so that the allocations do not have to be met
exactly but a penalty will be paid for deviations from the allocation
levels.3The other constraints are identical to their previous use in the
integer program making actual allocations.
100
APPENDIX B
101
102
UVEL4Y(lA.S,0,0) I
PROGRAM NAIN(iNPUT6 ,QUTPUTb5,TuP 9ZNPUT, TAPE50OUTPUT, 23
20 C THLS PROGRAM IS A nAPPIG RCUTIME WHICH CARRIES OUT A NUMBER OF 20
U SPECIALIZEO FUMc.T1SNS. MAPS A'ii3 THEIR ASSOCIAIEO DOCUMENTATION 21C ARE iNPUT £f1j THE P GRA1 AN3 THE MAP SYMBOLS ARE RE000EO £NTD 22. A WUMEKICA.. CUJE siiIGM A.jOW FURTHER OPERATIONS ON THE MAPS. 23C.
2.25 C THESE )PEr,ATIONS INCLUDEI 25
C t) OUTPUT £N 4 TAHJARJIZED FORM AT THE TERMINAL OR ON A 26
C LINE 4TEH FI.E, 27C 2) COMBiNATIN OF MAPS TO CREATE A MAP WITH A NEW SET OF 28C SYMBO REPRESENTING EAOH UNIOLIE MAP COMBINATION, 29
30 C 3) AGGREOATION OF lAP SYMBOLS lc.IO ASOCIA1EO GROUPS, 30
C ) SYMBO_ OHAES FROM A MP ASiC.NMENT, 31
C 5) REOROER HAP SYMBOLS ANtO LEc.EMO 32C 6) A OIRICTOw.Y OF MAPS AMCLAB..E O THE USER, 33
C 7) HIOH.ZGrIf OF SOME CIA CrE.iSTICS OF THE OUCUMENTATICH. 3'.
35 C 8) UTQ1..TCC .OUPING OF ADJACENT CELLS OF THE SAME SYMBOL 35
C INTO ?AGELS, 36C 9) UPOAT: Or NAP TILES AND LEGE4OS, 37C tO) IANPJLATION OF MAP CELLS. 38O 39
1.0 C THESE 3PERATON, NIL 3E 3ETTER DESCRIBED IN EACH SUBROUTINE. .30 THIS POGRA1 uPE,ATE ON AN INTEAOTIAE AASIS THE USER RESPONDS '.1C TO QUESTIONS AND ON MOST GASES CAN SELECT AMY MAP OPERATION .2
C AFTER COMPLETIDN OF THE PREVIOUS PERATION.C 1.'.
1.5 C THE TAPE uSAGE FO. Trill PRO1SIAAM IS) 1.5
O TAPES 1-20 TITLES AND LEGENDS FOR 20 lAPS 1.6C TAPES 2t- MUTUAL MAPS MSSOOITEO WITH OUCUHENTATION ON .7
C TAPES t-20 1.8
C TAPE .J ..INE PRINTER OUTPUT FILE 1.9
50 C TAPE '.1. cCOOEJ DATA OUTPUT FILE 53O TAPE 1.5 RECODED DATA INPUT Fi.E ASSIGNMENT INPUT FILE 51
o TAPE .0 RAW DATA INPUT FILE, ASIGNMENT INPUT FiLE 52
C WING FILE 53C TAPE .7 SYMBOL LIST LNPUT FILE. WORKING FILE 51.
tool FORMAT 4/,A,AMAPPING ROUT INE/CREATEO BY R. HAGESTEOTI 62toA,ASEPr. IB7t) 63
C 61.
65 wRIIE(50, 10321 651002 FDitMAT(/THE FOLLOWING FILE NUMBERS ARE SET./ 66
lATHE USER MUST OE CONSISTENT WTH THEIR OESIGNATION./ 072.3*,ALINE P..INTEP. OUTPUT Fi_E ('.31 Al33X,.tRECOOEU DATA OUTPUT FIiE ('.) Al 69
70 1.JA,tRE0OJEJ DATA iNPUT FILE ('.51(/ 7053A,ARAW (AIM t4PUT FILE (1.o) All 716AWNEN A AUE$TioN IS ASKED, ANSWER WITH A YES OP. NO.Al 727AWHEN M)(E TMMN ONE NUIIOER IS TO BE PLACED ON A SINGLE A, 73IALIME,t,ASEP..RAIE THEM HITH BLANKS OR COMMAS.AI 7'.
C 76CC THE MAPS WHICH ARE OUTPUT FROM THE PROGRAM CONTAIN A STANOAROIZED 78C SET OF 5YH3DLS. THE FO.LOWING COOE iNPUTS THIS SYMBOL LIST FROM 79
103
SO C TAPE'.?. IMIS FILE IS LATEI uSED 3! TiE PROCRAM AS A O.KING FILE. 80C 81
90 C THE FOLLOiI CUOE SET U? AHiS CAL.3 THE INPUT £UOUTIHE. THIS 90
L SU0WUTNE MILL. INUT MAPS IN ET THE OR ECOOEO DATA FORM. 91C IN SwE CAI. 'IiIERE A PREIOU. RU JUST lADE AND THE MAPS ARE 92C ALEAOY P.ADED ON LOCAL FILE5 t-'.0 THE NUM3ER OF AAILMdLE MAPS 93C IS SUPPLIED AND TH INPUT P,OGEDURE CAN 3E SKIPPED.
95 C 95C 96C 97
RITE(5U, 100.) 98t001 FUKIAT(/tIS THEr.E MN! MAP DATA TO SE INPUT. 99
110 C 110C 111C 112C THIS CODE ALLOWS SEtECTIN OF A MAP CPERATICN AND CALLS THE 113C APPOPIATE SU3ROUTIHE. 11'.
115 C 115C 116C 117110 WcITE(5Q,t007) 1161007 FORMAT(/tiST OF lAP OPERATIONS AND THEIR KES./ 119
120 AVTO A C.VEN 3PERTjN - TrP IN THE KEV.// 1203VLIST - TO GET THiS LISTV/ 121C*TERMIt.ATE - To TErNINATE NA? OPEIkATIDNS*/ 122UOUTPUI TO oUTPUT riAPSt/ 123EtGOMdZr4C - To COMBINE MAPS AND DDCUMENIATION*/ 12'.
125 FtAGGEA1E - TO AU,..,ECATE NAPS AND UPDATE 000UNENTAIIONX/ 125- TO iAKE lAP AS5ICNMtTS AND UPDATE DCGUMENTAIION/ 126
HtSEU,OEA. - TO EOOE THE lAP Stl8OLS AND LEGEND/ 127I*DIV.E,1(Y - TO EAANINE THE DLkEDTORVt/ 128JVHI,HLI,4T - 10 HI,HLiDMT CM MCTERISIIG(S) OF 0OCUP4ENTATI0N/ 129
ISO K:PACEL - 10 U.DUP AJONCENT CEL INTO PARCELSt/ 130LtUP)ATE - TO UPDATE MAP TTLE DR .GENO/ 131MIMANIPJ11E - TO IANIPULUTE PARCEL MAPS) 132
C 133120 IIE(5U,tQ0a) 13'.
135 1008 FOiATL//:,EECT A lAP OPERATIONs) 135REAO ('.9, 1009) KUOE 136
1009 FURlA1R3 137C 136
IF(KUOC.EU.3HLIS) GO TO 110 13011.0 IFKUUE.EQ.SHDU() DO TO 130 1.0
IF(KUOE.EU.3iCUH) CD TO 1'0 141IF(KUDE.LN.3H.UI CO TO 153 11.2
1F(KOOE.EU.3Hu) SD TO 150 1.3IF(KOOE.EQ.JNrEO) U TO 150 1'.'.
11.5 ZF(KOD.U.3HDIR0 GO To jc,JIF(KOO..E.3HIlCU0 GOb 170 1.6IF(KUOE.EU.3HPA) CU TO 183 1'.?IF(KOUE.EQ.SHUPO) CD TO L0 11.6IF(KOOE.EO.3H1AN) GO TO 20 11.9
150 IFOUE.EA.JlTER0 (J 10 210 150C 151
WRITE(50, 1010) 1321010 FCRMAT(f1rlA1S NOT ONE OF THE GHOICES? 153
GO 10 110 15'.
155 1.30 CALL DELAY('.P(MAPS,3,0) 155GO TO 120 186
11.0 CALL 0VELA!(1.P(MAPS,'.,0)TaO tO 120 158
104
150 CALL OLY(HMPO,5,3)C.O TO 123
toO CML. OIECT ttCo 10 120
170 C.L_ OVE.L.Y(,HMAPS.b,3) to3GO TO 123 tOL
165 180 (..LL O(LMY(M.iS.7,3 165GO TO 120 166t0 CALL OE(LAY(HMAPS,8,3 167GO TO 123 164
200 CALL 3V'LuY('HMAPS,9,3)170 GO 10 123 170
C 171172
C 173AFTER ALL IP J.Eb(4TLU H4T 3EEN C0IPLETEO OUT 0EFO1 PROGRAI t7'
175 (. TE,MZIsATLO, THE UOROUrLNE SELECT IS CALLEU. THIS U8OUTINE 175C 4LLON THE JE TO SELECT ECUJEO MAE'S TO 8E SEO IN THAT FO?i 176C FOR iNPUT Ol U8jEQUET L. THE SELECTEJ MAPS AE PLACEO 117C ON T4PE.',. 178C 179
180 C 180181.
210 CAL. OE.LAYL'MMAPS,2,0) 142C 153
WRITE(50,tatL) 13'.105 1011 HMPS HAVE 6EEN L4CEO ON FILE '.'./ 145
t$kEHdE TO VE THAT FILE/iTHE LINE PINTE. OUTPUT HAS 1862t8EEN PLCEJ ON FLLE '. /*(EMEMER TO OLTE THAT FILEt) 187
C 148STOP 139
190 190
105
t SU3UTjNE ECT t1uI1ENSIl Ti_E(6 192
193C 19'.
5 l5C t6C SI4CE T1E 'G*M IS INTEAC1IE, IT IS UIPORTAT HAT T Y POINT 17C IN Ji 1-sE KNOW HICl MAPS ME AILA3LE NICH MP 198c NUMBER C lrj 1T :OA)N. THiS IS C1IED aT uROuTINE 199
tO OECT. ir P00ES ri AP UME AO TITLE OF EACH AVAiLABLE MAP. 200C 201C 202C 203
WiITEI50,10Jfl 2U.t5 1001 FORlT(/OiET0RY FOi QV(.AYS, AGGREGATIONS, ANO C04IATIQNSt/) 205
C THIS J3OJtit OWS FO. lOE THA TEMT'f LAPS REPLCIM& OE OF 22C ThE TwET ILT1 rHE EAE i.P. T1E PLACEO P CA 3E Li$IMATE 22C O MY 3E P_A.EJ u ECJQED MAP OUTPUT FiLE. THIS FILE MA(E$ 233
to c THE$ iAP$ FO LATER 23t232
C THE TITLE )F TIE NCOMiN& hAP 4M TPE OiETOY OF MAPS ELI&I8E 233C FO EPuCiENT AE 2'.1..
2515 23ô
C237
WRtT(5O,tO01) (TITLEtI) IL,S) 238lOut F1T(/.tT1I 1API ,ôAII/.l1tJST 3E wITTEM 3ER AN 23g
ItEAiSTI i14P..) 240c 2kt
MFILE:ZI 2.2CALL. OECT 2.3
C2+.
wIT(5Q, 1Q2)25 lOU .AT(/t.ELEL THE APPP,UAT lAP MUME fl 2+â
1F 27C
2'.8L 24.9C
30 C IF TE EAisr1N MAP IS TO d $VE SU3ROUTIME COPY IS LALLEO. 25tC IT WLL Pk)E TE EXTIG AP a&FCE TPE MEW ?AP £S P4ACEO 252C ON THE FI.E. 53
C2.255
35 CWRjTE(O,1OU3I 257
1003 FtATL/O OL wH TO PLACE T$E EXISTING ?IAP ON TIE RECODED t, 251IAP UT3UT FE4fl 259
CCCC THIS SUKOuTIr4 15 CA..LEQ 5 OTHER jUROUTZNE WHEN THEY WANTC THE S'r430L5 UN 4AP ECODEO ACCOOZNO TO THE ODEk OF THEIRC OCCURrENCE ZN THE EGEUC. THIS URCUTINE ALSO kiCOkOS THEC CE*CE OF SYISOL IN THE HAP, so srrxsrios SUCH ASC NUHOER OF ACREAGE, AtD ERCETGE O ACREAGE CAN AEC C.UIPOT FOR EACH SYHAOL.CC1.
TOTAC=0.REWIND :
CCCC THE FULLCii CODE CHECKS EAC*i CEL IN EVERY MAP ROW AGAiNST THEC LEGEND OROEcING INFORMATION, RECODES THAT CELL ACCORDINGLY, AND1. WlITES THE CUUO QOW ON THE HAP FILE. A TALLY IS ALSO MADE OFC EACH SYMOUL OCCURRENCE.CCC100 READ (.o, loOt) £CO,IRUW, (MAP(I .1=1,36)1001 FORMATl2*.,36A2)
1F(IC.EQ.VV) GO TO 1901F(1Cu...T.t.ANO .MAP(t).E.2H AND.HARGIN(IROW).HE.ZH)
1 MAP(l):HAN(1ROW)IMMP :2H
CDO lQ 1 3oIF(HAPU) &.2H I GO TO 11)3IF('I.*P(I).E0.H) 6) TO 1.70IHAPMM?(i)GO TO 12)
110 IF(<OO.E.lHr) GO TO 160IF(LIAP.EU.2r1 P GO TO l0MAPI 11=1 lAP
120 DO 130 .l,IS13)3 1F(MAP(L).E9.ICOCE(Ll) GO TO 1*40
THE FOLLOWING CuOE COMP,..ETES THE OOCCMENTATIUM FILE FOR THE MAP.C FIRST IHE TLT4.E5 MRE TRANSFERRED, T*iEN SYMSOL OY Yl8OL, THEC LEGNa IS IAO ACREAGE AND PERCENTAGE ACREAGE IS CALCULATED,C AND THE EPANOED DOGUMEHIMIION Z WRITTEN ON THE HAP OOCUHEMTATIOMC FILE.
1(3 C tHIS 1$ THE FiRST SU9.)u1IE CALLED WHEN THE PKOGkAM IS EXECUTED. l+1C IT KEA3S 'S IN 80TH RECjJEO AND KAW DATA FDvl FROM FILES .5 ANO '.6. w20C THE o0CUMENrATEO FC. IlIESE MAPS IS PLACED ON FILES 1-20 +NLLE THE .21C MAPS IHEMSEdES ARE P..ACEJ ON FILES 21-.0. '+22C 1.23
15 C THE kE000EJ M4PS Mr.E .NFIJT 5I+CE THE RECODED FORM 15 USEO IN '.2'.C THE Pr0CRA11 (r(L COOE SIMPLY TRANSFE.j THE MAPS F.OM THE RE000EOC DATA £NPUT IsE TU EriE APP PP.IATE FLEA (TAPES 123 AND 2t1+0J,C MAKANIA THEM u4ILA5LE TO THE USER. 1+27C
2!) C 1+29C 1+30100 REAJ(.5,tQ01I NT,NC,l5,(TITLE(I),It,6j '+3110(31 FIAT(212,I.,oAtO) '+32
IF(EOF('.5 I t..Q,tt0 '+3325 c
CCC THiS CODE IA USED HENEER THE NUMBER OF AAILA8LE MAPS IS 1.37C INCREASED (EITHEi A NEW MAP I it+PUI U ONE £5 CREATED 1 A '+38
30 C MuP OERATI.IN). THE JM5E OF AMALE MAPS 15 iNCREASED, 1.39C THE FILE UMBERS Fi.K THE NEW HAP ARE UESIGNATEO, AND A C$EC1(C IS MADE TO SEE IF THE MEW IA? PUSHES THE NUMBER UF AWAILABLE 1.41C MAPS OEYOND THE FILE SPACE FOR TWENTY MAPS. IF THE NEW luP 1+42C IS NUMBER IWENTY-UME, 5uOUTINE EXESA IS CALLED To HANOLE +43
55 1003 FUR'IAT(j,,12M10 '+6'.IFUC0L.El.9o99) GO TO 1110 '+o5GO To 1311 '+56
C
6(3 0 l+9C THE MAPS IN Ri.W OuTA FORM ARF. REO IN. THEIR SYMBOL LIST AND LEGENO '471C ARE kE000ED. ..NO THIS INFORMATION LA CARRIED TO ROUTIN RECOOE '+71C WiEA.E THE RIDDOING POCES IS COMPLETED. THESE MuPS ARE PLACED ON '+72C THE UNUSED FILES AMONG TAPES 1-20 ANO 21-'.!). .73
C '+82C THIS CODE 1 uSED Wi1ENEER THE NUMBE OF AAILA3LE lAPS IS '+83
75 C INCREASED (EITHER A t+W MAP IS INPUT OR ONE IA CREATED BY A '48'.C MAP OPERATION). THE NUMBER OF AV,I.ABLE MAPS IS INCREASED, +85C THE FILE NJMBRS FOR THE NEW MAP ARE DESIGNATED, uNO A CHECK 1.86C IS MADE TO SEE IF THE NEW MA? PUSHES THE NUMBER OF A'+AILABLE '+87C MAPS BEYOM3 THE FILE SPACE OR TWENTY lAPS. IF THE NEW MAP 1.88
111
81) C IS NUPI3E 1WEiTY-ME, SUOROIjTINE EXES £5 TO HANOLEC TH .00CC.C150 MFiLEHL+t
MFMFIL.E '.05LF1F+Q *06iF(lFI.G1.2U1 Cui EXCE5 '.07
65 £COL2IJL 630IF(ICO..EQ.9999) CO TO 200 631.WcITE(l(I,1IJ10) ICOL 632
1010 FQRMAT(1A//1,COLUMN ,IL./) 633C
63L.
70 C 635C
636C THE FOLLUWENt. COQE CONVERTS TuE MAP CELLS TO SIM3OLS USINC THE 637C S1MoQ LiST INPUT AT THE 4EGNNiNC OF THE PROCRAM, THEN WAjTES 638C THE MAP ON TH ELECTEO OUTPUT OEiCE. 639
75 C6'.0
C 6.1C
64.2150 00 190 1=1 36 64.3
jF(NAPU).E.0) 00 TO 160
114
80 IF(iP(1.EQ.-99gJ GO 10 170 óW5LFMP(1J Ei.-d1 GO TO 130 o..6MAP( iijYM3&.(MAP(j 647GO TO 130 5.48
160 MAP(I2185 GO TO 100
170 MMP(.)21' 551GO TO 131 652
180 HAP(I21ee 653634
90 190 ..CNTINUEC 055
WITE(KZ,1.01U 6571011 F0.i.MAT (1( ,.ZA,3oAZ,,t
GO TO 13395
i:. 661662, THE F)LLQWIG GOOE WRITES THE LEGENO ANO ASSOCI.ATEO DOCUMENTATION 6O3
C ON THE 3E_ETD uTPu( EI.CE. 86'.100 C 885
C 6o6i:. a7200 W.LTE(KII1O12 6681912 FMMAT(L// NO. GELS,'.X,AEAGE,5X,tPCT.t, 859
WRITE(50,10201 6371020 FCRMAT(/0TMEl Ol.TPUTeJ 658RE0.3 10021 IU 655
1.35 IF(IU.E.1Hfl GO TO 100 700c 7012.30 RETURN 702
ENO 703
115
t 70L.PkOGQ H UM3O 75OIME4SI4 iMuP2Q),,l4Pj(3óJ,lAP(3),MAP3(3ô), 70677
5 2MUi.1123) ,MOC(2U) 7o709710
COMl//iYMr3L(2.ô 711CCMlONfl/TT..E() 712
tO C 713I;. 7t.C 715C Ti1L J3.OUTE CQM9I4E EEcAi.. HS INTO A SINGLE MAP Y 716C A3ir& NEW l3QL. TO ECi1 UNIQUE CJM0UATION CREATEU DURI4G 717
15 C T-4E PQCES. TilE EEJ5 OF T1E M..P IE LS COMINE, kESULTIN 718C IN A NW L.ENO WhiCh CJNTAINS ?lULTLNEJ UECkIPT1ONS FOk EACM 7tC SYM8O..C 71C 722
20 C 72310 WITE5O, LOOLJ 729lOol FQi..T(/tHu MANY NAPS O YU I3H TO COM1ME+fl 725EAO(9,) J. 7SIF(J.E.0 &.e TO .00 77
25 ciRLTE(5O.1U32
1002 FOiMT(/iJO YQU WLSM TO jE THE IETO.Y T AlO IN 7301YCU $E.ET4fl 731RE(,10Q3 J 73230 10J FOiMT(UJ 733
(J.E.1,iY CA.J. LRECT 73C 73
WR1TE(5l3,13.. 736100k F1MT(/tI EACh t4P £N THE UIWE OF CQM3LATION. 737
35 1PiJCE HE 4AP E5 ON A IE dITi1 EACH / 7382PlAP NUlE,k EAiMTJ a! OtiM. 739kEA3(',9,') (P(,Li.,J 740
C 741wRITE(0, 100 7.2
10115 FO4T(ItETE NAME 3F T1 C 1iNATIOl (5 CHARACTERS) 7.3REAO(49,10iI, (TITi(K),KL, 7.L
1000 FOi($AT(1OJC 7'.C 7.7C 7'.dC THIS COOE IS uEO WNEt1VE T1E UME F AiAE LAPS ZS 7+C ICREE (ELTHE. NE MP £5 INPUT O NE £ CETED A 70C MAP CPEs.ATLj. THE UN3E UF uIAE MAPS IS INCREASED 71
THE FILE NJlE FOR TilE $ 4P A ELNTE0, A3 A CIECK 752C 1$ MADE TO EE T1E Ew N.P PuHE.) TE PUE OF VILaLE 753
$AP EYON Tp,E FILE PAC O TEiTY MAPS. IF THE NEW MAP 7.C IS UM3E I TYO$, SUQUTIE ECES ZS CALLEO TO HANDLE 755C TH( PS3E1. 7
757C. 758C 7
HFiLEMFLE+I 763JCIFLLE 761MFMFIE 7,2
60 LFIF+fl 7iJ£F(lFIi..E.GT.20 CA..L EACS
C 75CC 7b7
65 O1LY TWO 1S ARE COrIIED IN ONE PS T}1OLGH THE SUROUTIS SO 7oC TIi FOLI O!I4& COL)E INiTLAl.I!E CONUITIONS FOR TH Two ips 8E1G 7C CJPtdLNEO II (H PASS. 770C 771.
£F(M4p1(:1 .E.-99.A 0. P2(1i.E.-9 G3 TO 17 l2110 GO TO 3U 8i3
0 TO 180M4P3(i114P1(j)'1U0OO4M4PI) 15LF(Ni.E.0 G& TO 1,000 1.J .:1,M a17
115 1.0 LF(MADJ(j.E(.ICOOE3(L)) GQ TO 160 d13150 NSMS+1 d19
1COUE3('1)14P3(II1sP3 (I)SVil(NS) 321GO To 1
t2 160 AP3(I)5YML(L) Ui&O TO 1 Ui
170 MAP3(I)H'&Q TO 13 U6
180 AP3(I)H125 l0 CONTINUE
C LTE(C+a,1039) ICL)Ll,LROW1(AP3(1),I1,36)1009 FCv14T(L4,3A2) 831
&U TO 12J 32130 833
0O 4RITE(.3,1013j N3 331010 FUT(.19,i' 835C 836
837135 C.
AFTEP. CETZNG T4E TIT.E F)R T4IS fE MAP, THE LEGEMO IS EVELPEO.T4 COQE IS ET UP TO NGE T4 EGENO It. ORD IT1 T1E LEGEND
C OF TIE EO'J NETE ITHIN ThE F$T. CE NUT EVERYCUM3INATIO' OF Q,EN 1Y EXZST £ THE MAP, TE EE0 IS OMPAREO 8.2
C TO TtE I$T u OINATIOS CREATED FM TE 4P ELL ANO THOSEC MXSSING 4kE SKIPPED THE REAIt'.E. OF T1E ..EGENO IS WR1TTN, ON A iOiK' Fj.E, TIlE INFORMsTOI 10 ECOJE THE MAP CELLS
C CONZSTN 4ZT T4E LEGEO T0RE), ANO SIJ8ROUTINE RECOOE ISC (A.LEO.15 CC(. 5a
COCC iF THE NM39. OF MAPS Ti) 3E OOMOINEO EXCEEOS TWO THIS SU6OU1NEO 1 P6SSED I iOu N AJJ1T0Qt9A TIME F3 EsGH IJiTIOHAL MAP. THEC PEV1CUSLY ATE0 OOM3INATiJi IS TEATED A. THE FLST MAP N0 THEO WELT MAP TO E CJMOINEO IS TETE0 95 THE SECJNO MAP. THE FOLLOWINGC COOE TESTS FO A T9.MLNAIION OF THE COMBINATION POCE$S ANO WILL SETC UP CONOITIOS FO ANY OF THE AOOITICNML RUNS THROUGH THE SU3ROUTINE.00C
IF(ML2.E.NMuPI.JAl) GO TO 1+06JEJE#1MLIMFML2r41A'(JE)NL1LFNLZZML220GO TO 110
5 COMM //ISYMoL(2'.9b 951COM1ON/0/TITL(i 932COHIUN/E/KOOE 953COMMCH/F/ikDATA ARE/2t.33/,ICT/t2960/ 935
10 C 936C 337
C958
C THIS SU3ROUIINE ALLOWS A USER TO AGG.EGATE OR A5IGN THE CELLS OF 9591 EACH SYHAUL £N10 KQU?S AND 3GUMENt tHOSE GROUPS. THI3 INFORMATION 960
15 C CAM AE PROIUE'3 jTHER IMTEACTIVELY FROM THE TERMINAL OR FROM TWOC FILES. IF IHE INFORMATION IS ON Fl_ES. TAPE45 CONTAINS THE ACTUAL 962C ASSICHIENIG AMO TAPE'.ã CONTAtH THE GROUP OUCUMENTATION. IT IS 9631 £i4PUl(TANT IN OAES WHERE THE tFOHATiOt4 IS ON FILE THAT THE MAPSC OE INPUT U' A PvEVIQUS RUN AHU THIS SUBROUTINE 3E THE FIRST 935
THIS CODE £5 USED WHENEJER THE NUMAER OF AAIL4aLE MAPS 15 987C IMCREASEO (EITHEK A MEW MAP IS INPUT OR ONE IS LREATEO AY A 998C MAP OPEiATtUN. THE NUMBER OF AWAILASLE MAPS IS iNCREASED 939
C THE FILE MUM8ERS FOR THE OEM MAP ARE OESIGNATED. AND A LEK 990
145 IS MLUE 3 EE IF THE NEW MAP PUSHES THE NUMBER OF AVAILAbLE 991
C MAPS BEYUIO3 THE FILE SPACE FOP TWENTY MAPS. IF THE NEW MAP 992
C IS NUH3E9. TWEHTYONE. SUBROLTIHE EXCESS IS CALLEO TO HANDLE 933
C rHE PR3BLEI. 90+
C995
50996
C3)7
MFILE:Mt*_Et 998MF.1FI_E 939
LFMFI23 1000
55 IF(MFIE.GT.20) CALL EXCESS 1001C
1002C
1003C
100.C THE FOLLOWING CuDE CLL5 THE 3USRCUT1ME WHICH WILL REORDER THE 1005
60 C MAP SYM9OLS IN THE LEGEHO ACCOROING 13 USER PREFERENCE CR BY 1036
C LuRGESr ACREAGE FIRST. THE UEP THEN INPUTS INFORMATION 1007
C OIRECTtNG lIE ASSIGNMENT/AGGREGATiON PROCESS ANO THE NEW 1003
105 C 1031C 1052C THE F0L0W1C COQE ALLOWS THE MAP GMMENTS/AGGkECATIONS AHO 1083C THE NEW L.EENO IU 3E INPUT FROM THE 1Eh1INAL. 105'.C 1035
110 C 10561. 1037120 WRITE(80,tQ0I NC 10581009 F0RiUr(/ENTEP. EACH GROUP NUrISEiR, ITS JOCUMENIATION , 1053
1ANJ THE AS5CIATE0 P1 .TH30L3./T1E EGENO FUR EACH t, t0Q115 2GiOUP MUST E £.IZ,* LINEt5I - 40 CHARACTERS PER LINE.t/ 1061
3tTHE P.)T YM8ULS SHOULO OF PLACEC £N QUOTATION MARKS.t/ 102'.WITH EA;H SE?AiRATEO OT A COMMA OR dLAHK.// 10635*MOW MA9 0RLiU*S 00 0U WIjH TO kEA1E4*I 106.REAO(49,) NO 1035
120 C 106thJi7
00 170 1068WRITE(50, 1013) 1069
1010 FORMAI(/.IEh4IER GROUP NUHOER ANO NUiI8EiR OF ASSOCIATED STMOOLS.*) 1070125 REAO(.9,') NP,ilP 1071
C 1072WRIIEtSO, 1011) NP 1073
1011 FUkMA1(/N1E.c LEGEHO FOR GROUP ,I3) 107'.00 130 £:t,f)C 1076
C 1092C 1033C 1094C THE FOLLOWInG COOE ASSIGNS EACH CEL IN THE MAP TO THE GROUP 1095
150 C WHICH WAS PREIQQL OESIONATEO. £r ANT CELI_ STMOOLS WERE 1096C II.LSSEO IN lIE INPUT POCES5 GiRCUP ASSIGNMENTS AHO OCCUMENTATIOM 1097C WILL 3E kEQUE1E0 ANO MUST PRQJISEO IMMEOIATELY AT THE 1098C TEiRMINAL. 1099C 1100
155 C 1101C 1102180 READ JR 1011.) ICOI.,LRON, (MAP (I),I=1,3S) 11331011. FQH.iAT(201s/8A,18j'.j 1101.
izi
IF(ICO..E.99J C,) TO 25Q 1105160 C 1106
00 4Q 1 3b 1107GO TO 2'.0 1108
IF1P(Z).E.-9) G TO 2'.00 1.0 _l ( 1110165 1 IF(P().k.RP(LJ) GO 10 220 11.11
11121113
WRItEO,jQ1) SY(MAP(IJ),?1U) 111'.1t5 Fc'UT/twNLGH GROUP DOES tiE Pi.T SYMBOL ,AZ,$ ($,11.,$I, 1115
5 UCMMUN/lSY.(2.'0ó) 1230C 1281C 1282C 1283U INIC SU8kOJI1NE miLL MIOMLIGMI GEA1N ASPECTS OF A MAP'S 124
10 U QOCUMEMTAT ION. iN A LO jNATQN MAP, THE uSE CAN FIX lE 0i 1285U MUE ACTEi.T1CS WHICrI MArE UP THAI AP AND LET OE O MORE 1286U CHAAAUTErISILLS A,Y. THIS WILL PDOUUE A .EGENO OF MAP CELLS 1287C WHICH 4MVE TH FIAST SET OF CHARACTERISTICS I COMMON BUT A 1288C DIFFE,cENT SSUUNU ET. 1289
15 C 1290U 1291c 1282100 WRITE(50, 1001) 12831001 FQNlAT(/tOO IOU WISH TO SEE THE DIRECTORy TO *10 iN SELECTION : 129'.
20 1OF A M*.'4) 1295REAO'.9 1082) JA 1206
1002 FQRMAT(61) 1297IF(JA.E.1HY) ALL QIRECT 1298
C1209
25 WTcITE(50. 1003) 13001003 PG T(/WhCM MAP 00 YOU WISH TO SELECT4) 1301
J8 1302IF(J3.E.0.0) t.O TO 310 1383
C130.
30 110 NR 1305TACk0. 1336,JF.f) 1307JFT0 1318
C1309
35 EWfl10 il 1310MEWINO J3 1311
C1312
C1313
C131+
.0 U THE FOLLOWINO OUOE PROVIDES FOP. THE SLECTICN OF TiOSE 1315
C CHAPACTEkISIlUS ICM dILL 3E FIXEU AND THOSE WHICH WILL VARY. 1316C 1317
L5 2O wRIT(,O,1U2'.i1I. FO,14T(/tT1Z WEtE O CELLS IN TE O48IMATIN WITH THAT
ttFiAEC 3Cu.1aATIUM.)C i.e300 wITE(5U,1OZ5) t+a.
190 1025 FQkrIAT(ft)O Vol. WISh T lAi(E AO)&TZONAL RU F THIS OPERATIONe/ t+QS1t(3O (ii YES, FROM THE SA1E M.P (2)YS, FROM A OIFFENT iPX)RZ4(,j JZIF(JE-li 313,110,100
C15 1O RETURN i.73
11.71
WRITE(Z , 1O23 TOTAC,TAC,PTAC t4,2F T(t/t,XTOTA_ MCAGE QF T1E EMTIRE AE4 = t,F1I.i/ 1.5311A,TT4. CE.GE iF THE PMTIL AIEA :,F11.I/21,tPET OF THE EMTE EA M THE PAQ.TAL AREA z X,F5.1)&o To .3U
C THI Su3Rour,E OETECTS Alu TEN 3UPS THOSE CEL.S Wl4ILH AEC. AJJACENT Tv UE ANUTHER NUT JIIGONAU uNi) ARE OF THE SAME 1..85
15 C. SYM9OL tHAE TrIE SAME G ,GTE,Q.ISTS). IF TWO CELLS SHAIE A 1+86C COIIMC:I 5YM3La Sul A AE NOT C ONNECT ED 5Y ADJACENT CEt. S OF THE 1.37C SAMI YM3O THESE TWO C.E.LS WILl. AE PLACED INTO DIFFERENTC. CiOUPS THIS PlL)CE55 WILL 3REA< THE ENTIRE MAP INTOC. PANCELS SOlE OF WHICH liAr SHARE CUIMUN ONAACTERISTICS SuT 1.10
213 . M,E SP.TIA..Y SEPARATEO.C 1.92CC. 1.0'.
1.9525 US0 1.06
C 11.97.kITE(50,i3Q1) 11198
11331 F.pelAI(/YOQ YOU WISH TO SEE THE D1RECTORY+*) 11109RE..C.0, 1002) j$ 1583
33 1002 FOlAT(1) 1501iF(IS.E0.IHY) CALL DIRECT 1582
C 1533WRITE(50, 1003)
1003 FU+MAT(/tWHICII MAP DO VOL WISH TO GROUP INTO PARCELS4Y) 150535 REu 118,) 1506
IFUA.E0.Q) CU TO 310 150?C 1508
159REWIND I 1510REWIND i 1511..EW4J +t, 1512REWINC + 1513
C 1)1111515
'IS C 1516rli:s C3OE tS USEC. WHENE9EP THE NUNSER OF AIAILAOLE NAPS IS 1517
C. INCREMSED EITliE9. A NEW MAP IM INPUT O ONE IS CREATED SY A 1518C MAP WEPTION). IHE NUHdER OF AVuIA8LE MAPS IS LNCREASEC, 1510C THE FILE IdRS FOR THE NEW IA? A DESIGNATED, AND A CHECK 1520
50 C IS MADE TO OEE IF EHE MEW MAP PUSHES THE NU1ISER OF AVAILASLE 1521C. MAPS EYUND THE FILE SPACE FD0 TWENTY MAPS. IF THE NEW MAP 1522C I NUMAER TWENTYONE, SUSROUTINE EADESS IS GAI.LEO TO HANDLE 1523C THE PRd3LEI. 152'.C. 1525
55 C 1526C 1527
MFILEMFILE+1 1528MF:MFILE 1529LFIF+Z0 1530
60 IFUIFILE.GT.20 CALL. EACESS 1531C 1532
1533C 153'.C THE FO1.LOWI;4G CODE CIIEDI<S EACH MAP DELL AGAINST THE CELLS ASOVE 1535
65 C AND TU THE ..EFT. IF IT lATCHES EITHER ONE IT IS PLACED IN THAT 1536t, PARCEL. IF IT MMTC.HES 00TH AND THEY ARE IN DIFFERENT PARCELS, 1537C THE GUNNECT ION £ NOTED AND THE GEL IS PLACED IN ONE OF THE 1533C PARCELS. TIE PARCEL NuIIOER OF EACH CELL IS WRITTEN OUT. 1539C 15.3
70 C 15+1C 154+2100 REA3LJ+. 10311) ICOL,IROW, HAP2U),IZ,37) 15.31001. FORMAT(3I./3X,i8I.)
IFUCO...EQ.9l,9) GO TO 22075 IF4ICOL.Ci.1) GO TO 135 i'.ó
15'.?NARGINURQW)Q 15.0
C 15.9135 MAPZi)'MARCINIROW) 1553
1Z8
NAP2W:1At(I4tIR0W) 1.551C 152
00 2O L2,JT 153IFIMMP2(I.EU.QI GO TO 1?IF(MMP2(I) .EU.-98I GO TO 18Q 155
85 GO TO i 15iC 157l8
£FiR0d.EQ.1) G TJ IIQ l55IF(?l..I(L) NE.M4iP1UI) O ro 15ó0
15ó11.10 IFb2(L).NE.MAP2(I-1)) GO TO 120 1Q2
C1.20 IF(K.E2.) GO TO IJG 15o5
IFIK.EQ.l) 2(I)IP4CL(NAP1(L)1 lb6IF(.EQ.) 1567
TO 135(O 10 2 15o
I; 15fl1tOO 13 N1I1 1571
ITYPEU.)lIAPZ(I)£PACE.('flN 1573NAP2Z)G TO 20) 1575
135 NPINAPICI)P2:4AP2 (jt) 15771.O IF(X..(cP1) .EQ.NPI) TO 1'. 17
1P1r ;E_ pij 15 gGO TO 1.J
t10 1'.5 IF(XP.UP2).EQ.N2) GQ TO 15Q 151NP2PE..(N'2)G) TO 1., 153
150 LF(NP112) 155,IoO,lôS i5+1,5 NAP2(i):4P1 185
t15 £PCEPZ)NP1Q TO OO 158,
ióO UAP2(X)P1GO TO 2Q 15o9
iô t4P2(I)1P2 1t1t2l IPAE.(eP1NP2 151
TO 2OJ 15170 NAP2(I)Q 193
co To 2180 N6P2(L)- 1595
US O T 2uJ 1561O NAP(I)- 157ZOQ 158C Mi(ILW)HMP2l31)
C THE FOLLu OOE CHECKS FOt ALL. CJNCTOS P1IOUSLY QETECTEO, 1ó1+C EAUS tHE PaCE. ANL <.LTES THE MP ONTO ORIG FIi_E 1615t5 ( MNY UiETED PARCELS THE A'1E SYML. THL SETS THE MAP 1616
UP TO 3E uutXE EGUE ACCOOIt To TE OROE OF 1617C THE LEGEND. 1618C 161.C 1o20
151) 16100 23 Ii,t 162IF(XPAL(I).NE.I GO TO 3O 1623NSM$I1 12JCOE(N) 165
. THE FCJ(.LUI.LNG CODE SETS UP THE tZTEL, WRITES THE LEGENC FC EACH 1657C PARCEL. LN JOE O A wO.iNG FILE TOE3 THE INFGMA1IOM TO PROPERLY 1653L ECOGE tHE P, ANO CAL.5 SU0OutiNE ECOOE TO COMPLETE TIlE PARCELG EtIOM PCE. 1660
190 C 1661C 1o62C 1b63
EA3CIa 1007) MT,NC,MO1007 FOlAt(2O2 i'+) 1665
195 itE(50,10) 1661008 FOlt(/EfE, THE AME OF THIS MEH AP (50 CHARACTERS)J 1667
EA3(.9,10U9) CTT.EC),K1,5)1009 FQIAT(5eL0) 1669
WjtE(',7,1010J NT,MC,5,CTTLE(,Kt5J 1670200 1010 M (212, ,0PACELS - ,5.10) 1671
C 167200 320 :1,T 1673EA3(I 1311) (tITLE(KJ,I=1,ã) 167g.
5 CUUN//YMdL42.I t7bC t77C. 1.703CC THIS 3U8OJ1iE aLLOWS THE u.ER To PUATE THE TITLES AM/Q LEGES 1.71.0
lii C L.F M. THE UPOTE MAY INLUCE AN MEAE O DECREASE IM THE 1.71.1.C NUMEK OF E OF EACH MS 4CLI. AS A ME IN THEIR COHTENTS. 1.71.2
C 17LC; t7tC t7L,1.00 WTE(0,t001) i.Ttó1.001 F1AT(/U YOU 415H TO SZE IHE OIECTORe) 1.71.7
READ ('9, 1002) .S 1.71.8
1002 FOpMATRfl t7LIFtIS.E.tHfl CALL DIRECT 1.720
20 C 1721.WRITE(S0,1003) 1.722
1003 F M.1(/wH;cH 1AP 00 YOU WISH TO SELECT") 1.723kEO(.9,) . t72.lF(I.E..J) t;Q TO 350 t72
25 1726110 EiINC o 1.727
cEWL0 I 1.728C 172U t7U
30 C 1731C THE FQ.LQWIM( U0E ALLO.tS THE USEk TO SELECT AN UPOATE OPTIUN AMO 1732U ASp(S TO HME THt 4PPOPI4TE £NFTION LNPUT FRUM THE TEMI44L. 1.733C AN OUUME1ATIUN 4HH IS HOl TO jE OHA4E0 1.3 SIMPLY MOE3 INTACT j734
C 10 THE W3AItiC FILE. 173535 C 1.736
C4 1.737
C 1.738EA3UR 100.I NT,HC,N5,(TITLE(II,I=t.6) 1.739
65IF(iT.E3.1 G TO 29) 1787IF(IT.E3.21 GI) TO toO 1783WTE(50. 1012J
1012 FOrJi.T(12HuW 1ANY LEGENOS 00 YOU WISH TO UPOATE4) 173090 EO(1+8. 1791
IF(L.E.0 U TO 3..0 1732C 1733
WjTE(0,t813 178k
951013 FOiMT(/tLjT THE SYi00L 45SQGITE0 .11TH THE .EGENOS TO
IdE UPUAIEU.t/YPLuGE EGH YHoO1. QUQT2TI3N lAi.KS/17951796
2t4t43 5E...uTE THE GYMOGL5 WITH A COjIMA O 0Li.Yl 1797Eu)(9.l LSYI(,I=1,NLI 1798
CU TO 183 1799C 1800
100 160 IF(IU.E.1HN) GO TO 180 1801jF(I.1.E.1Hfl GO TO 173 1802WIrE(,0,101.l NP 1303
tOIL. F MAT(fVELEGT THE ,I21t LINE(GI OF LEGENO TO AE RETAINEO./ 180'.ItTYPE 14 TMEI POGTLU N THE CUkTNT LEGENO.t/ 1805
105 ZtEP9kATE TH NUMOE'4.3 WLTH A GQMiA.l 1806EA0(1+9,1 (iPGi(,It,N1 1807GO TO 130 1803
C 1809170 WITE(5J,13151 NA. 181.0
110 1015 FO..aiAT(ItWHEN SELECTZNG THE .IHE(S) OF LEGENO,/ 18111TYDE IN TiE POSITION AFTER E..GH GUA.A.ENT LZGENO IS LISTEO./ 181.22SEPAR.rE THE Nui1i3E.S .ITH A COMIA.tI 1313
1.70 C THE FOLLOiiNG 000E EAOS THE lAP, .00TE CELLS IN THE PAICELS TO 20.C SE ELIIINATEO, GHANI..E5 THOSE CE.L SYHSOi.S, ANO ITES THE MAP OUT 205
ON A W3pi(j'4, FILE. THE OUCUrENt4TI0i MIUS rHAT ASOCIATEO WLTH 2056C THE ELIMINATEu PACEj I. uLO T 5FEeEO TO OciNO FILE. 207C TMLS iNF3MTION IS P3VICEO TO S H.OuIINE EGOOE SO THE MAP 4r40 2053
1.75 C OUGuMENIATLON CAN SE UPO4IEO. 2053C 2050C 2061C 2062V.0 E.J(J,l0t2J 2063
18)3 1F(ICO_.E).995) GO TO 300 200L.C 20o5
00 290 1 3o 206IF(i4AP(I) .&Q.U) GO TO 270 2067IF(MAP(.EQ.-98) GO TO 260 2063
185 00 250 J1,:lPu. 2069250 IF(M(L).EQ..PA(J)I GO TO 260 2070
1AP( Z)zI51M9(MAP(I) I 2071GO TO 250 2072
260 MAP(i)isYM3L(JPAIJ1I 2073190 GO TO 20 207L.
270 ?4AP(I)2l 2075GO TO 200 2076
280 MAP(I)2l 2077290 CONIINUI 2076
195 C 2079ITE(.S,1Ot3) ICOL,ICW,MAP(I),It,36) 2010
225 C 2109C THE FOLLOWI'G 00E AL.OW5 THE USE TO IP.]ICATE THE PMRCEL TO 9E 2110C SPLIT, THE NUNSEI% OF NEW PACEL )A4TE0, 5180 THE OI,ECTOt. OF TIE 2111.C iPLIT. TIlE MA? £5 THEN .EAO IN TO FINO THE LOCATION OF EACH CELL 2112O IN THE PIIEL. 2113
230 C 211'.0' " 2115C 2116370 W.ITE(5),1.022) 21171022 FC)14T(/tWMLCr$ PAiCEP. 00 YOU WANT TO OIVIGE, INTO HC t, 2118
235 ItMANY PIECES, 4180 IN WHAT OIECTICN.t// 21192tLIST THE PAHCE SVM3OL IN UUTATIJN MARKS, THE NUM0ER OF PIECES,t 21203/*ANO TIE OLREGTION OF 014i51018 (1) NorHsoUrH (2) EAST-WEST.t/ 2121
133
5EPAkAT1 EACH OF THESE WITH A COMMA O.R 3LAMR.) 2122RE..O('.9,') ISYM,NPC,IOIR 2123
V.0 00 3C0 J1,126 212.380 1F(1SM.EQ.I.VML(J)) GO TO 390 2125
C 2176C 2177C THE F0LLOIhG CO,,E USES THE CELL LOCATON PREIOUSL! COLLECTEO TO 2t78
295 C CALCULATE W-IErE T.iE PARCE.. SPLiTS '.I, SE MOE. I THEN REOOS THE 2179C MM? ANO E 5i4 THE CE.LS OF THE £NOIGMTEO PARCEL U5ING THE 2180C CALCULT1ON OF THE 5PL1T. 2181C
2182C
2183300 c 218'.
1=1 218500 520 3:2,250 2136L0C(J)..0 (j)+LOC41) 2187IF(Luc(J).Gr.a...Nu.ocJ).Eo.LoC(11) GO TO 530 2138
305 520 1:j 219C 2193530 h0I.O0(J)/NC 219k
REWINO J 2132C 2133
3L0 5'.0 REAOIJR, 1012) ICOL,IROW, (MAP(I) .1=1,36) 219'.IF(ICOL.E.3.9999) GO TO 20 2135
C 213600 610 Ii,36 2197IF(MAP(I).EO.0) GO TO 560 2198
315 ;F(MAP(I).Eu.-999) GO TO 570IF4MAPIL).EQ.'(PAR) GO TO 583 2200
134
5,0 MAP(I)IYML(MAP(LJJ 2201CO TO 611 2202
560 MAP(1)24 2203320 GO TO DL0
570 HAP(I)24" 2205GO TO otU 2206
530 LF(IOik.EO.2) G(i TO 590 2207JUCO.-ti'3b4I 2203
325 L011 O(J)/N)Il 2200GO TO 600 2210
590 LOLOII'JW)/NCI 2211600 IF(.Ji.E'..0) u TO 550 2212
2213330 1FI00.GT.) NaNS 221'.
M4Pt)lJYi8L(NQ 2215010 CONIZNOI 2216C 2217
WRITE('+6,1013) IGOL,IROM,tMAP(1),I1,36) 2218335 GO fo 5.9 2210
C2220
b20 W.ITE(*ô,101'.i iS 222100 ó30 I1.,250 2222
30 LOC(I):0 22233i.l) C
222'.2225
C2226
C THE SU0i3ufINE RECODE IS ALLEO FCR LL THREE .PERATIONS. IT 2227( CHANGES AC..E LAUSEO 8! 101 iNc. FOM PAEL TO PARCEL, 222
3'.5 C ELIM1NATE 00 .fATON FOR THOSE PARCELS AGOREC.ATED WiTH OTHERS, 2220C ANO EAPANOS uHENTATON FOP. THSE PMRCELS CREATED O'f A SPLIT. 2230C
2231.
C 2232C 2233
350 '.0 .ElNO .5 223'.
CALL REOOE 2235C
2236jRs1F 2237
2238355 C 2239
WRITECSO,1.02., 22'+0
1.024 FORMAT(YOO 'fOLj WISH TO PERFORM ANY MORE OPERATIONS X, 22'.1
lION THE ACELS OF THIS 22'.2
REAO('.0 1002) IS 22.3360 IF(IS,E.1e1'f) GO TO 103 224'.
C 22.5oSU RETURN
END 22'.?
135
OELAY(l,11,0) 1385
PNOGRArI lut4LP 1885OiIENSIOI8 1837
1888
5 COMM /9#MF_E,HF,LF 1839C3/3/IE(t23S,45,NC,NT 1390
COON/O/TIT-E(5) 1892OATA L3;/uao/ 1893
10 C139'.
C1395
C1896
C THIS su3oJrIiiE A...OWS THE USER TO MANIPULATE MAP CELLS AuT DOES 137C NOT ALLOW CONTENT GHACES TO TiE LECEN4O. THERE A,cE rHREE TYPES 18
15 C OF oPRArI3 PRUAIOEDI ZNOIIOUAL CEL.S CAN E lOVED FROM ONE 1393C PARCEL TO A9UTHER, CELLS OF ONE PAW,CEL CN dE ACOREOATEO WITH 1930
C THOSE OF A'4JTriE, ANU T1E CELLS IN ONE PARCEL CAN SE SPLIT INTO 1901
C TWO Ui MORE PARCE. 19(12
C1933
20 CC
1905WRIIE(50,1A01. 1906
1001 FO-.IAT(/tOU YOU WISrI TO SEE THE OIRECTORY4A( 1907
RE43(.9,1002( IS 1908
25 1002 FOMATU1( 1.909
IF(tS.E9.INY( CMLI. DIRECT 1910
C1911
WRITE(S0, 1.0031 191.2
1003 FOR1At(/WHIH MAP 30 YOU WISH TO SELECT4Y1 191.3
30 EA('.9 '1 IR 1.31.'.
jF(jR.E1.Ui GO TO b50 191.3
C131.5
C1917
C1313
35 C THIS CCOE IS uSED WHENEE THE NUMSER OF AVAILASLE MAPS IS 1919
C INCREA3EO (EITHE A NEW MAP IS INUT O OnE 13 CREATEO SY A 1920
C MAP OPENATJNl. TiE nIUMSER OF AILA3LE MAPS E.G INCREASED, 1321C THE FtLE MUM8E.S FOR THE NEW MAP aRC DESII.UATED, aND A CHECK 1922
C IS MACE TE SEE IF THE NEW MAP PUE3 THE n4UMSER CF AVAILASLE 1923'.0 C MAPS 3EYO) T4E FILE SPACE FOR TWENTY MAPS. IF THE NEW MAP 192'.
C 13 NUM3E TWEr.TYOME, SUSROLTI9E EXCESS IS CALZO TO HANDLE 1925
o THE PROSLEI. 1325
C1.927
C1923
'.5 C1929
MFLLEMFLLEG1 1930
MFMFI_ 1931.
LFMFG?J 1932IF(MFILE.CT.201 CALL EXCESS 13
50 C193a1935
C1935
WRITE(,0, 10051 1937
1005 FORMAT(/EiTE* THE NAME OF THE NEW MAP ('.5 CHAHACTERSI1 1938
55 (TITLE(,K1,Sl 10391006 FUrciAT(5101 19.3
C
100 REWINO60 REWZND J( 19..
REWIND 13.5EWINO .7 19'+ó
c65 REA3(IR,10071 r.T,NC,MS 13'.9
1007 FQIAT(212,IWI 1950c
1951WRITE(50,1004)
1.332
1000 FORIAT(#WHiCr4 OPERATID14 DO IOU WISH TO USEet/ 195370 lt(ti MOVE UOI4LOUAL ELLS FROM ONE PARCEL TO ANQTHER/ 195'.
2(21 AUGECTc QME PARCELS WITS OTHEW.SA/ 1955
3(3l OIZOE A CINCLE PAROE INTO TWO OR MORE PMRCELS) 19,6
REA3('.9,'l 101957
IF(1.O-21 110,220,370 1.958
75 C195919o0
C1961
ç THE FOLLOWING COOE ALLOWS THE oSER TO LOCATE INOIVIOUAL CELLS IN A 1962C MAP THEN IUMATE THE SYMOOLS FUH. THOSE CELS. 1963
135
SO C 19&.C 19,5C 1966110 WSITE(,Q,1039) 1967lOGY FCriAT(/ti1.jW MMNY It4QLVIDUAL CE..5 DO OU WANT TO 3E 19o8
55 ItMOVEO F.OM PARCEL TO PAACE.et) 1969RE..D('.Y,) ELL 1.973
C 1971WRIIE(53,jQlU) 1972
1010 FURMAT(/tL5T THE COUlN NU$BEF ROW MUM9EP., LOCATION t, 197390 hIM THE ç0W $/tANO THE MEW PAkC. YM50L (IN UuOTATIQN MAR5) t,
2tFO EO1 cL. 6EING MOVEO./ 13753t$EPAcAfE EAh OF 1HEE WITH A COMMA OR .t) 19l00 130 £1,ItE4j. 1977cEAD U.s. ) J..J.( II ,JAOI4( II ,JSPOT( I) ,JS(M (I) 1975
100 CC THE FOLLOWING CODE REAOS THE MAP, ..00A1E3 THE INOICATED CELLS, 1935C CHANGES THDSE OEL. 5ya9J...5, ANO WRITES THE MAP OUT ON A WC(ING 196O FL.E. THE JouN64r4TijN IS A.GO 1M.3FER9.EO TO A WoW(ING FILE 1987C AND INF0RMArIOM IS PRUVIOEJ To ROUTINE RECQOE jO CHANE5 IN 1938
105 0 THE AC.EA)E FOR EAGH SYM8JL CAN SE JPOATEO. 1989C 1*90C. 1991C 19321.'.Q READ(JR 1012) LGOC,I.OW,(MAP(I),t.1,J6) 1993
113 1012 FOMMT(2Jj,/3A,L5I'+) 1994.IF(IOO.U.998) GO TO 190 1995
C 199600 150 1937
150 iF(JC(J).E4.O.ANO.JO(J1EQ.iROW) MAP(JSPOT(JI).JCELL(JI 1398115 C 1399
DO 180 t1 30 2000GO TO 160 2001
F(lAPIZ) .U.-999) 00 TO 170 e 2092MAI(L)I3YI1dL(ItA?(L) ) 2003
120 oo To too 230.160 MAP(1I.2l 2005
GO TO 139 2006170 MAP(j)r?'l' 2007180 CONTINUE 2003
125 C 2019WRITE(I.6,10131 ICOL,IRCW,(MAP(hi,I1.J6) 2010
IS COMlON/N/I5YMBL(12g6),AS(1296),tJ5$(20,'.j,IT(1296J,IACR( 1296) ,PCACR(1296J ,00C(1296 ,'.e
10II
C 12WRIIE(l0,1001I 131001 FORMAT(/ICOMFLICT DETECTION ROLTINE/1X,CREATEO BY R. )IAGESTEDT/ 1'.
15 18,OEC. 1978e 15C 16
WRIJE(l0, 1002) 171002 FORIAI(/IIHE FOLLOWING FILE NUMBERS ARE SET./ 13
tITHE USER MUST BE CONSISTENT WITI THEIR OESIGNATION./ 1920 23A,151M331_ LIST CIJ$13X,IACTIVITY ASSIGNMENTS t2)$/ 20
33,1AC1LwITY NAME LIST (3)I/3X,ACIIVIIY SPILLOVER MATRIX (kIll 21'.3X,IADJACENT MANAGEMENT UNITS (5)X/ 2253X,IMANAGEMENT UNIT MAP DOCUMENTATION (6)1/ 2363X,IMINIIUM MANAGEMENT UNIT ACREGES (7)1) 2'.
25 c 25REWIND I 26REwIND 2 27REWIND 3 23REWIND 29
30 REWIND 0 30REWIND 5 31REWIND 7 32
C 33WRITE(1IJ,117031 3'.
35 003 FORMAT(1IOU YOU WISH TO INPUT ACTIVITY NAMES, MINIMUM ACREAGES,$/ 35IIOR SPILLOVERS ONTO A FILEeY) 36READ(9,10IJ'.) IA
100'. FORIAT(A1I 38IF(LA.E.iNY) CALL IN 39
'.0 C '.0REAOU,1005) (ISYMBL(I),It,1296) '.11005 FORIAT(IOAZI '.2C '.3
25 1006 FORMAT('.AlO) 198100 WRLTE(.3, 1007) I, (USE(J),J21.l.) 199
1007 FORNAT(C,,kAtOl 200
REWIND 3 201
C 202
30 110 WRITE(tO,1008) 2031008 FO.NAT(YOO YOU WISH TO INPUT MINIMUM MANAGEMENT UNIT A, 20'.
IACREAGESe) 205REAO(9 1002) IA 206IF(IA.U.tHN) GO TO 130 207
35 C208
WRITE(7 1005) NLU 209WR1TE(ti. 1.009) 210
10119 FORMATI/AENTER THE IINIMUI ACREAGE OF MANAGEMENT UNITS AFTER Al 211IAEACN ATIVITY NUMBER IS LISTEO.A) 212
1,0 00 120 L:1,NLU 213WRITEIIO iooi x 21'.
REuGlO,') MIi4ACR 215ACR$INFLOAT(M1NACR) 216
120 WRITE(7 1010) ACRHIN 217
1.5 1010 FORMAT(Fti.t) 218
REWIND 7 219C
220
130 WRITE(10,t011) 221
1011 FORMAT(AO0 YOU WISH TO INPUT LANDUSE SPILLOVERSeA) 222
50 REb.O(9 1002) IA 223IF(LA.h.IHM) GO TO 170 22'.
WRITE('.,1005) NLU 225
C225
WRITE(10,1012) 227
55 1012 FORNAT(tENTER EACH SPILLOVER BY LISTING THE TWO ACTIVITY A, 2281ANUMBERA/AFOLLOWEO BY THE TYPE (.,) AND LEVEL OF THE SPILLOEEi.A 2292,'ASEPARATE EACH OF THE NUMBERS WITH A COMMA.A/TERMINATE THE A, 2303:ENTRY EOUENCE WITH A SERIES OF ZEROS.A) 231