EPA-450/3-76-012-C
September 1976
GROWTH EFFECTS OF MAJOR
LAND USE PROJECTS:
VOLUME III - SUMMARY
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
-------
EPA-450/3-76-012-C
GROWTH EFFECTS OF MAJOR
LAND USE PROJECTS:
VOLUME III - SUMMARY
In-
Frank Benesh, Peter Guldhcrg, and Ralph D'Agostino
Walden Research Division of Abcor
201 Vassar Street
Cambridge. Massachusetts 02139
Contract No. 68-02-2076
EPA Project Officer: Thomas McCurdy
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
September 1976
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations - in limited quantities - from the
Library Services Office (MD35) , Research Triangle Park , North Carolina
27711; or, for a fee, from the National Technical Information Service,
5285 Port Royal Road, Springfield, Virginia 22161.
This report was furnished to the Environmental Protection Agency by
Walden Research Division of Abcor, Cambridge, Massachusetts 02139,
in fulfillment of Contract No. 68-02-2076. The contents of this report
are1 reproduced herein as received from Walden Research Division of
Abcor. The opinions, findings, and conclusions expressed are those
of the authors and not necessarily those of the Environmental Protection
Agency. Mention of company or product names is not to be considered
as an endorsement by the Environmental Protection Agency.
Publication Nb. EPA-450/3-76-012-C
ii
-------
ACKNOWLEDGEMENTS
Special appreciation goes to Mr. Thomas McCurdy, the U.S. Environmental
Protection Agency project officer for this study, whose extensive assistance
and advice was indespensible in the execution of this study.
In addition, we wish to express our appreciation to the cooperation of
the more than one hundred individuals who were contacted and provided infor-
mation during the data collection phase of this project.
STAFFING
Mr. Frank Benesh was the project manager of this study at Walden Research.
Dr. Ralph D'Agostino, contributed to the overall study design and analysis.
Mr. Peter Guldberg conducted much of the causal analysis, calibration, and
cross-validation of the land use model, fir. Kenneth Wiltsee contributed
the chapters explaining the motor vehicle emissions calculations and Mr.
Mahesh Shah assisted in the data collection. The data collection was
coordinated by Mrs. Allison B. Goodsell and the manuscript was prepared under
the direction of Ms. Gail Kelleher.
Metcalf and Eddy of Boston were subcontractors to Walden Research, assis-
ting in model specification, sample selection* and data collection. Mr.
Richard Ball, Mrs. Elizabeth. Levin, Mr. Stephen Koop, Mrs. Nancy Lundgren,
and Mr. Gerald Takano contributed to Metcalf and Eddy's effort. Mr. Gilbert
Nelson conducted much of the development of the traffic model.
m
-------
TABLE OF CONTENTS
Section Title Page
I INTRODUCTION, ORGANIZATION OF REPORT, AND SUMMARY . 1-1
A. Introduction 1-1
B. Organization of Volume III 1-2
C. Summary 1-3
: 1. Phase 1 . 1-3
2. Phase 2 1-5
3. Phase 3 1-5
4. Phase 4 1-7
5. Phases 5 and 6 1-10
II LAND USE MODEL SPECIFICATION AND CAUSAL ANALYSIS . 2-1
A. General Approach 2-1
1. Theory of Induced Development 2-1
2. Selection of General Approach 2-1
3. Statement of Fundamental Model 2-2
B. Land Use Model Specification 2-3
1. Objective 2-3
2. Definition of Major Project 2-4
3. Model Specification Methodology . 2-4
4. Model Description 2-5
C. Sample Selection 2-10
1. Purpose and Initial Criteria 2-10
D. Data Collection 2-10
E. Causal Analysis 2-12
1. General Approach 2-12
2. Data Transformations 2-12
3. Theory Trimming . 2-19
III DEVELOPMENT OF LAND USE MODEL PREDICTIVE EQUATIONS 3-1
A. Approach 3-1
B. New Variable Definitions 3-6
C. Discussion of Results 3-13
D. Summary of Predictive Equations 3-19
1. English Units 3-19
2. Metric Units 3-21
E. Disaggregation and Aggregation of Land Use
Categories 3-22
F. Cross-Validation Analysis 3-29
IV
-------
TABLE OF CONTENTS (CONTINUED)
Secti on Title Page
IV LAND USE BASED EMISSION FACTORS ... ....... 4-1
A. Approach to Land Use Based Emission Factors . . 4-1
1. Emission Factor Structure ......... 4-2
2. Variance of Energy Requirements,
Efficiency, and Emission Factors ..... 4-4
B. Compilation of Land Use Based Emission Factors 4-8
V THE TRAFFIC MODEL . .< ...... . ........ 5-1
A. Introduction ................. 5-1
B. Summary .................... 5-2
C. Principal Elements of the VMT Model ...... 5-4
1. Introduction ............... 5-4
2. Vehicle Trip Generation Rates ....... 5-4
3. Trip Lengths ............... 5-8
4. Duplicated Trips ...... . ...... 5-12
5. Speed Ranges ....... ^ ....... 5-13
6. Vehicle Class ............... 5-18
VI MOTOR VEHICLE EMISSION FACTORS ....... '. . . 6-1
A. Description of Emission Factor Components . . . 6-3
1. Mean Emission Factor (Cipn) ........ 6-3
2. Weighted Annual Travel (mjn) ..... • • 6-4
3. Speed Correction Factor (Vips) .' ..... 6-4
4. Ambient Temperature Correction Factor
5. Operating Temperature Correction Factor
(riptwx) ................ • 6"5
6. Crankcase Hydrocarbon Emission Factor (fj) 6-5
7. Evaporative Hydrocarbon Emissions (ei ) . . 6-6
VII COMPUTATION WORKSHEETS AND INSTRUCTIONS ...... 7-1
A. Land Use Model ........ ........ 7-1
1. Residential Land Use Model ........ 7-3
2. Industrial/Office Land Use Model ..... 7-4
3. Computing Confidence Intervals ...... 7-5
B, Calculation of Vehicular Traffic ...... . 7-11
1. Worksheet No. 1 (VMT-1) .... ...... 7-11
2. Worksheet No. 2 (VMT-2) ..... ..... 7-12
3. Worksheet No. 3 (VMT-3) .......... 7-12
4, Worksheet No. 4 (VMT-4) .......... 7-13
V:
V
-------
TABLE OF CONTENTS (CONTINUED)
Section Title Page
C. Calculation of Motor Vehicle Emissions .... 7-14
1. Motor Vehicle Emissions Worksheet No. 1
(VEM-1) 7-14
2. Worksheet No. 2 (VEM-2) 7-16
D. Calculation of Emissions 7-18
1. Worksheet No. 1 (EMI-1) 7-18
2. Worksheet No. 2 (EMI-2) 7-18
3. Worksheet No. 3 (EMI-3) 7-19
VIII EXAMPLES OF USE OF COMPUTATION WORKSHEETS .... 8-1
A. Example of Land Use Model Calculations .... 8-1
B. Example of Traffic Model Calculations .... 8-12
1. Worksheet VMT-1 8-12
2. Worksheet VMT-2 8-12
3. Worksheets VMT-3 and VMT-4 8-13
C. Example of Calculation of Motor Vehicle
Emission Factors 8-18
1. Automobiles 8-18
2. Light-Duty Trucks 8-20
3. Heavy-Duty Gasoline Vehicles (HDG) .... 8-21
4. Heavy-Duty Diesel (HDD) 8-21
5. Calculation of Composite Emission Rate . . 8-22
D. Calculation of Emissions 8-28
1. Worksheet No. EMI-1 8-28
2. Worksheet No. EMI-2 8-28
3. Worksheet No. EMI-3 8-29
IX REFERENCES 9-1
APPENDIX A - STATISTICAL OUTPUT FOR THE PREDICTIVE
LAND USE EQUATIONS A-l
APPENDIX B - GRAPHS OF ACTUAL VERSUS PREDICTED
LAND USE FOR THE CROSS-VALIDATION
ANALYSIS B-l
vi
-------
LIST OF TABLES
Table Title Page
2-1 MODEL VARIABLES AND DEFINITIONS . 2-9
2-2 ORIGINAL SPECIFICATION OF THE RESIDENTIAL MODEL 2-11
2-3 ORIGINAL SPECIFICATION OF THE INDUSTRIAL-OFFICE MODEL . . 2-13
3-1 INSTRUMENTAL VARIABLES USED IN THE FINAL PATH ANALYSIS . 3-3
3-2 EXOGENOUS VARIABLES INCLUDED IN THE ORIGINAL MODEL SPECI-
FICATIONS BUT TRIMMED PRIOR TO THE FINAL PATH ANALYSIS . 3-4
3-3 SUMMARY STATISTICS OF PREDICTIVE EQUATIONS 3-15
3-4 COMPARISON OF SUMMARY STATISTICS FOR PREDICTIVE EQUATIONS
IN WHICH LAND USE OF THE MAJOR PROJECT IS INCLUDED OR
EXCLUDED FROM THE DEPENDENT VARIABLE . . 3-18
3-5 AVERAGE PERCENTAGE VALUES FOR DISAGGREGATED LAND USE
VARIABLES . 3-26
3-6 AGGREGATED TOTAL LAND USE PREDICTIVE EQUATIONS AND
SUMMARY STATISTICS 3-28
3-7 COEFFICIENTS OF VALIDITY BETWEEN ACTUAL AND PREDICTED
LAND USE FROM THE CROSS-VALIDATION ANALYSIS OF ALL LAND
USE PREDICTIVE EQUATIONS 3-31
4-1 TYPICAL EMISSION FACTORS FOR ELECTRIC UTILITIES 4-9
4-2 SINGLE FAMILY, RESIDENTIAL, LAND USE BASED EMISSION
FACTORS 4-10
4-3 MOBILE HOME, RESIDENTIAL, LAND USE BASED EMISSION
FACTORS 4-11
4-4 LOW RISE MULTI-FAMILY, RESIDENTIAL, LAND USE BASED
EMISSION FACTORS 4-12
4-5 HIGH-RISE MULTI-FAMILY, RESIDENTIAL, LAND USE BASED
EMISSION FACTORS 4-13
4-6 RETAIL ESTABLISHMENTS, WAREHOUSES, WHOLESALING ESTABLISH-
MENTS, LAND USE BASED EMISSION FACTORS . 4-14
4-7 OFFICE BUILDING LAND USE BASED EMISSION FACTORS 4-15
4-8 NON-HOUSEKEEPING*, RESIDENTIAL, LAND USE BASED EMISSION
FACTORS 4-16
4-9 HOSPITAL LAND USE BASED EMISSION FACTORS 4-17
4-10 CULTURAL BUILDING LAND USE BASED EMISSION FACTORS .... 4-18
-------
LIST OF TABLES
Table Title Page
4-11 CHURCH BUILDING LAND USE BASED EMISSION FACTORS .... 4-19
4-12 SCHOOL BUILDING LAND USE BASED EMISSION FACTORS .... 4-20
4-13 ESTIMATED NATIONAL INDUSTRIAL LAND USE BASED EMISSION
FACTORS BY TWO DIGIT 1967 STANDARD INDUSTRIAL CLASSIFICA-
TION CODE 4-21
5-1 EFFECT OF CAR OWNERSHIP ON AVERAGE 'NUMBER OF TRIPS PER
HOUSEHOLD BY TRIP PURPOSE, CINCINNATI URBANIZED AREA . . 5-6
5-2 DEFAULT VEHICLE TRIP GENERATION RATIOS FOR VARIOUS LAND
USE CATEGORIES 5-7
5-3 AVERAGE TRIP DISTANCES AND AUTOMOBILE TRAVEL BY RESIDENCE
LOCATION - 1968 . . . 5-10
7-1 SUMMARY OF COMPUTATION WORKSHEETS 7-2
7-2 LIST OF DEPENDENT VARIABLE NAMES FOR EACH FINAL PROJECTED
LAND USE CATEGORY 7-7
7-3 LIST OF PREDICTOR VARIABLE NAMES FOR THE RESIDENTIAL
MODEL 7-8
7-4 LIST OF PREDICTOR VARIABLE NAMES FOR THE INDUSTRIAL/
OFFICE MODEL 7-9
vm
-------
LIST OF FIGURES
Figure Title Page
1-1 FLOW DIAGRAM OF GEMLUP TECHNICAL EFFORT 1-4
2-1 FINAL PATH ANALYSIS FOR THE RESIDENTIAL MODEL . . . 2-20
-2-2 FINAL PATH ANALYSIS FOR THE INDUSTRIAL-OFFICE MODEL 2-21
4-1 NORMAL SEASONAL HEATING DEGREE DAYS (BASE 65°F)
i 1941-1970 4-5
4-2' NORMAL SEASONAL COOLING DEGREE DAYS (BASE 65°F)
1941-1970 4-6
4-3 ANNUAL AIR CONDITIONER COMPRESSOR-OPERATING HOURS
FOR HOMES THAT ARE NOT NATURALLY VENTILATED .... 4-7
5-1 EXISTING ANALYSIS RINGS, WASHINGTON METROPOLITAN
REGION 5-11
5-2 PLOT OF AVERAGE TRIP LENGTH FREQUENCY DISTRIBUTION
BY TRIP TYPE 5-14
5-3 THEORETICAL RELATIVE USE OF LOCAL STREETS, ARTERIAL
STREETS, AND EXPRESSWAYS WITH A TWO MILE RAMP
SPACING 5-15
5-4 THEORETICAL RELATIVE USE OF LOCAL STREETS, ARTERIAL
STREETS, AND EXPRESSWAYS WITH A FOUR MILE RAMP
SPACING . 5-17
5-5 RELATIONSHIPS BETWEEN V/C RATIO AND OPERATING
SPEED, IN ONE DIRECTION OF TRAVEL, ON FREEWAYS
AND EXPRESSWAYS, UNDER UNINTERRUPTED FLOW
CONDITIONS 5-19
5-6 TYPICAL RELATIONSHIPS BETWEEN V/C RATIO AND AVERAGE
OVERALL TRAVEL SPEED, IN ONE DIRECTION OF TRAVEL,
ON URBAN AND SUBURBAN ARTERIAL STREETS 5-20
IX
-------
LIST OF EXHIBITS
Exhibit Title Page
7-1 WORKSHEET RLUM-1 .'.,.'.. . . 7-20
7-2 WORKSHEET RLUM-2 7-21
7-3 WORKSHEET RLUM-3 7-22
7-4 WORKSHEET RLUM-4 7-23
7-5 WORKSHEET RLUM-5 . 7-24
7-6 WORKSHEET RLUM-6 7-25
7-7 WORKSHEET RLUM-7 7-26
7-8 WORKSHEET IOLUM-1 . . . , , 7-27
7-9 . WORKSHEET IOLUM-2, 7-28
7-10 .WORKSHEET IOLUM-3 7-29
7-11 WORKSHEET IQLUM-4 7-30
7-12 WORKSHEET IOLUM-5 7-31
7-13 . WORKSHEET. 10 LUM-6 7-32
7-14 r WORKSHEET IOLUM-7 7-33
7-15 :.. WORKSHEET LUM-1 7-34'
7-16 , ,. WORKSHEET LUM-2 7-35;.
7-17 WORKSHEET LUM-3 7-36
7-18 WORKSHEET VMT-1 7-37
7-19 WORKSHEET VMT-2 7-38
7-20 WORKSHEET VMT-3 7-39
7-21 WORKSHEET VMT-4 7-40
7-22 WORKSHEET VEM-1 7-41
7-23 WORKSHEET VEM-2 7-42
7-24 WORKSHEET EMI-1 7-43
7-25 WORKSHEET EMI-2 . 7-44
7-26 WORKSHEET EMI-3 7-45
8-1 WORKSHEET RLUM-1 . 8-2
8-2 WORKSHEET RLUM.-2 .'.'.' 8-3
8-3 WORKSHEET RLUM-3 8-4
-------
LIST OF EXHIBITS
Exhibits Title Page
8-4 WORKSHEET RLUM-4 ... ., . ,. . .. . . 8-5
8-5 WORKSHEET RLUM-5 . .. , , .. . , . ., .. , , .. . , . . . .8-6
8-6 WORKSHEET RLUM-6 ...... ..',.,.-.. 8-7
8-7 WORKSHEET RLUM-7 8-8
8-8 WORKSHEET LUM-1 ..,...., .. , , ...... 8-9
8-9 WORKSHEET LUM-2 .,'., , .. . ., . ... . . . .8-10
8-10 WORKSHEET LUM-3 . .......... 8-11
8-11 WORKSHEET VMT-1 8-14
8-12 WORKSHEET VMT-2 ... 8-15
8-13 WORKSHEET VMT-3 8-16
8-14 WORKSHEET VMT-4 . .. . . . . .. , . ., .... , . 8-17
8-15 WORKSHEET VEM-1 8-23
8-16 WORKSHEET VEM-1 ................ . . . .. 8-24
8-17 WORKSHEET VEM-1 8-25
8-18 WORKSHEET VEM-1 . . . 8-26
8-19 WORKSHEET VEM-2 ....... 8-27
8-20 WORKSHEET EMI-1 .'.....' 8-30
8-21 WORKSHEET EMI-1 ....... 8-31
8-22 WORKSHEET EMI-2 8-32
8-23 WORKSHEET EMI-3 . . . 8-33
XT
-------
I. INTRODUCTION, ORGANIZATION OF REPORT, AND SUMMARY
A. INTRODUCTION
This report documents the results of a study of the Growth Effects
of Major Land Use Projects (GEMLUP). The principal objectives of the GEMLUP
study were to formulate a methodology to predict air pollutant emissions
frjom:
• Two types of major land use developments: • large concen-
trations of employment such as office or industrial
parks, and large residential developments,
• Land development that is induced by the two types of
major land use development projects,
• Motor vehicular traffic associated with both the major
project and induced development.
GEMLUP relates to a number of EPA programs, including air quality
maintenance plan (AQMP) development [1], environmental impact statement (EIS)
review [2], the indefinitely suspended portions of indirect source review
[3], and the prevention of significant air quality deterioration, or nonde-
gradation [4]. Explicit or implicit in these programs in an evaluation of
air quality impacts of land use plans or project developments. GEMLUP is
designed to formulate and test a method of evaluating land use impacts at the
project scale, and, in the process, develop a set of land use based emission
factors potentially useful at the regional scale.
The study was divided into six phases:
Phase 1 - Specification of a preliminary model and generation
of a list of data requirements,
Phase 2 - Data collection,
Phase 3 - Causal analysis of the land use model using path
analysis,
Phase 4 - Development of predictive equations for the land
use node! and development of a traffic model,
Phase 5 - Development of indices of fuel consumption,
Phase 6 - Translation of fuel consumption indices into land
use based emission factors.
1-1
-------
The first three phases of the study (i.e., model specification,
sample selection, data collection, and causal analysis of the land use model)
are documented in the first volume of this report, Growth Effects of Major
Land Use Projects, Volume I: Specification and Causal Analysis of Model [5].
Two .of the appendices to Volume I (C and D) were published separately [6].
Appendix C contains listings of the data files and simple correlation mat-
rices. Appendix D contains the computer output of the statistical applica-
tion packages used in the path analysis of the final causal model. The
fifth and sixth phases of the study are documented in the second volume of
this report, Growth Effects of Major Land Use Projects, Volume II: Compila-
tion of Land Use Based Emission Factors [7].
This final volume of the report, Volume III, summarizes Volumes I
and II, documents the fourth phase of the study (i.e., development of the
predictive equations of the land >use model and development of the traffic
model), and serves as a guideline document for the application of the models
developed to the task of predicting land use and emissions associated with
major land use projects.
B. ORGANIZATION OF VOLUME III
The remainder of this introductory chapter provides an overall
summary of the GEMLUP study. Chapters II and III are devoted to the land
use model; the first summarizing the model specification, sample selection,
data collection, and causal analysis while the latter documents the calibra-
tion and validation of the model (i.e., the translation of the causal model
into a predictive model. Chapter IV summarizes the land use based emission
factors and indices of fuel consumption on which they are based. Chapters
V and VI are devoted, respectively to the development of the traffic model
and the estimation of motor vehicular emissions. The entire GEMLUP methodo-
logy is codified in Chapter VII as a set of computation worksheets and
instruction for their use. Finally, Chapter VIII provides an example of
these guideline procedures.
1-2
-------
There are two appendices to this volume. The first, Appendix A,
provides the data necessary to develop confidence intervals for the predic-
tions of the land use model while the second, Appendix B, further documents
the cross-validation analysis discussed in Chapter III.
C. SUMMARY
As discussed previously, the GEMLUP study was divided into six
phases, each of which is summarized briefly below. A schematic flow diagram
of the technical effort is shown in Figure 1-1.
1. Phase 1
The first phase of the study consisted of developing the pre-
liminary hypothesis of induced land use development. This was an elabora-
tion of the following theory:
Constructing a large source of employment like an
industrial/office complex generates jobs which
result in the nearby construction of dwelling units;
these induce retail development to locate near them
and generate demand for community, cultural, and
religious facilities (schools, recreation areas,
libraries, churches, theaters, fire and police
stations, etc.). All of this requires the construc-
tion of streets and highways that then improve
accessibility to the area. Better access fosters
continued urban development, particularly highway-
oriented commercial and office land uses. Addi-
tional sources of employment come into the area
as secondary (and tertiary) industry or services
locate near the original major project, spurring on
another round of residential development, and so
forth.
Concurrently, a determination was made of the input requirements of the
traffic model, the estimation of emissions with land use based emission
factors, and the availability of data. With these three elements of infor-
mation, the preliminary hypothesis of induced land use were translated into
a specification of a land use model. The model was specified in two separ-
ate forms to represent induced land use growth associated with large
1-3
-------
I
-pi
ACTIVITY
FACTOR
REVIEW
DEVELOPMENT
PRELIMINARY
HYPOTHESIS
TRAFFIC
MODEL
REVIEW
\f \f
GENERATE
LIST OF
VARIABLES
GENERATE
LIST OF
POTENTIAL
CASES
X
^
!
CASE STUDY
SELECTION
\
f
PRELIMINARY
CONTACTS,
TRAINING
CASES
DEVELOP
ACTIVITY
FACTORS
GENERATE
EMISSION
FACTORS
EMISSION
FACTOR
REPORT
(vol. II)
\
^
->
CAUSAL
ANALYSIS
PATH
ANALYSIS
REPORT
(vol. I)
TRAFFIC
MODEL
PREDICTIVE
EQUATIONS
SUMMARY
REPORT
(vol.Ill)
DATA
COLLECTION
Figure T-l
Flow Diagram of GEMLUP Technical Effort
-------
residential developments, and large Industrial or Office parks in the fol-
lowing 12 land use categories:
Residential Hotels/Motels
Commercial Hospitals
Office Cultural
Manufacturing Churches
Highways Education
Wholesale/Warehouse Recreation
(
( •
The'models predict the land use in a 10,000 acre acrea of influence ten
years after construction of the Major Project. Mote that the models predict
the total land use in the area of influence in each of the twelve categories,
not just the induced land use.*
Concurrent with the specification of the land use model, forty
case studies were selected (twenty of each major project type), based on
various criteria relating to geographical location, major project size and
phasing, and data base availability.
2. Phase 2
The second phase of the study was the collection of the requi-
site data, as identified by the specification of the model. After a test-
training case, this data collection phase was composed of two distinct tasks.
The first was a site visit which included interviews with individuals at the
regional planning agency and, if feasible, the developer. During the site
visit, aerial photograph interpretation was performed of the area of influ-
ence. The second task was the collection of requisite Census data.
3. Phase 3
The assumption of a single basic causal structure for induced
development, and the use of cross-sectional data from diverse locations
*The change in land use over the ten year period can be found by subtracting
the current land use from the estimated land use ten years in the future.
This change in land use would include the land use induced by the major
project as well as land use change due to general regional growth.
1-5
-------
throughout the United States, allowed a static approach to the testing of
the theoretical models, using path analysis. Paith analysis is a set of sta-
tistical techniques useful in testing theories and studying the logical con-
sequences of various hypotheses involving causal relations. It is not capa-
ble of deducing or generating causal relations, only testing them.
The causal analysis of induced land use development in the cur-
rent study involved the use of two basic statistical techniques: two-stage
least squares and stepwise ordinary least squares (multiple regression). The
first technique was required to produce consistent estimates of the path coef-
ficients in a system of simultaneous equations involving feedback loops or
reciprocal causation in the models. The second technique was used to solve
the remaining recursive portions of the models. The dependent variables in
these regression analyses represented the total land use in the previously
noted 12 categories. Both linear and non-linear forms were tested and the
linear form was found to produce the best fit. Specific statistical criteria
were developed to identify model paths that were insignificant or redundant,
and these criteria were used to trim unneeded and undesirable paths from the
models. A second complete path analysis was performed, and the trimming pro-
cess repeated several times until the final path models were decided upon.
The trimming process eliminated almost half of the paths in the models as
originally specified.
The final models of land use development show that strong sta-
tistical relationships exist between the variables representing the 12 cate-
gories of total land use and the other model variables representing induced
and non-induced land use growth processes. Only in the case of cultural
land use did the path analysis reject the hypothesized causal relationships.
2
Excluding this category, the R statistic for the model equations in the
simultaneous block ranged from 0.43 to 0.81 with an average value of 0.66
2
and the R values for the model equations in the recursive block ranged
from 0.12 to 0.86 with an average value of 0.41. These statistics can be
interpreted as the amount of variance in the dependent variables (total
land use) of the model equations that can be explained through the linear
1-6
-------
relationships in the final causal model. These results indicate a good
verification of the hypothesized land use development model.
There were several problems encountered in the path analysis,
involving multicol linearity, suppressor variables, choice of instrumental
variables, available degrees of freedom, and coefficient instability. The
first two problems were eliminated through the approach used for theory trim-
ming of the models. The last three problems were caused principally by a
common element: the small number of data samples (20) available for analysis.
In the model equations, as originally specified, there were sometimes as many
independent variables as data samples. Since at least several degrees of
freedom should be reserved for the error term in any multiple regression,
some model paths had to be trimmed prior to, and in order to perform, the
first path analysis. Thus, the limited data sample did preclude.the testing
of causal relationships in some instances. Also, an analysis of the stability
of the model path coefficients revealed appreciable instability in the indi-
vidual model coefficients when i,t was applied to different subsets of the
original land use data set. We note that this instability does not invali-
date.the strong causal relationships confirmed by the path analysis..
4. Phase 4 .
a. Model Calibration
The development of predictive equations for land use devel-
opment, separate from the model equations obtained in the causal analysis,
was. necessitated by the simultaneity of the causal relationships, i.e., the
causal equations include independent, variables whose values will not be
known in the future. Therefore, it was necessary to develop predictive
equations in which., the endogenous variables appeared only as the dependent
variables. Such an assumption defined a system of equations which was
solved with ordinary least squares analysis. Because of the poor perform-
ance of some of the exogenous variables in the causal analysis, the repre-
sentations of these variables were reconsidered and several new independent
variables formulated for use in the development of the predictive equations.
In order to systematically decide which variables to include in the predic-
1-7
-------
tlve equations, stepwise regression techniques were employed and objective
statistical criteria applied to obtain predictive equations for the 12
categories of total land use analyzed in the causal analysis.
Summary statistics indicate the predictive equations
explain the majority of variance in the dependent variables. The overall
F statistics indicate practically all of the predictive equations are sign-
ificant at or below the one percent level. The results for the coefficient
of variation, however, were less encouraging indicating that the average
error encountered in the use of these predictive equations will be +. 87
percent of the predicted value. In an attempt to reduce the coefficients
of variation for some of the predictive equations, the dependent variables
RES, COW, OFFICE, and MANF were defined in a second manner which did not
exclude the dwelling units or land use of the major project from the varia-
bles. Predictive equations using these dependent variables were found to be
less statistically significant, however.
Predictive equations were also developed for land use at a
finer level of detail, where, in addition to the 12 types of land use, the
size range (or density) of development for each type was used to categorize
the land use being predicted. The equations for disaggregated land use were
found to be in general not statistically significant. Therefore, average
percentage figures for these subcategories were developed instead. In addi-
tion to attempting to disaggregate the 12 categories of land use, predictive
equations were developed for an aggregated variable representing the total
developed floor area in the area of influence (including the major project).
The statistical results for these equations indicate that there is a certain
advantage to predicting total land use using an aggregated predictive equa-
tion, as opposed to summing together the predicted levels of 12 individual
predictive equations and a projection of the major project size.
The validity of any set of predictive equations depends
upon their generality. The preferred test of an equation's validity is an
external validation, viz., applying it on a test case basis to an independ-
ent sample of data (i.e., independent of the sample on which the coefficient
1-8
-------
values were based) and observing its predictive ability. In the current
study a separate, independent sample was not available. Therefore, the
analytical technique of cross-validation was used to simulate the existence
of such a test sample. This procedure involved splitting both the Residen-
tial and Indus trial/Office data samples of 20 into two random samples of 10
each. The first 10 samples were used to recompute the coefficient values
of the predictive equations and the second 10 samples were used as the inde-
pendent test sample. Statistical comparisons were then made between actual
and predicted values for the dependent variables in the second sample of 10.
The results indicate that about half of the 24 predictive equations are
generalized enough to produce good predictions using an independent sample.
Considering the extremely small sample size (10) used in the cross-valida-
tion and the large amounts of variance in the dependent variables, it was
encouraging to obtain significant correlations between actual and predicted
values in as many equations as we did. The equations that did not perform
well in the cross-validation are not necessarily useless. The poor correla-
tions obtained could simply be due to nonhomogeneity in the data sample
caused by the extremely small sample size. It was not possible to ascertain
if this was the case without a larger, independent sample.
b. Traffic Model
The development of the traffic model included both the
development of a methodology to predict the vehicle miles traveled (VMT)
by motor vehicular activity induced by the major project and induced land
uses as well as the specification of default values for use in the metho-
dology.
The basic methodology for predicting VMT is well known, viz.,
• Estimate vehicle trips by multiplying the amount of land
use by vehicle trip generation rates, and
• Estimate VMT by taking the product of the vehicle trips
and an estimated trip length.
1-9
-------
The important decisions in developing the methodology were then the amount
of disaggregation to employ. These included,
• What kinds of trips s'hould be treated separately with
respect to trip lengths and trip rates,
• What vehicle classes should be considered,
• How mass transits should be considered,
• How to estimate average route speed.
The final methodology considered two types of trips, (i.e., work and non-
work) in four vehicle classes with six different average route speeds i.e.,
(local streets, arterial, and expressways in peak hour and off peak condi-
tions). The impact of mass transit was assumed to be negligible.
5. Phases b and 6
Phases 5 and 6 were devoted to the compilation of a set of
land use based emission factors appropriate for use in the GEMLUP metho-
dology. For this reason, manufacturing emissions received less attention
than otherwise might have been appropriate.
The land use based emission factor was specified as grams of
pollutant per building floor area. This ratio may conveniently be expressed
as the product of two ratios, a fuel based emission factor (i.e., the
typical emission factor presented in AP-42 [ 8]), and an activity factor or
fuel consumption per building floor area. As the former ratio is well
known, the emphasis of Phases 5 and 6 were devoted to quantifying the fuel
consumption per unit of building floor area in each of the categories of
induced land use.
1-10
-------
II. LAND USE MODEL SPECIFICATION AND CAUSAL ANALYSIS
This chapter discusses the specification and causal analysis of the land
use model, as well as sample selection and data collection. It is a summary
of the first volume of the GEMLUP Final Report [5].
A. GENERAL APPROACH
1. Theory of Induced Development
Taking the industrial/office major land use project type as
the more general case of the two types investigated, we adopted the follow-
ing theory of induced development.
Constructing a large source of employment like an industrial/
office complex generates jobs which result in the nearby construction of
dwelling units; these induce retail development to locate near them and
generate demand for community, cultural, and religious facilities (schools,
recreation areas, libraries, churches, theaters, fire and police stations,
etc.). All of this requires the construction of streets and highways that
then improve accessibility to the area. Better access fosters continued ...
urban development, particularly highway-oriented commercial and office land
uses. Additional sources of employment come into the area as secondary
(and tertiary) industry or services locate near the original major project,
spurring on another round of residential development, and so forth.
2. Selection of General Approach
The selection of an approach for testing this theory was tem-
pered by programmatic considerations. The approach that was selected con-
sisted of a cross-sectional model that predicts the total land use in the
vicinity of a major project ten years after development of a major land use
project.
2-1
-------
3. Statement of Fundamental Model
The basic theory of Induced development may be restated as
follows: the amount induced land use is some function of the size of the
Major Development Project and certain other variables,, viz.,
induced land use = f (major project, other variables)
As indicated previously, the approach used in this study limits one to the
use of endogenous variables that measure the total land use at the end-of
the ten year time period. Conceptually, one can disaggregate the total land
use in the area of influence into three components, land use existing prior
to the-development of the major project, new land use induced by the major
project and certain other variables, and new land use not induced by the
major project (that is, attributable to some other phenomena such as general
regional growth). This may be expressed as:
total land use = prior land use +
project induced land use change +
non-project induced land use change
In predicting the total land use in an area of influence, one
can identify two types of exogenous variables:
Type I - Those used for predicting the induced land use
component, such as,
• The size of the major project,
• The induced component of the endogenous variables of
other land uses,
• Other independent variables influencing the effect of
the major project (i.e., housing vacancy t+0).
Type II - Those used for predicting the prior or non-
induced land use component,
• The prior and non-induced component of the endogenous
land use variables,
2-2
-------
• Those used for predicting the prior land use
component (i.e., 1960 housing density),
• Those used for predicting the non-induced
component (i.e., regional population, growth).
Accordingly, the fundamental model that we have assembled is
land uset+10 = priort+0 + 1nducedt+(J „ t+1Q
+ non-induced
t+0
= f(Type I variables, Type II variables).
We note that both the distinctions between the three land use
components and the two variable types are unavailable to this study. The
three land use components are not measurable; also, several of our
independent variables are possibly of both types.
B. LAND USE MODEL SPECIFICATION
1. Objective
The objective of this phase of the project was to specify an
initial land use model explaining induced or associated land use ten years
after the construction and operation of a major project. The basic theory
underlying the development of the model was that major projects have certain
associated or induced land uses and these land uses can be predicted based
on the characteristics of the major project and the area in which it locates.
However, because of the approach to testing this model, it is necessary to
include all land uses in the vicinity of the major project, whether they were
induced or not induced, or even existing prior to the construction of the
major project.
Two types of major projects were to be considered in the formu-
lation of the model, residential projects and office/industrial projects.
Because of the differing land uses associated or induced by these types of
projects, a separate model was constructed for each. Thus, two models were
developed, one explaining induced or associated land uses ten years after
2-3
-------
!
the construction and operation of a major residential project; the other
(<
explaining induced or associated land uses ten years after the construction
!>
and operation of a .major industrial/office project.
2. Definition of Major Project
For purposes of the model specification, a major residential
project was defined as housing facilities,, planned unit developments or new
towns containing a minimum population of 4,500; a major industrial/office
project was defined as an office or industrial park or a research and
development complex with a minimum employment of 2,250. Both types of pro-
jects were initially assumed to reach nearly 80 percent occupancy within two
years of operation. However, during case study selection, the definition of
major project was somewhat relaxed to permit phased projects.
In addition, for purposes of calibrating the model, the case
studies to be analyzed were required to be projects built between 1954 and
1964. The induced or associated land uses were those as of the year 1970;
i.e., the year by which it was assumed that the land use impacts of the pro-
ject had stabilized.
Based on a consideration of the typical size of potential major
projects relative to the potential size of the area of influence, it was
deemed appropriate to specify a fixed size for the area of influence. A
10,000 acre (4.0 x 10 square meters) area of influence was selected.
3. Model Specification Methodology
The specification of the model was based on (1) a literature
search to identify methodologies and case studies which had been used to
determine land uses associated with major projects and (2) the prior experi-
ence of personnel with land use planning and forecasting, land use models,
impact analyses, and large development projects.
2-4
-------
4. Model Description
a. Endogenous Variables
Due to the requirements of the emission factors, the units
of the endogenous variables for both the residential model and the industrial/
office model are building floor area (except for residential and outdoor
recreation land uses and highway lane miles) in each of 12 land use categories.
These land case categories are residential, retail, office, manufacturing,
wholesale and warehousing, hotel, hospital, cultural, churches, public
education, outdoor active recreation, and highway lane miles. These partic-
ular categories evolved from a process which balanced the following considera-
tions:
• What land use output was needed for
estimating emissions,
• What land use output could most effectively
be predicted using a causal model, and
• What land use output would be available
during data collection to calibrate the
model.
The model endogenous variables are defined in Table 2-1.
b. Exogenous Variables
The model consists of 23 independent variables. These
variables represent (the numbers refer to' the order in Table 2-1):
• Population housing and employment characteristics
(variables 13, 14, 16, 18, 19, 20, 21, 25, 26,
27, 28, 29, 30, 34),
• Accessibility measures (variables 17, 22, 24)
• Developability measure (variable 15)
• Regional influences (variables 23, 31, 32, 33,
35)
The independent variables for each equation were selected
because of their perceived causal relationship with the dependent variables.
2-5
-------
TABLE 2-1
MODEL VARIABLES AND DEFINITIONS
1. RES = Number of housing units in area of influence in 1970
(excluding major project).
2. COMM = Commercial land use in area of influence in 1970 in 1,000
square feet
Commercial land use includes the following land use codes
(LUC) as used by the Public Service Administration Service
in its 1962 Land Use Classification Manual.
LUC 52-59 Retail trade
61 Personal services
63 Automobile service
64 Miscellaneous repair service
65 Indoor amusement service
3. OFFICE = Office land use in area of influence (excluding major
project) in 1970 in 1,000 square feet
LUC 60 Finance, Insurance, Real Estate
62 Business services
67 Medical, Health, Legal services
68 Other professional services
4. MANF = Manufacturing land use in area of influence (excluding
major project) in 1970 in 1,000 square feet
LUC 2 Nondurable goods manufacturing
3 Durable goods manufacturing
5. WHOLE = Wholesale/warehouse land use in area of influence in 1970
in 1,000 square feet
LUC 50 Wholesale
46 Warehousing
6. HOTEL = Hotel and motel land use in area of influence in 1970 in
1,000 square feet
LUC 07 Hotels, Motels, Tourist Homes
7. HOSPTL = Hospital, etc., land use in area of influence in 1970 in
1,000 square feet
LUC 77 Hospitals, Sanatoria, Convalescent Homes and Rest
Homes
2-6
-------
TABLE 2-1 (CONTINUED)
MODEL VARIABLES AND DEFINITIONS
8. CULTUR « Cultural land use in area of influence in 1970 in 1,000
square feet
LUC 76 Museums, Libraries, Art Galleries, except,
Churches (764)
Arboreta (762)
i Cemeteries (767)
9. CHURCH = Religious land use in area of influence in 1970 in 1,000
square feet
LUC 764 Churches
765 Other religious services
10. EDUC = Public educational land.use in area of influence in 1970
in 1,000 square feet
LUC 74 Public Schools
11. REC = Active outdoor recreational land use in area of influence
in 1970 in acres
12a. HWLMNX = Highway lane miles in area of influence in 1970, excluding
limited access highways ' ;
12b. HWLM = Highway lane miles in area of influence in 1970
13a. MPR70 = Residential land use in major project in 1970 in dwelling
uni ts
13b. MPR68 = Residential land use in major project in 1968 in dwelling
uni ts
13c. MPRt2 = Residential land use in major project in base year plus
two (t+2) in dwelling units
14. DUACRE = Dwelling units per acre in area of influence in 1960
15. VACACR = Percent vacant developable acreage in area of influence
in year (t+0)
16. VACHSG = Percent vacant housing in area of influence in 1960
17. HWYINT = Highway interchanges in area of influence in year (t+5)
2-7
-------
TABLE 2-1 (CONTINUED)
MODEL VARIABLES AND DEFINITIONS
18. MINCC = Median income of families and individuals in area of
influence relative to U.S. median income in 1960
19. INCMP = Variable indicating the median income level of major
project compared to surrounding community in year (t+2)
20. OFFVAC = Percent office buildings vacant in metropolitan area in
year (t+0)
21. OFFACR = Office employment per acre in area of influence in year
(t+0)
22. DISCBD = Distance from center of major project to CBD in year (t+0)
!
23. ENERGY = Cost factor for electricity (S/1500 kWh) for commercial
users in the metropolitan area in year (t+0) divided by
the average U.S. commercial rate in 1960
24. RRMI = Railroad mileage in area of influence in year (t+0)
25. WWEA = Warehouse and wholesale employment per acre in area of
influence in year (t+0)
26. EMPACR = Total employment per acre in area of influence in year
(t+0)
27. NONHSE = Nonhousehold population per acre in area of influence in
1960
28. MPKIDS = School-age children per dwelling unit in major project
in year (t+2)
29. ENRACR = Public school enrollment per acre in area of influence
in 1960
30. MANACR = Manufacturing employment per acre in area of influence
in year (t+0)
31. DELPOP = Growth factor for total regional population between 1960
and 1970 (county data)
32. DELEMP = Growth factor for total regional employment between 1960
and 1970 (county data)
2-8
-------
TABLE 2-1 (CONTINUED)
MODEL VARIABLES AND DEFINITIONS
33. MINCR = Median income of the region in year (t+0) relative to the
median U.S. income in 1960
34a. MPE70 = Number of employees in major project in 1970
34b. MPE68 = Number of employees in major project in 1968
34c. MPEtZ = Number of employees in major project in base year (t+2)
35. AUTO = Automobile drivers per acre in county in 1960
2-9
-------
Prior to selecting these variables a .complete 11st of all possible factors
influencing the dependent variable was prepared. From this list the most
significant variables were identified.
The specific format for each independent variable was
developed based on 1) availability of data, 2) consistency of data among case
studies, and 3) the appropriate time period for the data.
c. Equations
Twelve equations were specified to predict each of the twelve
endogenous variables. In both models (i.e., the residential model and the
industrial/office model), five of these equations are simultaneous. The
remaining seven equations are recursive.
C. SAMPLE SELECTION
1. Purpose and Initial Criteria
The purpose of the sample selection process was to identify for
each type of major project a sample of case studies which could be used in
the testing and calibration of the model. Once a list of qualified case
studies was prepared, the actual selection of the final sample took place.
This selection process involved consideration of factors such as availability
of information, particularly aerial photographs and geographic location of
the project. The final list of case studies is shown for the industrial/
office and residential sample in Table 2-2.
D. DATA COLLECTION
Following the specification of the model, a list of data items
required for the model was prepared. This was supplemented with additional
items potentially useful in the model calibration phase of this project.
The data collection process consisted of two simultaneous phases.
The first was an on-site visit which primarily consisted of interviews with
the local and regional planning agencies, the developer (if available), and
2-10
-------
TABLE 2-2
CASE STUDIES
Industrial
Residential
ro
i
Farmington Park, Farmington, CT
Western Electric, North Andover, MA
Avco, Wilmington, MA
IBM, Kingston, NY
Ft. Washington, Philadelphia, PA
Keystone Park, Scranton, PA
Crestwood Park, Wright Turnpike, PA
General Electric, Salem, VA
Cummings Park, Huntsville, AL
IBM, Lexington, KY
Collins Radio Park, Cedar Rapids, IA
Ford, Woodhaven, MI
Western Electric, Columbus, OH
White-Westinghouse, Columbus, OH
Little Rock Industrial Park, Little Rock, AR
Chrysler, Fenton, MO
Western Electric, Omaha, NE
Motorola, Phoenix, AZ
Tektronix, Washington County, OR
Honeywell, Phoenix, AZ
Joppatown, Hartford County, MD
Montgomery Village, MD
Kings Park, Fairfax County, VA
Vienna Woods, Fairfax County, VA
DeltOna, FL
Miami Lakes, Miami, FL
Town'n Country, Tampa, FL
Montclair-Starmount, Charlotte, NC
Weathersfield, Schaumberg, IL
Oak Park, Blaine, MN
Cottage Grove, MN
Clear Lake City, Harris County, TX
Meyerland, Houston, TX
Westwood Heights, Omaha, NE
Northglenn, CO
Sun City, Maricopa, AZ
Foster City, CA
Huntington Harbour, Orange County, CA
Sun City, Perris, CA
Rancho Bernado, CA
-------
the collection of locally available data. This data consisted primarily of
building floor area of various categories of land uses in the area of influ-
ence obtained by aerial photograph interpretation. The second phase con-
sisted of the collection of 1960 and 1970 Census of Population and Housing
Data for the area of influence.
E. CAUSAL ANALYSIS
1. General Approach
The approach to path analysis in the current study involved
the use of two basic statistical techniques: two-stage least squares and
ordinary least squares multiple regression. The first technique was used
to solve for path coefficients in the system of five equations connected
by feedback loops. For a given dependent variable, the first stage of the
two-stage process involved estimating the values of the other four endogenous
variables through linear combinations of so-called instrumental variables
which are chosen for their causal relationships with the endogenous variables.
To solve for path coefficients in the other model equations that
were not interconnected, ordinary least squares regression techniques were.
used.
2. Data Transformations
The land use and demographic data collected in the field program
were loaded into a field data file on our computer system for processing.
Computations were performed on these data to create the model variables
chosen for path analysis. These data transformations are summarized in a
list of variable definitions in Table 2-3.
Analyses were subsequently performed on model variables to test
for multicollinearity, possible suppressor variable problems, and the suita-
bility of instrumental variables (used in solving feedback loops). The
model was trimmed as a result of the investigations and the first path
analysis performed.
2-12
-------
TABLE 2-3
PATH ANALYSIS MODEL DATA TRANSFORMATIONS
Model Variable
1. RES = Residential land use in area of influence in 1970 (excluding
major project) in dwelling units
RES = du70t - mpr70
where: du70t = dwelling units in area of influence in 1970
mpr70 = residential land use in major project in 1970
in dwelling units
2. COMM = Commercial .land use in area of influence in 1970 in 1,000
square feet
COMM = (comml + comm2 + commS + comm4)/10
where: comml = TOO square feet commercial in area of
influence in 1970 (<25K)
comm2 = 100 square feet commercial in area of
influence in 1970 (25-50K)
commS = 100 square feet commercial in area of
influence in 1970 (50-1OOK)
comm4' = 100 square feet commercial in area of
influence in 1970 (>100K)
3. OFFICE = Office land use in area of influence (excluding major project)
in 1970 in 1,000 square feet
OFFICE = (offl + oof2 + off3)/10
where: offl = 100 square feet office in area of influence
(excluding major project) in 1970 (<50K)
off2 = 100 square feet office in area of influence
(excluding major project) in 1970 (50-1OOK)
off3 = 100 square feet office in area of influence
(excluding major project) in 1970 (>100K)
4. MANF = Manufacturing land use in area of influence (excluding
major project) in 1970 in 1,000 square feet :
MANF = manf/10
where: manf = 100 square feet manufacturing in area of
influence (excluding major project) in
1970
2-13
-------
TABLE 2-3 (CONTINUED)
PATH ANALYSIS MODEL DATA TRANSFORMATIONS
Model Variable
5. WHOLE = Wholesale/warehouse land use in area of influence in 1970 in
1,000 square feet
WHOLE = whole/10
where: whole = 100 square feet wholesale/warehouse in area
of influence in 1970
6. HOTEL = Hotel and motel land use in area of influence in 1970 in
1,000 square feet
HOTEL = (hotel! + hotelZ + hotels + hote!4)/10
where: hotel 1 = 100 square feet hotel in area of influence
in 1970 (<25K)
hote!2 = 100 square feet hotel in area of influence
in 1970 (25-50K)
hotels = 100 square feet hotel in area of influence
in 1970 (50-100K)
hotel4 = 100 square feet hotel in area of influence
in 1970 (>100K)
7. HOSPTL = Hospital, etc., land use in area of influence in 1970 in
1,000 square feet
HOSPTL = (hospl + hosp2 + hosp3)/10,
where: hospl = 100 square feet hospitals in area of
influence -in 1970 (25-50K)
hosp2 = 100 square feet hospitals in area of
influence in 1970 (50-1OOK)
hosp3 = 100 square feet hospitals in area of
influence in 1970 (>100K)
8. CULTUR = Cultural land use in area of influence in 1970 in 1,000
square feet
CULTUR = cultur/10
where: cultur = 100 square feet cultural in area of
influence in 1970
9. CHURCH = Religious land use in area of influence in 1970 in 1,000
square feet
CHURCH = church/10
where: church = 100 square feet religious in area of
influence in 1970
2-14
-------
TABLE 2-3 (CONTINUED)
PATH ANALYSIS MODEL DATA TRANSFORMATIONS
Model Variable
10. ED = Educational land use in area of influence in 1970 in 1,000
square feet
ED = (edl + ed2 + ed3)/10
where: edl = 100 square feet education in area of ..
influence in 1970 (<25K)
ed2 = 100 square feet education in area of
influence in 1970 (25-50K)
ed3 = 100 square feet education in area of
influence in 1970 (>100K)
11. REC = Active outdoor recreational land use in area of influence
in 1970 in acres
12. HWLM = Highway land miles in area of influence in 1970
12a. HWLMNX = Highway lane miles in area of influence in 1970 without
expressways
13. MPRT2 = Residential land use in major project in year t+2 in
dwelling units
13a. MPR68 = Residential land use in major project in 1968 in dwelling
units
13b. MPR70 = Residential; land use in major project in dwelling units
14. DUACRE = Dwelling units per acre in census tracts in 1960
DUACRE = (du60c - mpr60)/ac60c
where: du60c = dwelling units in census tracts
ac60c = census tract acreage in 1960
mpr60 = dwelling units in major project in
1960
15. VACACR = Percent vacant developable acreage in area of influence
in year (t+0)
VACACR = vacdev/(10,000-vacund)
VACUND = Vacant undevelopable acreage in area of influence in
year (t+0)
where: vacdev = vacant developable acreage in area of
influence in year (t+0)
2-15
-------
TABLE 2-3 (CONTINUED)
PATH ANALYSIS MODEL DATA TRANSFORMATIONS
Model Variable
16. VACHSG = Percent vacant housing in census tracts in 1960
VACHSG = vac60c/du60c
where: vac60c = Vacant available housing units in census
tracts in 1960
17. HWYINT = Highway interchanges in area of influence in year (t+0)
18. MINCC = Median income factor for families and individuals in
census tracts relative to average U.S. income in 1960
MINCC = mincc/$5,650
where: mince = Median income for families and
individuals
19. INCMP = Variable indicating the median income level of major
project compared to surrounding community in year (t+2)
INCMP = incmpa - incmpb
where: incmpa = +1 if major project median income
>.15 percent above that of surrounding
community
incmpb = +1 if major project median income
2ll5 percent below that of surrounding
communi ty
20. OFFVAC = Percent office buildings vacant in metropolitan area in
year (t+0)
21. OFFACR = Office employment per acre in area of influence in year
(t+0)
OFFACR = offemp/10,000
where: offemp = Office employment in area of influence
in year (t+0)
22. DISCBD = Distance from center of major project to CBD in year
(t+0) in miles
23. ENERGY = Cost factor for electricity ($/1500 kWh) for users in the
metropolitan area in year (t+0) relative to the average
U.S. commercial rate in 1960
ENERGY = energy/$51.59
where: energy = Dollars per 1500 kWh for commercial
users in metropolitan area in year
(t+0)
2-16
-------
TABLE 2-3 (CONTINUED)
PATH ANALYSIS MODEL DATA TRANSFORMATIONS
Model Variable
24. RRMI = Railroad mileage in area of influence in year (t+0)
25. WWEA = Warehouse and wholesale employment per acre in area of
influence in year (t+0)
WWEA = wwemp/10,000
where: wwemp = Warehouse and wholesale employment in
area of influence in year (t+0)
26. EMPACR = Total employment per acre in area of influence in year
(t+0)
EMPACR = : totemp/10,000 .- • ,
where: totemp = total employment in area of influence .
.in year (t+0)
27. NONHSE = Nonhousehold population per acre in. census tracts in 1960
NONHSE = (p60c - hp60c)/ac60c
where: p60c = Total population in census tracts in 1960
hp60c = Household population in census tracts in
1960
28. MPKIDS = School-age children per dwelling unit in major project in
year (t+2)
29. ENRACR = Public school enrollment per acre in census tracts in 1960
ENRACR = p!460c/ac60c
where: p!460c = Population under 14 years of age in
census tracts in 1960
30. MANACR = Manufacturing employment per acre in area of influence in
year (t+0)
MANACR = manemp/10,000
where: manemp = Manufacturing employment in area of
influence in year (t+0)
31. DELPOP = Growth factor for total regional population between 1960
and 1970 (county data)
DELPOP = (p70cty - p60cty)/p60cty
where: p70cty = Total population in county in 1970
p60cty = Total population in county in 1960
2-17
-------
TABLE 2-3 (CONTINUED)
PATH ANALYSIS MODEL DATA TRANSFORMATIONS
Model Variable
32. DELEMP
DELEMP
33. MINCR
MINCR
34. MPET2
34a. MPE68
34b. MPE70
35. AUTO
AUTO
Growth factor for total regional employment between 1960
and 1970 (county data)
(e70cty - e60cty)/e60cty
where: e70cty = Total employment in county in 1970
e60cty = Total employment in county in 1960
Median income factor for the region in year (t+0) relative
to the average U.S. income in 1960
mincr/$5,660
where: miner = Median income for the region in the year
(t+0) (county data)
Number of employees in major projects in year (t+2)
Number of employees in major projects in year 1968
Number of employees in major projects in year 1970
Automobile drivers per acre in county in 1960
au60cy/ac60cy
where: au60cy = Automobile drivers in county in 1960
ac60cy = County acreage in 1960
2-18
-------
3. Theory Trimming
The results of the first path analysis revealed many path coef-
ficients that were too low or of the wrong sign as predicted from theory.
In addition, the overall t-statisties of several of the model equations
indicated no statistical significance at the five percent level. Thus,
criteria were developed which when applied to the numerical output of the
first path analysis trimmed many model paths. A second path analysis was
performed, and the process repeated several times until a final path model
was decided upon. This recursive procedure was necessitated by the fact that
trimming one variable often causes significant changes in the path coeffi-
cients of the remaining variables.
In trimming the path analysis model, the following rules were
applied. A path was trimmed if:
• |t| or F < 1.0, and 3 < 0.1, and loss of the variable
would not cause the loss of a significant instrumental
variable in the first stage estimations,
• The sign of path coefficients (3) was wrong and
counter to the original path model hypothesis.
In addition, some paths whose statistics indicated they should
be trimmed were kept if the sign of the path coefficients (3) was correct
and the variables were very important causally in the path model. Finally,
the discovery of 3 > 1.0 in a few model equations indicated correlations
between the independent variables to the extent that some paths were redun-
dant. In these instances, simple correlations between the various path
coefficients were computed and the most highly correlated variable was
trimmed. Correlation coefficients were calculated from the variance-
covariance matrix of the estimated coefficients by dividing the covariance
of the coefficients of the two variables in question by the product of their
individual standard deviations.
The results of the two-stage least squares and the ordinary
least squares regressions performed for the final path analysis are presented
in Figures 2-7 and 2-8 for the Residential and Industrial/Office Models,
2-19
-------
RCSIDENTIAL
MAJOR PROJECT
1970
Figure 2-7 Final Path Analysis for the Residential Model
-------
.J5&,
EDUCATIONAL
.25
c-18
NO. OF EMPLOY
IN MAJOR PROJECT
(1970,
Figure 2-8 Final Path Analysis for the Industrial-Off ice Model
-------
p
respectively. The path coefficients (3) are shown on each path and the R
for each regression is displayed next to the associated dependent variable.
For each equation, the effect on the dependent variable of all residual
causes can be quantified by the path coefficient for residual causes,
defined as:
6e - 0-R2)1'2
a. Summary of Final Model Equations in Unstandardized Form
The final model equations are summarized below with unstand-
ardized coefficients.
(1) English Units
(a) Residential Model
RES = -.1.38 OFFICE + 168 HWLMNX - 0.808 MPR70 + 3930 DUACRE + 2730
COMM = 0.0814 RES +0.649 OFFICE +18.4 HWLMNX + 0.0976 MPR70 - 692
OFFICE = - 0.0319 RES + 5.74-HWLMNX + 0.0572 MANF-- 0.0127 VACACR
- 5.26 OFFVAC + 690 OFFACR - 15.4.DISCED + 765 DELEMP + 421
MANF = 1470 MANACR +316 HWYINT + 4100 ENERGY - 0.176 MPR70 +79.8 RRMI
- 3940
HWLMNX = 0.00247 RES - 3.31 HWYINT - 1.73 DISCBD + 47.9
WHOLE = 60.4 HWYINT + 0.0608 MANF + 26.5
HOTEL = 0.00151 RES + 0.0140 MANF + 49.3
HOSPTL = 0.0106 RES + 0.0246 MPR70 - 61.6
CULTUR =0.00014 RES + 0.00154 MPR70 - 0.447 DISCBD + 19.5
CHURCH = 0.0134 RES + 0.00716 MPR70 + 13.3
EDUC = 0.0392 RES + 209 MPKIDS + 0.0203 MPR70 - 137
REC = 329 INCMP - 563 MINCC + 826
(b) Industrial/Office Model
RES =3.96 OFFICE - 0.859 MPET2 + 0.392 MPE70 + 4560 DUACRE
+1.05 VACACR - 4860
2-22
-------
COMM = 0.0785 RES + 0.413 MANF + 0.0367 MPE70 + 39.1 HWYINT
- 2270 DELEMP - 290 MINCC - 45.2
OFFICE = 0.0110 RES + 0.0178 MANF - 0.0199 MPE70 + 0.0195 MPET2
- 285 OFFACR - 8.11 DISCED + 902 DELEMP + 174
MANF = 4.55 HWLMNX + 0.124 MPET2 - 0.0974 MPE70 + 55.6 RRMI
+ 8690 DELEMP + 1020 MANACR - 22.5
HWLMNX = 0.00178 RES - 0.0209 OFFICE + 0.00980 COMM - 4.59 HWYINT
- 0.246 DISCED + 32.6 AUTO +14.3. ;
WHOLE = 10,500 WWEA +34.7 HWLMNX + 29.0 RRMI + 64.6 HWYINT - 1120
HOTEL = 0.0524 MANF + 0.105 OFFICE + 0.0182 MPE70 - 0.0156 MPET2
- 4.04 DISCED - 26.0
HOSPTL = 508 NONHSE - 5.66 DISCED + 108
CULTUR = 0.0023 RES - 0.249 DISCED + 13.6
CHURCH = 0.00513 RES + 51.1 . '
EDUC = 0.0405 RES + 304
REC = 0.0149 RES + 242 MINCC - 184
(2) Metric Units. As can be:seen in Tables 2-1 and 2-3,
the model variables were defined and the path analyses carried out in English
units. Thus, the path regression coefficients given for each equation in,
the previous section reflect this fact. The path diagrams in this section,
however, display the path coefficients (3) which are independent of the units
chosen.
This section summarizes the final model equations for
variables RES, COMM, OFFICE, MANF, WHOLE, HOSPTL, CULTUR, CHURCH, AND EDUC.
2
The variables REC and VACACR also use m but in place of acres. Distances
in miles in HWLMNX, DISCED, and RRMI are converted to km. Finally, the
-1-2
units of acre are replaced by m in the variables DUACRE, OFFACR, WWEA,
EMPACR, NONHSEy ENRACR, MANACR, DELEMP, and AUTO.
The conversion factors used were:
1 ,000 ft2 - m2
acres •*• m
*by 92.90
*by 4,047
2-23
-------
2
miles ->• m *by 1.609
acre" •*• m" *by 4047
(a) Residential Model
RES = - 1.38 OFFICE = 9700 HWLMNX - 75.1 MPR70 + 1,480,000,000 DUACRE
+ 254,000
COMM = 0.0814 RES + 0.649 OFFICE + 1,060 HWLMNX +9.07 MPR70
- 260,000,000 DELEMP - 12,900
OFFICE = - 0.0319 RES +331 HWLMNX + 0.0572 MANF - 0.000292 VACACR
- 489 OFFVAC + 260,000,000 OFFACR - 889 DISCBD + 2
+ 288,000,000,000 DELEMP
MANF = 553,000,000 MANACR + 29,400 HWYINT + 381,000 ENERGY - 16.4 MPR70
+ 4,610 RRMI - 366,000
HWLMNX = 0.0000428 RES - 5.33 HWYINT - 1.73 DISCBD + 771
WHOLE = 5,610 HWYINT + 0.0608 MANF + 2,460
HOTEL = 0.00151 RES + 0.0140 MANF + 4,580
HOSPTL = 0.0106 RES + 2.29 MPR70 - 5,720
CULTUR = 0.00014 RES + 0.143 MPR70 - 2.58 DISCBD + 1.810
CHURCH = 0.0134 RES +0.665 MPR70 + 1,240
EDUC = 0.0392 RES + 19,400 MPKIDS + 1.89 MPR70 - 12,700
REC =1,330,000 INCMP - 2,280,000 MINCC + 3,340,000
(b) Industrial/Office Model
RES = 3.96 OFFICE - 79.8 MPET2 + 36.4 MPE70 + 1,710,000,000 DUACRE
+ 0.0241 DUACRE - 451,000
COMM = 0.0785 RES + 0.413 MANF + 3.41 MPE70 + 3.630 HWYINT
- 853,000,000 DELEMP - 26,900 MINCC - 4,200
OFFICE = 0.0110 RES + 0.0178 MANF - 1.85 MPE70 + 1.81 MPET2
- 107,000,000 OFFACR - 468 DISCBD + 339,000,000 DELEMP + 16,200
MANF = 263 HWLMNX + 11.5 MPET2 - 9.05 MPE70 + 3,210 RRMI
+ 3,270,000,000 DELEMP + 383,000,000 MANACR - 2,090
HWLMNX = 0.0000308 RES - 0.000362 OFFICE + 0.000170 COMM - 7.39 HWYINT
- 0.246 DISCBD + 212,000 AUTO + 23.0
2-24
-------
WHOLE = 3,948,000,000 WWEA + 2,000 HWLMNX + 1.670 RRMI + 6,000 HWYINT
- 104,000
HOTEL = 0.0524 MANF + 0.105 OFFICE + 1.69 MPE70 - 1.45 MPET2
- 233 DISCED - 2,420
HOSPTL = 191,000,000 NONHSE - 327 DISCBD + 10,000
CULTUR = 0.0023 RES - 14.4 DISCBD + 1,260
CHURCH = 0.00513 RES + 4,750
EDUC = 0.0405 RES + 28,200
REC = 0.649 RES + 979,000 MINCC - 745,000
2-25
-------
III. DEVELOPMENT OF LAND USE MODEL PREPICTIVE EQUATIONS
A, APPROACH
The development of predictive equations for land use development,
separate from the model equations obtained in the causal analysis, was neces-
sitated by the simultaneity of the causal relationships. Since the endo-
genous variables (to be predicted for some future time period) appear as
both independent and dependent variables in the causal model equations,
these equations clearly can not be used for operational, predictive purposes.
Or more simply, these causal equations include independent variables whose
values will not be known in the future. Therefore, it was necessary to
develop predictive equations in which the endogenous variables appeared
only as the dependent variables. Such an assumption defines a system of
equations which can be solved with ordinary least squares analysis.
When predictive equations are developed using ordinary least squares
for variables which are known to be effected by simultaneity, the individual
regression coefficients have expectations equal to the true structural para-
meter, plus a function of the variables left out of the regression; that is,
they are biased estimates of the true structural parameters (see Section
VI.A of Volume I [ 5]). However, the final prediction of the equation
is an unbiased estimate of the dependent variable. Since these equations
are to be used for predictive and not analytical purposes, the fact that the
individual coefficients are biased estimates was not of great concern. It
must be emphasized therefore, that the regression coefficients obtained for
the predictive equations should not be examined to judge the effects of
independent variables on the dependent variable, nor should these coeffi-
cients be compared with those obtained in the causal analysis. Rather the
appropriateness of a predictor can only be determined by examining its
performance with regard to a set of objective statistical criteria.
3-1
-------
In order to systematically decide which variables to include in
the predictive equations, stepwise regression techniques were employed.
The dependent variables in this stepwise regression analyses were the 12
categories of total land use analyzed in the causal analysis:
. Residential • Hotels/Motels
• Commercial • Hospitals
• Office • Cultural
• Manufacturing • Churches
• Highways • Education
• Wholesale/Warehouse • Recreation
The independent variables included in the stepwise regression analysis for
predictive equations were separated into two sets. The first set (1) in-
cluded all instrumental variables used in the causal model which best pre-
dicted the endogenous variables in the first stage of the 2-stage least
squares procedure (see Table 3-1) and those exogenous variables which were
significant in the final path analysis. The second set (2) of independent
variables included all other exogenous variables from the original speci-
fication of the models (shown in Table 3-2} as well as many new independent
variables, discussed in the next Section III.B.
The approach to developing predictive equations involved two step-
wise regression analyses. The first analysis performed allowed all vari-
ables from set (1) with a statistically significant F statistic to enter
the regression equation before any variables from set (2) were considered.
The second analysis included both sets of independent variables (1) and (2)
at the same level of consideration. In the analyses, an F ratio test was
used at each step of the multiple regression procedure to determine whether
the reduction in the residual sum of squares due to the added variable was
statistically significant. The critical F value used was 1.8, i.e., an
independent variable was added into the regression if its F ratio equalled
or exceeded 1.8. This critical value corresponds approximately to the
3-2
-------
TABLE 3-1
INSTRUMENTAL VARIABLES USED IN THE FINAL PATH ANALYSIS
Residential Model
Industrial/Office Model
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
DUACRE
OFFACR
DISCBD
MPR70
VACACR
OFFVAC
HWYINT
DELEMP
MANF
DUACRE
MINCC
OFFACR
DISCBD
MANACR
MPE70
VACACR
RRMI
OFFVAC
AUTO
HWYINT
DELEMP
MPET2
3-3
-------
TABLE 3-2
EXOGENOUS VARIABLES INCLUDED IN THE ORIGINAL MODEL SPECIFICATIONS
BUT TRIMMED PRIOR TO THE FINAL PATH ANALYSIS
Model Equation
Exogenous Variables
Residential Model
RES
COMM
OFFICE
MANF
HWLMNX
WHOLE
HOTEL
HOSPTL
CULTUR
CHURCH
EDUC
REC
Industrial/Office Model
RES
COMM
OFFICE
MftNF
HWLMNX
WHOLE
HOTEL
HOSPTL
CURTUR
CHURCH
EDUC
REC
VACACR, VACHSG, DELPOP, MPRT2, MPR68
INCMP, MINCR, MINCC, HWYINT, MPRT2, MPR68
HWYINT, MPRT2, MPR68, MPR70
DELEMP, MPRT2, MPR68
EMPACR, AUTO, MPRT2, MPR68, MPR70
DELEMP, RRMI, WWEA
EMPACR, DISCBD, MPRT2, MPR68, MPR70
DISCBD, NONHSE, MPRT2, MPR68
MPRT2, MPR68
MPKIDS, MPRT2, MPR68
ENRACR, MPRT2, MPR68
MPRT2, MPR68, MPR70
DELPOP, VACHSG, MPE68
MINCR, MPET2, MPE68
OFFVAC, VACACR, HWYINT, MPE68
HWYINT, ENERGY, MPE68
EMPACR, MPET2, MPE68, MPE70
DELEMP
EMPACR, MPE68
MPET2, MPE68, MPE70
ENRACR
3-4
-------
10 percent level of significance, i.e., there was at most a 10 percent
chance of accepting a variable as significant in the regression when it
actually was not.
For each dependent variable, the final predictive equation chosen
from the many possible forms generated by the stepwise regression analyses
P
was the one with the highest adjusted R value which had six or less predic-
tors, each with an F statistic significant at the 10 percent level or better,
and none having a $ value greater than 1.0. The adjusted R is defined as:
2O t_l 9
_ n<- /"Ji M n DM
" a " R lN-k} U"R '
where:
R = coefficient of determination
k =,number of independent variables
N = number of data samples
2
The conventional R statistic can yield deceptive results when the signifi-
cance of specific predictors is in question. For example, simply adding a
variable to any regression equation, whether it is at all correlated with
the dependent variable or not, will raise the R2 and indicate additional
variance has been explained. The R2 gives a more conservative, unbiased
Q
estimate of the amount of variance explained in the dependent variable
through the regression equation. In cases where two forms had nearly
identical R values, the one containing more causally important variables
(i.e., set (1) variables) was chosen. The maximum limit of six variables
was set to keep the predictive equations simple, avoid possible degrees of
freedom problems, and keep the confidence intervals small. In addition,
examination of the results of the stepwise analyses showed that in most
cases, the first six variables explained most, if not all, of the variance
of the dependent variable that could be explained by including more inde-
pendent variables. The requirement that each individual predictor have a
significant F statistic eliminated the situation where the inclusion of a
3-5
-------
new variable in an equation caused the individual F statistics of previously
entered variables to drop below the critical value of 1.8. The restriction
on 3 values was used to avoid multicollinearity in the regression analyses
which might cause instability in the regression coefficients. One further
criterion used in simplifying the equations was that an independent variable
was not included unless its presence caused a noticeable reduction in the
coefficient of variation from the previous step (viz., a minimum of two-
three percent change). The coefficient of variation is defined as the ratio
of the standard error of estimate of the regression to the mean value of the
dependent variable. A detailed discussion of the results of the regression
analyses is given in Section III.C.
B. NEW VARIABLE DEFINITIONS
Based upon the poor performance of some of the exogenous variables
in the causal analysis, the representations of these variables were reconsi-
dered and several new independent variables formulated for use in the
development of the predictive equations.
The variable DELPOP was defined in a second manner, on a growth
per acre basis:
DELP2 = (p70cty-p60cty)/ac60cy*
We believe that whenever regional population growth is large, there is in-
creased demand for all types of land use development. Thus, this new
variable was included in all of the predictive equations.
Due to the variation in the cost of living across the U.S., areas '-
with the same MINCC or MINCR values may have different standards of living.
To eliminate this problem, a new variable MINC was defined to indicate whether
the median income of the area of influence was above or below that for the
region. This variable was defined as:
MINC = mincc/mincr
*See Table 2-3 for a definition of all lower-case variables in this and
subsequent equations which are not defined.
3-6
-------
This new variable was included in all of the predictive equations but RES,
MANF, and WHOLE. Since RES represents total dwelling units in the current
model, and not housing value or acreage, it is doubtful that a relationship
exists between RES and MINC.
Due to the poor performance of HWYINT in the causal analysis, the
data collected was examined and it was discovered that in 40 percent of the
samples HWYINT = 0, with a maximum value of 6 occurring. HWYINT is an
indicator of the presence of limited-access expressways in the area of in-
fluence. Due to the small area of influence in the current study (10,000
acres), it is doubtful that the presence of multiple highway interchanges
induces more land use development than the presence of just one. Therefore,
a new variable HWYLIM was defined to indicate whether a limited-access
highway is present or not .in the area of influence.
HWYLIM = ° if HWYINT =0
1 if HWYINT > 0
This new variable was included in the predictive equations for RES, COMM,
OFFICE, MANF, WHOLE, HOTEL, and HWLMNX.
The variable AUTO was included in the model equation for HWLMNX
on the theory that the density of auto drivers per acre effects the construc-
tion of highway facilities. Since AUTO was not found to be causally impor-
tant in the HWLMNX equation in the Residential model, a new variable AUT02
was defined to represent motor vehicle density in a different way as auto
drivers per dwelling unit:
AUT02 = au60cy/du60cy
where: duSOcy = total dwelling units in county in 1960.
(input data card 7, columns 11-16)*
This new variable was included in the predictive equations for HWLMNX.
* See Sections V and VLB in Volume I [5] for a discussion of the data col-
lection and formatting tasks.
3-7
-------
Another new variable related to highway development , TRIPS, was
defined to represent the amount of trip generation per dwelling unit from a
residential major project. Since specific trip generation data were not
available, a metric variable of arbitrary units related to the income level
of the major project was used instead (see Section V). The new variable
was defined as:
1 if INCMP = -1
TRIPS = 3 if INCMP = 0
5 if INCMP = 1
The variable TRIPS was only included in the predictive equation for HWLMNX
in the Residential model.
The indicator variable INCMP as defined for the causal analysis
assumed that residential major projects with a median income more than
15 percent above or below that of the surrounding community had equal but
opposite effects in inducing development,relative to a major project with
a median income equal to that of the surrounding community. This metric
assumption placed an unrealistic constraint on the variable. Therefore, two
new indicator variables INCMPL and INCMPH were defined to represent the low
and high income effects separately. These new variables were defined as:
1 if INCMP = -1
INCMPL = o if INCMp = o or 1
INCMPH = ] if INCMP B ]
0 if INCMP = 0 or -1
The variables INCMPL and INCMPH were only included in the predictive equa-
tions for COMM and REC in the Residential model.
The variable DELEMP was included in the model equations to account
for development associated with regional employment growth. Due to its
3-8
-------
poor performance in the causal analysis, a new variable EMP60 was defined
to represent the absolute level of regional employment in 1960, rather than
the growth in employment from 1960 to 1970. This new variable was defined
as:
EMP60 = e60cty
The variable EMP60 was included in the predictive equations for COMM, OFFICE,
MANF, and WHOLE.
The variable VACACR was included in the model equations for RES and
OFFICE to represent developable acreage in the area of influence (excluding
the major project). Vacant developable land is a prime factor encouraging
new land use development. Unfortunately, VACACR performed poorly in the
causal analysis. A possible explanation for this variable's failing in the
causal analysis may be ascribed to vacant acreage in the major project,
developed between the year t+0 and 1970. Note that the total land use vari-
ables RES and OFFICE for which relationships were hypothesized in the causal
analysis exclude the acreage of the major project. Therefore, any variable
introduced into a model equation to explain recent development included in
these land use totals must be corrected for all influences of the major
project. In the causal analysis a correction was made by defining VACACR to
exclude all acreage under the control of the major project developer. How-
ever, the major project probably had a secondary effect on new land use
development that was not corrected for. Suppose for example that an area
contained a large amount of developable acreage outside the major project in
year t+0. Even in an area favorable to new development, little if any may
have occurred between the year t+0 and 1970 if the major project grew sub-
stantially in this time period. In other words, the area had a certain fixed
development potential in residential or office land use for the period t+0
to 1970. Any group considering development in this area in the year t+0
probably took account of the effect of planned development announced for the
3-9
-------
already existing major project on this .potential market. Therefore, a new
vacant acreage variable VAC2 was defined in which the proportion of major
project land developed between the year t+0 and 1970 is subtracted out from
non-major project developable acreage to correct for this effect. This new
variable was defined as:
vflrArR MPR70-MPRT2 .
VACACR -- - mpdev
in the Residential model
VAC2 =
unrnrn MPE70-MPET2 mn,m,
VACACR - — MP'FVQ - mpdev
in the Industrial/Office model
The variable VAC2 was included in the predictive equations for RES and
OFFICE.
One other new vacant acreage variable VACS was defined that took
account of land area zoned for other than residential, commercial, etc.,
uses. This variable was defined as:
VAC3 = 10,000-mpdev-mpund-(100 zother)
where: zother = percent of total acreage in the area of influence
zoned for other than residential, commercdal,
industrial, or office use in year t+5
(card input 4, columns 56-58)
Note that the variable vacund was not included in this equation because of
possible overlap between land areas classified as undevelopable and zoned
for other uses. The variables VACS was included in the predictive equations
for RES and OFFICE.
Two new variables were developed to represent the size of a
major project differently from the MPR and MPE variables used in the causal
analysis. The first of these, DENSE, measures the housing density of resi-
dential major projects •, and is defined as:
DENSE = MPR70/mpdev
3-10
-------
The second variable, MPACRE, measures the land area of the major projects,
and is defined as:
MPACRE = mpdev
The variable DENSE was included in all predictive equations in the Residential
model. The variable MPACRE was included in all predictive equations. One
additional new variable MPTIME was developed to represent the relative amount
of time a major project had to induce development before 1970. This variable
was defined as:
MPTIME = 1970 - base yr
where: base yr = the base year t when the major project
first opened (input card 1, columns 13-14)
The variable MPTIME was included in all predictive equations.
The variables MANF and OFFICE were included in the model equations
for RES and COMM to represent development demand associated with growth
in manufacturing and office employment. Unfortunately these variables
did not perform very well in the causal analysis. Therefore, two exogenous
variables, MANACR and OFFACR, representing the amount of manufacturing and
office employment, respectively, in the area of influence in the year t+0
were included in the predictive equations for RES and COMM, and additionally
in the HOTEL equations.
The equations for HOTEL, HOSPTL, CULTUR, and REC had fairly low R2
values in the causal analysis, suggesting that the variables included in
the equations did not represent the principal causes of these forms of
development. Therefore, additional relationships were hypothesized and
included in the predictive equations for these categories of land use. The
new variables are defined below:
DISAIR = Highway distance in miles to nearest major airport in
year t+0 (input card 4, columns 29-32)
AIRPRT = 1 if an airport existed within 3.23 miles of the center
of the major project in the year t+0
0 otherwise (input card 3, column 66)
3-11
-------
PVTSCH = 1 if a private, school existed within 3.23 miles of the
center of the major project in the year t+0
0 otherwise (input card 3, column 64)
UNIV = 1 if a university existed within 3.23 miles of the center
of the major project in the year t+0
0 otherwise (input card 3, column 62)
WATER = 1 if a five square mile body of water existed within 3.23
miles of the center of the major project in the year t+0
0 otherwise (input card 3, column 70)
COAST = 1 if a sea coast existed within 3.23 miles of the center
of the major project in the year t+0
0 otherwise (input card 3, column 68)
The variables DISAIR, AIRPRT, UNIV and MINC were added to the HOTEL equations.
Airport accessibility and area income level can both directly effect hotel
development. The presence of a university can also effect hotel development,
but indirectly through the accommodation requirements of families visiting
students and professionals attending conferences. The variable UNIV was
added to the HOSPTL equation since the presence of a nearby medical school
favors the creation of research and hospital medical facilities. The vari-
ables PVTSCH, UNIV and MINC were added to the CULTUR equations. The presence
of a private school or university, and the area income level can all directly
effect the development of cultural activities. The variables WATER and COAST
were added to the REC equations to account for the presence of these major
recreation-attracting natural features.
One type of data gathered in the data collection phase of this study
that was not utilized in the causal analysis was land use zoning classifi-
cations. New variables representing the proportion of acreage zoned in
various categories were defined as follows:
ZRES = percent of total acreage in the area of influence zoned for
residential use in the year t+5
ZRES = zressf + zresmf
where: zressf = percent of total acreage in the area of influence
zoned for single-family residential use in the year
t+5 (input card 4, columns 45-47)
3-12
-------
zresmf = percent of total acreage in the area of influence
zoned for multi-family residential use in the
year t+5 (input card 4, columns 48-49)
ZCOMM = percent of total acreage in the area of influence zoned
for commercial use in the year t+5 (input card 4, columns
50-51)
ZOFF = percent of total acreage in the area of influence zoned
for office use in the year t+5 (input card 4, columns 52-
53)
ZIND = percent of total acreage in the area, of influence zoned for
industrial use in the year t+5 (input card 4, columns .54-
55)
The variables ZRES, ZCOMM, and ZOFF were included in the predictive equations
for RES, COMM, and OFFICE, respectively. The variable ZIND was included in
the predictive equations for MANF and WHOLE. The rationale in each case
was that development is directly controlled by the zoning classifications
set in each area.
In many areas of the country, the construction of public sewers by
cities and towns is used like zoning to guide land use development. There-
fore, a new variable SEWER was defined as follows:
SEWER = percentage of land in the nearest municipality which had
public sewerage available in the year t+5. (input card
4, columns 42-44)
The variable SEWER was included in all of the predictive equations.
C. DISCUSSION OF RESULTS
The predictive equations obtained by applying the previously dis-
cussed objective criteria to the stepwise regression analyses are summarized
in Section III.D. Summary statistics for these equations are shown in
Table 3-3. The number of predictors in each equation varies from one to six
with an average of from three to four. R^ values indicate the predictive
equations are explaining the majority of variance in the dependent variables.
The mean value for this statistic of 0.54 can be considered quite good in
3-13
-------
view of the fact that the regressions were performed on a relatively small
(20) sample of cross - sectional data. The overall F statistics indicate
practically all of the predictive equations are significant at or below the
one percent level. The results for the coefficient of variation, however,
are less encouraging. The values of this statistic range from 0.34 to 1.73
with a mean of 0.87. Since this statistic .expresses the standard error of
estimate of the regression relationship as a percentage of the dependent
variable mean value, a value as close as zero as possible is desirable.
The summary statistics indicate that the average error encountered in the
use of these predictive equations will be +_ 87 percent of the predicted
value. An examination of the statistics in Table 3-3 show that high values
2
for the coefficient of variation often occur when the R „ of the regression
o
is low, as would be expected.
In an attempt to reduce the coefficients of variation for some of the
predictive equations, the dependent variables RES, COMM, OFFICE and MANF
were defined in a second manner which did not exclude the dwelling units or
land use of the major project from the variables. In the Residential model,
the major project represents residential land use and so the only new depen-
dent variable tested was RES* defined as:
RES* = total dwelling units in the area of influence in 1970.
(input card 5, columns 49-53).
In the Industrial/Office model, three new dependent variables were created,
defined as:
OFFICE* = total office floor area in the area of influence in 1970
in 1,000 ft*.
MANF* = total manufacturing floor area in the area of influence
in 1970 in 1,000 ft2.
WHOLE* = total wholesale and warehouse floor area in the area of
influence in 1970 in 1,000 ft .
3-14
-------
TABLE 3-3
SUMMARY STATISTICS OF PREDICTIVE EQUATIONS
Dependent Number of Sample
Variable Predictors R2 R?=
a
Residential Model
RES
CO MM
OFFICE
MANF
HWLMNX
WHOLE
HOTEL
HOSPTL
CULTUR
CHURCH
EDUC
REC
Industrial/Office
RES
COMM
OFFICE
MANF
HWLMNX
WHOLE
HOTEL
HOSPTL
CULTUR
CHURCH
EDUC
REC
* A sianificance
6
6
5
3
5
5
3
2
2
2
4
1
Model
5
4
5
3
4
4
3
3
4
4
3
6
level of
0.72
0.82
0.81
0.30
0.69
0.79
0.70
0.38
0.49
0.41
0.65
0.43
0.82
0.78
0.66
0.47
0.65
0.77
Q.46
0.51
0.43
0.48
0.46
0.75
0.01 in
0.62
0.76
0.76
0.22
0.61
0.73
0.66
0.34
0.47
0.38
0.58
0.43
0.77
0.73
0.57
0.41
0.59
0.73
0.40
0.46
0.32
0.38
0.39
0.66
dicates tf
Significance Level Coefficient
of Overall F Statistic* of Variation
0.005
0.001
0.001
0.15
0.005
0.001
0.001
0.025
0.005
0.025
0.005
0.005
0.001
0.001
0.01
0.025
0.005
0.01
0.025
0.01
0.10
0.05
0.025
0.005
lere is^at most
0.81
0.49
0.67
1.44
0.74
0.79
0.54
1.34
0.94
0.91
0.58
1.63
0.34
0.48
0.69
0.84
0.47
0.95
1.04
1.73
1.31
0.73
0.60
0.88
a one percent chance
(using the two-tail test) that the population R^ for the regression equation
is 0.
3-15
-------
where: OFFICE* = OFFICE +43.56 mpdev (percent OFFICE +
percent R&D)/100
MANF + 43.56 mpdev (percent MANF/100)
WHOLE + 43.56 mpdev (percent WHOLE/100)
MANF*
WHOLE*
and where,
Percent R&D = percent of the developed land in the major
project used for research and development
purposes.
Percent OFFICE = percent of the developed land in the major
project used for office purposes
(input card 1, columns 73-73)
Percent MANF
Percent WHOLE
mpdev
percent of the developed land in the major
project used for manufacturing purposes
(input card 1, columns 67-69)
percent of the developed land in the major
project used for wholesale and warehousing
purposes (input card 1, columns 76-78)
total developed land area of the major project
in acres
Due to the exclusion of major project land use from the dependent
variables RES, COMM, OFFICE, and MANF, these variables experienced a large
variance in value with zero values occurring a significant percentage of
the time, indicating in these cases that the major project contained all
of a certain type of land use in the area of influence. By including the
major project land use in the variables being predicted, we had hoped to reduce
the resultant coefficients of variation in two ways. First, the effect
of including major project land use was to eliminate zero values and so
significantly reduce the variance that needed to be explained in the regres-
sion relationships, and secondly, this inclusion also raised the mean value
of these dependent variables. Predictive equations for the four total land
use variables RES* COMM* OFFICE* and MANF* were obtained by applying the
techniques used discussed in Section III.A.
3-16
-------
A comparison of the summary statistics for the predictive equations
in which land use of the major project is included or excluded from the
dependent variable is shown in Table 3-4. The results show that for two
of the four modified variables, RES* and WHOLE*, lower coefficients of vari-
ation are obtained. An examination of the regression statistics, however,
reveals that most of this change is due solely to the increase in the mean
value of the dependent variable, and not due to lower standard errors of
estimate. In addition, three of the four overall F statistics indicate less
significance (higher percent level) for the "*" regressions, and in general,
2
the R values are lower. Based on these results, therefore, use of the
a
original predictive equations excluding major project land use are recom-
mended. The equations for RES*, OFFICE*, MANF*, and WHOLE* are, however,
summarized in the following Section III.D, along with the equations for the
other dependent variables, for reference purposes.
2
Due to the relatively large coefficients of variation and low R
a
values for some of the predictive equations (e.g., the MANF equation in the
Residential model and the CULTUR equation in the Industrial/Office model), it
may be tempting for the user to selectively substitute the equation forms
developed in the causal analysis. Examination of the statistics for these
equations, however, show them to be no more significant. Due to the fact
that the predictive equations have the smallest possible variance of all
linear estimators and only they produce unbiased estimates of the dependent
variables, the causal equations should not be used for predictive purposes.
Use of the causal equations would also probably result in cumulative errors of
estimation from the use of endogenous variables as predictors.
The predictive equations shown below in Section I.D, constitute a
set of equations applicable to areas where a major project (Residential or
Industrial/Office) will be or already has been-built. They do not constitute
a general land use predictive model. The data values that will be used for the
-3-17
-------
u>
I
00
TABLE 3-4
COMPARISON OF SUMMARY STATISTICS FOR PREDICTIVE EQUATIONS IN WHICH LAND USE
OF THE MAJOR PROJECT IS INCLUDED OR EXCLUDED FROM THE DEPENDENT VARIABLE
Dependent
Variable +
Residential Model
RES*
RES
Industrial /Off ice Model
OFFICE*
OFFICE
MANF*
MANF
WHOLE*
WHOLE
Number of
Predictors R
5
6
2
5
4
3
5
4
0
0
0
0
0
0
0
0
Sample
2 R2a
.66
.72
.35
.66
.46
.47
.87
.77
0.
0.
0.
0.
0.
0.
0.
0.
60
62
32
57
35
41
83
73
Significance Level
of Overall F Statistic
0.
0.
0.
0.
0.
0.
0.
0.
01
005
025
01
05
025
001
001
Coefficient of
Variation
0.
0.
1.
0.
0.
0.
0.
0.
47
81
87
69
73
84
68
95
A suffix of "*" on a variable indicates that it includes the floor area of the major project.
-------
major project variables MPR70 and MPE70 where they appear in the predictive
equations will, therefore, correspond to the projected final size of the
major project in the area of influence. The data values for MPET2 and MPR68
will correspond to the size of the major project two years after initiation
and two years before completion, respectively. Examination of the predictive
equations reveals that the major project variables do not appear as often
as one would expect, based upon the relationships verified in the causal
analysis. This fact indicates that for some types of land use, just the
presence of a large major project, and not necessarily its size, induces a
certain amount of land use development. The predictive equations developed
in this study are based upon data collected in areas containing a large major
project, viz., one containing several thousand dwelling units (Residential
Model) or employing several thousand employees (Industrial/Office Model).
Specifically, based upon the mean and standard deviations of the variables
MPR70 and MPE70 representing final major project size:
• A Residential project should be in the range of
1,100-5,300 total dwelling units, and
• An Industrial/Office project should be in the range
of 3,600-9,100 employees.
The use of these equations should be limited to situations in which the major
project is in this size range. Also, any application of the predictive
equations should be qualified by the error range indicated by the coefficients
of variation shown in Table 3-3.
D. SUMMARY OF PREDICTIVE EQUATIONS
1. English Units
a. Residential Model
RES = 8,910 DUACRE + 6,790 DELP2 - 351 DISCED - 1,360 HWYINT
+41.2 SEWER - 0.682 MPR70 + 7,200
RES* = 8,880 DUACRE + 6,110 DELP2 - 343 DISCED - 1,346 HWYINT
+ 42.5 SEWER + 8,270
COMM = 791 OFFACR - 73.7 DISCBD + 656 DUACRE - 200 HWYINT
+ 0.00327 EMP60 + 0.0647 VACACR + 1,380
OFFICE = 845 OFFACR + 601 DELEMP - 14.1 DISCBD - 400 DUACRE
+ 85.8 HWYINT +355
3-19
-------
MANF = 1,050 MANACR + 761
HWLMNX = - 2.79 DISCED + 40.5 DELP2 - 135 AUT02 + 46.6 MINC
+0.00595 MPR68 + 78.8
WHOLE = 97.1 HWYINT - 0.0736 MPR70 - 269 DUACRE - 11.4 DISCED
+15.0 OFFVAC + 488
HOTEL = - 0.968 ZRES + 230 AUTO - 150 INCMPL +81.7
HOSPTL = 191 OFFACR + 0.0196 MPR70 - 23.9
CULTUR = 60.2 UNIV + 0.00175 VACACR + 2.54
CHURCH = 202 MINC - 4.42 DISCED - 14.7
EDUC = 0.0408 VACACR + 2.46 SEWER - 25.5 DISCED + 184 MPKIDS + 244
REC = 0.103 MPACRE - 33.5
b. Industrial/Office Model
RES = 2,480 DUACRE + 205 VACACR + 563 OFFVAC - 128,000 VACHSG
- 406 DISCED - 9.530
COMM = 869 DUACRE +119 ZCOMM - 2,090 OFFACR + 0.0553 MPE70 - 838
OFFICE = 11.5 RRMI + 68.9 ZOFF - 0.0273 MPE70 + 507 MINC + 254 MANACR .
- 326
OFFICE*= 13.2 MPACRE +8,120 DELPOP - 2,540
MANF = 10,100 DELEMP - 2,620 MINCC + 0.252 VACACR + 1,120
MANF* = 0.911 MPE70 - 7,010 DUACRE + 466 RRMI - 417 DISCED + 9,430
HWLMNX = 0.00385 VACACR - 6.08 HWYINT + 19.6 DUACRE + 1.70 OFFVAC^
- 25.7
WHOLE = 7,470 WWEA +90.8 ZIND +11.4 SEWER + 726 DUACRE - 1,650
WHOLE* = 111 ZIND + 955 DUACRE + 2,420 HWYLIM - 435 HWYINT
- 0.00227 EMP60 - 1,240
HOTEL = - 11.5 DISCED + 1,180 DELEMP - 249 OFFACR + 145
HOSPTL = 478 NONHSE + 443 MANACR - 283 OFFACR - 20.5
CULTUR = - 34.6 ENERGY + 0.00004 EMP60 + 0.0411 MPACRE + 12.6 PVTSCH
+ 23.6
CHURCH = 8.65 RRMI + 0.0314 VACACR - 1.07 SEWER - 0.0146 MPET2 - 148
EDUC = 0.0802 MPET2 + 0.0974 VACACR + 34.8 RRMI - 925
REC = 17.6 OFFVAC + 1,440 DELEMP + 387 MINCC - 615 AUTO + 14.8 DISCED
+ 188 MANACR - 604
3-20
-------
2. Metric Units
This section summarizes the predictive equations for variables
defined in metric units. Specifically m2 replaces 1,000 ft2 in the variables
RES, COMM, OFFICE, MANF, WHOLE, HOTEL, HOSPTL, CULTUR, CHURCH, and EDUC.
The variables REC, MPACRE, VACACR, and VACS also use m2, but in place of
acres. Distances in miles in HWLMNX, DISCBD, and RRMI are converted to km.
-1 2
And finally, the units of acre are replaced by m in the variables DUACR
OFFACR, WWEA, EMPACR, NONHSE, ENRACR, MANACR, DELEMP, AUTO and DELP2.
RES
RES* =
COMM =
OFFICE =
MANF =
HWLMNX =
WHOLE =
HOTEL =
HOSPTL =
CULTUR =
CHURCH'=
EDUC =
REC
The conversion factors used were:
1,000 f
acres
miles
acre
t2-
->
->•
->•
m2
m2
km
m"2
*by
*by
*by
vby
92.90
4,047
1.609
4,047
a. Residential Model
3,350,000,000 DUACRE + 255,000,000 DELP2 - 20,300 DISCBD
- 126,000 HWYINT+ 3,830 SEWER - 63.4 MPR70 + 669,000
3,340,000,000 DUACRE + 2,297,000,000 DELP2 - 19,800 DISCBD
- 125,000 HWYINT + 768,000
297,000,000 OFFACR - 4,260 DISCBD + 247,000,000 DUACRE
- 18,600 HWYINT +0.304 EMP60 + 0.00149 VACACR + 128,000
318,000,000 OFFACR + 226,000,000 DELEMP - 814 DISCBD
- 150,000,000 DUACRE + 7,970 HWYINT + 33,000
395,000,000 MANACR +70,700
- 2.79 DISCBD + 264,000 DELP2 - 217 AUT02 + 75.0 MINC
+ 0.00957 MPR68 + 127
9,020 HWYINT - 6.84 MPR70 - 101,000,000 DUACRE - 658 DISCBD
+ 1,390 OFFVAC + 45,300
- 89.9 ZRES + 86,500,000 AUTO - 13,900 INCMPL + 7,590
71,800,000 OFFACR + 1.82 MPR70 - 2,220
5,590 UNIV + 0.0000402 VACACR + 236
18,800 MINC - 255 DISCBD - 1,370
0.000937 VACACR + 229 SEWER - 1,470 DISCBD + 17,100 MPKIDS
+ 22,700
417 MPACRE - 136,000
3-21
-------
b. Industrial/Office Model
RES = 932,000,000 DUACRE + 0.0471 VACAGR + 52,300 OFFVAC
- 11,900,000 VACHSG - 23,400 DISCED - 885,000
COMM = 327,000,000 DUACRE + 11,100 ZCOMM - 786,000,000 OFFACR
+5.14 MPE70 - 77,900
OFFICE = 664 RRMI + 6,400 ZOFF - 2.54 MPE70 + 47,100 MINC
+ 95,500,000 MANACR - 30,300
OFFICE*= 0.302 MPACRE + 754,000 DELPOP - 236,000
MANF = 3,800,000,000 DELEMP - 243,000 MINCC + 0.00579 VACACR + 104,000
MANF* = 84.6 MPE70 - 2,640,000,000 DUACRE + 26,900 RRMI
- 24,100 DISCED + 876,000
HWLMNX = 0.00000153 VACACR - 9.78 HWYINT + 128,000 DUACRE + 2.74 OFFVAC
- 41.4
WHOLE = 2,810,000,000 WWEA + 8,440 ZIND + 1,060 SEWER + 273,000,000 DUACRE
- 153,000
WHOLE* = 10,300 ZIND + 359,000,000 DUACRE + 225,000 HWYLIM
- 40,400 HWYINT - 0.211 EMP60 - 115,000
HOTEL = - 664 DISCBD + 444,000,000 DELEMP - 93,600,000 OFFACR + 13,500
HOSPTL = 180,000,000 NONHSE + 167,000,000 MANACR - 106,000,000 OFFACR
- 1,900
CULTUR = - 3,210 ENERGY + 0.00372 EMP60 + 0.000944 MPACRE + 1,170 PVISCH
+ 2,190
CHURCH = 499 RRMI + 0.000721 VACACR - 99.4 SEWER - 1.36 MPET2 - 13,700
EDUC = 7.45 MPET2 + 0.00224 VACACR + 2,010 RRMI - 85,900
REC = 71,200 OFFVAC + 23,600,000,000 DELEMP + 1,570,000 MINCC
- 10,100,000,000 AUTO + 37,200 DISCBD + 3,080,000,000 MANACR
- 2,440,000
E. DISAGGREGATION AND AGGREGATION OF LAND USE CATEGORIES
As part of the transformation of collected field data to the
desired variables in the causal analysis, land use in many subcategories
representing different size ranges were aggregated to form the endogenous
variables. For example the endogenous variable OFFICE is defined* as:
OFFICE = office land use in area of influence (excluding major
project) in 1970 in 1,000 square feet
*See Table 2-1.
3-22
-------
OFFICE = (off! + off2 + off3)/10
where: offl = 100 square feet office in area of influence
(excluding major project) in 1970 (<50K)
off2 = .100 square feet office in area of influence
(excluding major project) in 1970 (50-100K)
off3 = 100 square feet office in area of influence
(excluding major project) in 1970 (>100K)
Once predictive equations were developed for the 12 categories of total land
use, our attention turned to the possibility of predicting land use at a
finer level of detail, where, in addition to the type of land use, the size
range of development was used to categorize the land use being predicted.
In other words, we attempted to predict land use at the disaggregated level
of offl, off2, and off3, for example.
Of the 12 types of land use analyzed, data were available to disag^
gregate six of these, corresponding to the variable RES, COMM, OFFICE, HOTEL,
HOSPTL, and EDUC. Due to the overall lower R2 values of the predictive equa-
a
tions for HOTEL, HOSPTL, and EDUC (see Table 3-3), disaggregation predictions
were not attempted for these variables. In addition, the subcategories of
OFFICE were not analyzed due to the results from Volume II [7] indicating that
energy consumption per floor area (and hence emission rates) are not signi-
ficantly different for office buildings in different size ranges. Disaggre-
gation analysis was, therefore, performed only for the remaining variables
RES and COMM.
Residential land use (dwelling units) was disaggregated into the
subcategories.
• Single Family Detached
• Single Family Attached
• Mobile Home
• Multifamily low rise
• Multifamily high rise
3-23
-------
Commercial land use (1,000 ft2) was disaggregated into the subcategories of:
• < 50,000 ft2
• 50-100,000 ft2
• > 100,000 ft2
In order to avoid the problem of having the predicted land use in the sub-
categories not add up to the total land use, variables representing the
percent of land use of a certain type in'each subcategory were used. The
idea was to use the previously discussed predictive equations to project
| total land use of a certain type and then use the disaggregation equations
to predict the percentage breakdown of the total land use among the various
subcategories. The variables analyzed were defined as follows:
Percent COMM1 *"| the percent of total commercial land use in
Percent COMM2 r the area of influence in 1970 occurring in the
Percent COMM3 J subcategories of <50,000 ft2, 50-100,000 ft2, and
>100,000 ft2, respectively.
Percent RESSFD
Percent RESSFA
Percent RESMO
Percent RESML
Percent RESMH
The percent of total census tract residential
dwelling units in 1970 related to single family
^ detached and attached homes, mobile homes, multi-
family low rise structures, and multi-family high
rise structures, respectively.
Stepwise regressions were performed to develop predictive equations
for the subcategory percentage variables. The objective statistical criteria
discussed previously (Section III.A) were used to evaluate the results.
Regressions were performed with percent COMM1, percent COMM2 and percent COMM3
as the dependent variables and variables representing median income level
(MINC), distance from the central business district (DISCBD), the amount of
land zoned for commercial use (ZCOMM), the amount of land with access to a
sewer system (SEWER) and the commercial employment per acre (COMEMP) as the
3-24
-------
independent variables. For the dependent variables percent RESSFD, percent
RESSFA, percent RESMO, percent RESML, and percent RESMH, the independent
variables in the regressions represented median income level (MINC), distance
to the central business district (DISCED), residential density (DUACRE),
amount of land with access to a sewer system (SEWER) and the amount of land
zoned for single family and multifamily residential development (ZRESSF and
ZRESMF, respectively).
An examination of the summary statistics of the predictive equations
for the disaggregation variables indicated that no statistically significant
predictors could be found in about half of regressions. In the regressions
where significant predictors were chosen, the relationships developed were
counter to what one would reasonably expect. For example, MINC (median .income
level) was found to be negatively related to percent RESSF (the percent of
single family structures in total residential land use). Due to the poor
results obtained, average percentage figures for each subcategory were comr
puted; these are summarized in Table 3-5. Since these data are representative
of the year 1970, and not future years, it is recommended that a projected
percentage split for the land use being predicted be developed in each par-
ticular case, taking account of both the area and future time period involved.
It is doubtful that the average percentages in Table 3-5 take into account
the parameters or effects that will guide the densities of commercial and resi-
dential development in future years. In the absence of other data, however,
the average percentages shown here do provide an estimation of the possible
disaggregation of land use development.
In addition to attempting to disaggregate the 12 categories of land
use for which predictive equations were developed, these land use variables
were aggregated to form a total land use variable TOTUSE, defined as:
TOTUSE = total developed land use in the area of influence (incl
ing major project) in 1970 in 1,000 ft2.
ud-
3-25
-------
TABLE 3-5
AVERAGE PERCENTAGE VALUES FOR
DISAGGREGATED LAND USE VARIABLES
Disaggregation
Variable
Subcategory
Description
Major Project Type
Residential Industrial/Office
Percent COMM1
Percent COMM2
Percent COMM3
Commercial
30,000 ft2
50-100,000 ft2
100,000 ft2
51%
20%
29%
100%
66%
14%
20%
100%
Percent RESSFD
Pervent RESSFA
Percent RESMO
Pervent RESML
Percent RESMH
Residential
single family detached 61%
single family attached 3%
mobile home 6%
multi family low rise 29%
multifamily high rise 1%
100%
68%
2%
6%
23%
1%
100%
3-26
-------
TOTUSE = RES* ((1.6 Percent SF/100) + (0.9 Percent MF/100))
+ COMM + OFFICE* + MANF* + WHOLE* + HOTEL +HOSPTL
+ CULTUR + CHURCH + EDUC + 43.56 REC
where:
Percent SF = percent of total residential dwelling units in the
area of influence in 1970 related to single family
and mobile homes.
Percent MF = percent of total residential dwelling units in the area
of influence in 1970 related to multifamily structures.
1.6 = Average 1,000 ft2 of floor area for a single family
dwelling unit
0.9 = Average 1,000 ft2 of floor area for a multifamily
dwelling unit
43.56 = Conversion factor from acres to 1,000 ft2
Percent SF = resl/restot
Percent MF = (resl + res2 + res3 + res5 + res50)/restot
where:
resl = single family home dwelling units in census tracts in
1970 (input card 5, columns 68-73)
res2 = two-family structure dwelling units in census tracts
in 1970 (input card 5, columns 74-78)
res3 = three and four-family structure dwelling units in
census tracts in 1970 (input card 6, columns 4-8)
res5 = five through 49-family structure dwelling units in
census tracts in 1970 (input card 6, columns 9-13)
res50 = 50-and greater family structure dwelling units in
census tracts in 1970 (input card 6, columns 14-18)
restot = total year-round dwelling units in census tracts in
1970 (input card 5, columns 62-67)
Stepwise regressions were performed in which TOTUSE was the depen-
dent variable and a composite set of all independent variables used in the
predictive equation analysis was used for the independent variables. The
regression relationships developed and the associated statistics are sum-
marized in Table 3-6. The results indicate that when the major project type
3-27
-------
TABLE 3-6
AGGREGATED TOTAL LAND USE PREDICTIVE EQUATIONS
AND SUMMARY STATISTICS
Residential Model
TOTUSE = 2.42 MPACRE + 2,700 ZCOMM + 12,100 ZOFF +
638 DISAIR - 23,100 MINC - 8,390 DUACRE +
4,120
TOTUSE = 0.0556 MPACRE + 251,000 ZCOMM + 1,120,000 ZOFF +
36,800 DISAIR - 2,150,000 MINC -
3,150,000,000 DUACRE + 383,000
2
R = 0.73 Significance level of overall F statistic = 0.005
R2 = 0.64 Coefficient of variation = 0.96
a
Industrial/Office Model
(English TOTUSE = 31.2 MPACRE + 8,650 OFFACR +2.06 MPE68 +
Units) 28,400 MINCR - 1.85 MPET2 - 25,600
(Metric TOTUSE = 0.716 MPACRE + 3,250,000,000 OFFACR + 191 MPE68 +
Units) 2,640,000 MINCR - 172 MPET2 - 2,380,000
R2 = 0.84 Significance level of overall F statistic = 0.001
D2 _ n on Coefficient of variation = 0.27
K - U.oU
a
3-28
-------
is Residential, there is a certain advantage to predicting total land use
using an aggregated predictive equation, as opposed to summing together the
predicted levels of 12 individual predictive equations and a projection of
the major project size. Although the coefficient of variation for the TOT-
USE predictive equation in this case, 0.96, is slightly higher (more error)
than the average coefficient of variation for the 12 predictive equations
n
(see Table 3-3) of 0.91, the R = value of 0.64 is also higher (more explained
a
variance) than the 12 equation average of 0.55. For the Industrial/Office
model, the advantages are more clearcut in using the TOTUSE predictive
equation. The aggregated predictive equation coefficient of variation is
p
only 0.27, compared to a 12 equation average of 0.83, and the R value is
a
0.80 compared to a 12 equation appropriate average of 0.53. Therefore, the
use of the TOTUSE predictive equation is recommended whenever total land
use projections are desired.
F. CROSS-VALIDATION ANALYSIS
The usefulness of any set of predictive equations depends upon their
generality. Only if a regression equation is based upon a data sample which
is representative of the general data population can it provide accurate pre-
dictions in different situations. In the current study, where only a small
data sample (20) was available for the development predictive land use
equations, the question of validity was especially important. The preferred
test of an equation's validity is an external validation, viz., applying it.
on a test case basis to an independent sample of data (i.e., independent of
the sample on which the coefficient values were based) and observing its
predictive ability. In the current study a separate, independent sample was
not available. Therefore, the analytical technique of cross-validation was
used to simulate the existence of such a test sample.
The procedure of cross-validation involved splitting both the
Residential and Industrial/Office data samples of 20 into two random samples
of 10 each. In the Residential model, the first sample of 10 consisted of
case numbers* 4,6,7,10,15,16,19,20,25, and 36; the second sample of 10
* See Table C-l in the separately published Appendix C to Volume I [6] for
a definition of case numbers.
3-29
-------
consisted of case numbers 14,21,22,23,27,29,30,35,37 and 40. In the Indus-
trial/Office model the first sample of 10 consisted of case numbers 1,2,3, 5,
9,11,12,13,26 and 34; the second sample of 10 consisted of case numbers
8,17,18,24,28,31,32,33,38 and 39. In each model, the coefficient values
in the predictive equations were recomputed using the first data sample of
10 and the set of independent variables decided upon previously. These
coefficient values, specifying 12 predictive equations in each model, were
used in conjunction with the data for the independent variables from the
second sample of 10 to predict the values of the dependent variables in the
second sample of 10. Statistical comparisons were then made between actual
and predicted values for the dependent variables in the second sample of 10.
Unlike the jackknifing technique used in the causal analysis to
examine the stability of individual model coefficients, cross-validation pro-
vided a measure of the overall validity of the predictive equations..
The correlation coefficients (R) between actual and predicted values are shown
in Table 3-7 and graphs of the actual versus fitted (predicted) values are
summarized in Appendix B. In this application, the R statistic represents a
coefficient of validity of the predictive ability of each equation. Values
in Table 3-7 exceeding 0.63 are statistically significant, at the five
percent confidence level (two-tailed test). The results of the cross-vali-
dation indicate that about half of the 24 predictive equations are generalized
enough to produce good predictions using an independent sample. The COMM,
WHOLE, CHURCH, EDUC, and REC equations validated well in both models, as well
as the RES and HWLMNX equations in the Industrial/Office model. Considering
the extremely small sample size (10) used in the cross-validation and the large
amounts of variance in the dependent variables, it was encouraging to obtain
significant correlations between actual and predicted values in as many
equations as we did.
Although coefficients of validity >0.63 indicate equations with
good predictive ability, coefficients <_ 0.63 do not necessarily indicate the
associated predictive equation is useless. The poor correlations obtained
3-30
-------
TABLE 3-7
COEFFICIENTS OF VALIDITY BETWEEN ACTUAL AND PREDICTED LAND USE FROM THE CROSS-
VALIDATION ANALYSIS OF ALL LAND USE PREDICTIVE EQUATIONS
Independent Variable
in Equation
Residential Model
RES
COMM
OFFICE
MAIMF
HWLMNX
WHOLE
HOTEL
HOSPTL
CULTUR
CHURCH
EDUC
REC
Industrial/Office Model
RES
COMM
OFFICE
MANF
HWLMNX
WHOLE
HOTEL
HOSPTL
CULTUR
CHURCH
EDUC
REC
Coefficient of
Validity (R)
-0.11
0.85
0.07
0.43
0.20
0.73
0.46
0.44
0.35
0.69
0.80
0.74
0.81
0.77
-0.61
0.25
0.71
0.86
0.46
0.17
-0.12
0.62
0.70
0.67
Statistical
Significance Level
0.005
—
--
—
0.02
—
--
--
0.05
0.01
0.02
0.005
0.001
—
__
0.05
0.005
—
—
—
0.10
0.05
0.05
3-31
-------
could simply be due to nonhomogeneity in the data sample caused by the
extremely small sample size (10). However, without an additional indepen-
dent sample to use as a test case, it was not possible to ascertain if this
was the case, or if the predictive equations simply lacked generality. In
many instances where low R values were obtained, one or two particular
samples consistently had the largest residuals (actual minus predicted),
i.e., were outliers. It can be seen from the graphs of actual and fitted
values (summarized in Appendix B) that in the Residential model, case number
30, and in the Industrial/Office model, case number 8, are recurrent out-
liers. Further work could be done in the validation analysis by assessing the
effect of excluding certain outlier samples on the coefficients of validity
obtained, and examining whether the actual values of such samples are indeed
outliers in the dependent variable populations. Such efforts, however, were
not attempted in the current study since it is quite probable that in a sample
of only 10 that a single outlier may represent a valid 10 percent of the data
population. Such possibilities, again, would be less likely if a larger sample
were available for analysis. Further work could also be done in assessing the
stability of the predicted land use estimates by computing overall confidence
intervals for the predictive equations. All of the necessary data for com-
puting confidence intervals is contained in Appendix A, which summarizes the
statistical output for the predictive land use equations, including the
variance - covariance matrices. Given an equation of the form:
Y = bo + ^ Vi
where:
Y is the dependent variable
X-, Xo, . . .X are the independent variables
confidence intervals can be specified as:
Y + tV(Y)1/2
3- 32
-------
where t is the t-statistic for the regression of the model equation and
V(Y) is the variance of Y. V.(Y) can be expressed as:
n n
V(Y) = I E X-X. Covariance (b,b,)
1=1 j=l T J 1 J
Worksheets for using the predictive equations and computing confidence
intervals are given in Section VII.
3-33
-------
IV. LAND USE BASED EMISSION FACTORS
This chapter discusses the approach used to develop the land use based
emission factors and presents a tabular compilation of the emission factors
that were developed. It is a summary of Volume II of the GEMLUP final
report [7].
A. APPROACH TO LAND USE BASED EMISSION FACTORS
The objective of this phase of the GEMLUP study was to develop a
set of land use based emission factors to permit the estimation of air
pollutant emissions resulting from the construction and operation of a major
land use project. These emission sources may be principally categorized
as follows:
• Stationary source emissions occurring on the site of
the major project (e.g., the on~site combustion of
fuel oil for space heating needs),
• Stationary source emissions occurring at the land use
induced by the major project (e.g., the on-site
combustion of fuel oil for space heating needs),
• Secondary (i.e., occurring off-site) stationary source
emissions (e.g,, the combustion of fuel oil at the
local electric utility to serve the electricity demand
of the major project and induced land uses),
• Mobile source emissions (e.g., emissions due to motor
vehicular traffic generated by the major project and
induced land uses).
The latter category, mobile sources, is treated separately in Chapter VI of
this report.
The estimation of emissions from the first three categories, all
stationary sources, is the subject of this chapter. The means of this
estimation is the use of land use based emission factors, that is, emissions
per unit floor area of a particular land use category. Given the size of the
major project and the amount of floor area, air pollutant emissions may then
be estimated by taking the product of the appropriate land use based emission
factor and the floor area of a particular land use category.
4-1
-------
1. Emission Factor Structure
The land use based emission factors, emissions per unit floor
area, may be disaggregated into two factors, an activity factor (i.e., fuel
throughput, etc., per unit floor area), and the "Standard" emission factor
(i.e., emissions per unit fuel). For example, in the case of fuel oil
space heating consumption, this would be,
emissions, gr. oil consumption, gals. * emissions, gr.
3 3
floor area, 10 sq.ft. floor area, 10 sq;.ft. . oil consumption, gals.
Given this structure, a complete set of land use based emission factors
would consist of an n-dimensional array, with' specific values given for a
pollutant species, fuel or process type,, building category, and, in some
cases, energy requirements (e.g., region of the country).
Ignoring the solvent evaporation, solid waste disposal, and
other miscellaneous emissions,* the energy consumption related emission
factor can be generalized as follows:
emissions. . . Btu. Btu. Btu.
T »J>K _ [(__] + —1 + J )*
sq.ft. year sq.ft. year sq.ft. ht.d.d1. sq.ft. cl.d.d.
, , emissions.
I * 3 * J I
j
heat content. seasonal efficiency. unit fuel.
where ht.d.d. = heating degree days per year
cl.d.d. = cooling degree days per year
and for a particular fuel type i, pollutant species j, and
building category k.
*Emissions fr&m these sources are mot considered in this report, since there
•is both more limited infprmation about their characteristics and that they
may be expected to display more variation in per unit floor area emissions
between parts of the country. However, the emission factor structure dis-
cussed above is amenable to their inclusion. It is recommended that they
be included in areas where there are significant emission sources and/or
better information concerning their characteristics is available.
4-2
-------
The fourth term 1n this equation, emissions per unit fuel, are
the commonly used values determined directly from the EPA Compilation of
Air Pollutant Emission Factors [8]. Hence, the focus of this project is
generating the first three terms (i.e., the activity factor).
The first three terms identify the fuel consumption per build-
ing floor area, given the heating and cooling degree days. The heat content
of a fuel in British thermal units is approximately constant and is well
known [9]. It does display some variation for every fuel, especially for
natural gas in different regions of the country [10].
The values for the efficiency for various building types and
fuels are less well known. Efficiency can be defined in a variety of ways.
The purpose of this application is to account for the differences in the
amount of energy consumed by a building depending on the fuel type selected
to provide that energy.
The desired efficiency measure is the ratio of heat loss from
a structure to the energy input to the structure, variously defined as effi-
ciency of utilization or seasonal efficiency.
The term in brackets, the energy requirement per square foot
and per square foot degree day, represents the energy requirements of a
building. It is divided into three components:
• Process use of energy that is not related to
climate; examples include:
Lighting
Elevators
Refrigeration
Water heating equipment
Cooking equipment
Ventilation
4-3
-------
• Energy requirements for space heating, as a function
of heating degree days*
• Energy requirements for air conditioning, as a function
of cooling degree days.t
In lieu of cooling degree days for residential buildings, this
study will use the estimated compressor operating hours of residential air
conditioning units as compiled by Oak Ridge National Laboratory. A map of
iso-compressor operating hours is shown in Figure 4-3.
2. Variance of Energy Requirement, Efficiency, and Emission
Factors
The energy requirement factor, the efficiency of utilization,
and the standard emission factors are all estimates of the mean of popula-
tion values and can be expected to display a large variation. In general,
these factors are not precise indicators of energy requirement, efficiency,
or emissions of a single source. They are more valid when applied to a
large number of sources.
*Early this century, heating engineers developed the concept of heating
degree days as a useful index of heating fuel requirements. They found
that when the daily mean temperature is lower than 65 degrees, most build-
ings require heat to maintain an inside temperature of 70 degrees. The
daily mean temperature is obtained by adding together the maximum and
minimum temperature reported for the day and dividing the total by two.
Each degree of mean temperature below 65 is counted as one heating degree
day. Thus, if the maximum temperature is 70 degrees and the minimum 52
degrees, four heating degree days would be produced. (70 + 52 = 122,
122 divided by 2 = 61; 65 - 61 = 4). If the daily mean temperature is 65
degrees or higher, the heating degree day total is zero. A map of iso-
heating degree days for the United States is shown in Figure 4-1 [11].
tThe cooling degree day is a mirror image of the heating degree day. After
obtaining the daily mean temperature by adding together the day's high and
low temperatures and dividing the total by two, the base 65 is subtracted
from the resulting figure to determine the cooling degree day total. For
example, a day with a maximum temperature of 82 degrees and a minimum of
60 would produce six cooling degree days. (82 + 60 = 142; 142 divided
by 2 = 71; 71 - 65 = 6). If the daily mean temperature is 65 degrees or
lower, the cooling degree day total is zero [11].
4-4
-------
Figure 4-1: NORMAL SEASONAL HEATING DEGREE DAYS ( BASE 65°F ) 1941-1970
« co^^nvY
• lndi.n«po»- * .' _ , V
-------
->
St. Louis* I / L»Kington
• '
' Honolulu c=± Kihulul
o?CX3732
PM.RTO B1CO HMD V IRC IS 1SLAV.S
Figure 4-2: NORMAL SEASONAL COOLING DEGREE DAYS ( BASE 65°F ) 1941-1970
-------
FIGURE 4-3. ANNUAL AIR CONDITIONER COMPRESSOR-OPERATING HOURS FOR HOMES THAT ARE NOT
NATURALLY VENTILATED. Source: Oak Ridge National Laboratory [35j.
-------
B. COMPILATION OF LAND USE BASED EMISSION FACTORS
This section presents a tabular summary of the land use based
emission factors. The emission factors are presented in units of pounds
of pollutant emitted per "measure" for oil and gas combustion. For
electricity consumption, the factors are in terms of kilowatt-hours per
"measure". The measure, depending on the activity involved, may be per
square foot of building floor area, per square foot heating degree day,
per dwelling unit, etc.
The quantity of secondary, i.e., off-site, emissions occurring due
to electricity consumption depends on the nature of the local electricity
utility generating station. It is suggested that the local utility be con-
tacted to determine the appropriate emission factor. Default values of
pounds of pollutant emissions per kilowatt-hour sold are presented in
Table 4-1 and are based on data in References 8, 12, and 13. It should
also be pointed out that the emissions due to increased electrical demand
do not necessarily occur at the nearest generating plant.
Tables 4-2 through 4-13 present the land use based emission
factors for residential, commercial, institutional, and industrial land
uses. The industrial factors do not include process emissions. Metric
equivalents of these emission factors are given in Volume II of this report
[7].
4-8
-------
TABLE 4 -1
TYPICAL UNCONTROLLED EMISSION FACTORS FOR ELECTRIC UTILITIES
Ibs. emissions per kWh sold to customer
coal
oil
gas
PM
5.23
6.34
1.19
x
x
X
10" 3A
io-4
io-4
SO
x ?
1.53 x 10 S
1.26 x
7.13 x
10"2S
io-6
CO
4.03 x
2.38 x
2.02 x
io-4
io-4
io-4
1
1
1
HC
.21 x
.58 x
.19 x
ID'4
io-4
io-5
NO
x 9
2.21 x 10"*
8.32 x 10"3
8.32 x 10~3
Note: A 33.3% overall plant efficiency is assumed for coal-fired plants [12].
A 31.6% overall plant efficiency is assumed for oil- and gas-fired plants [12].
A 10% transmission loss is assumed [13].
'S1 and 'A1 represent, respectively, the sulfur and ash percentage of fuel
by weight.
-------
TABLE 4-2
SINGLE FAMILY RESIDENTIAL LAND USE BASED EMISSION FACTORS
pound of pollutant (or kilowatt-hours) per measure
PM
SO.
CO
HC
NO
kWh
Measure
Space Heating
Electricity
Gas
Oil
Air Conditioning
Central
Electricity
Gas
Room
Electricity
Process
Hot Water
Electricity
Gas
Oil
Cooking
Electricity
Gas
Miscellaneous
2.2 x 10"
2.5
3.8
2.6 x 10"4 1.5 x 10"5
,-3
5.1 x 10
-4
3.2 x 10"2S 1.1 x 10"3
2.0 x 10"4 2.6 x 10"3
6.6 x 10"4 2.6 x 10"3
4.7
l.SxlO"4 1.1 x 10"5 3.5 x 10"4 1.4xlO~4 l.SxlO"3
dwelling unit^ht.d.d.
dwelling unit-ht.d.d.
dwelling unit-ht.d.d.
dwelling unit-op.hr.
dwelling unit-op.hr.
3.0 x 10"1 l.SxlO"2 6.0 x 10"1 2.4 x 10"1 3.0
3.7 x 10~]S 1.2
7.5 x 10"1 3.0
l.lxlO"1 6.6 x 10"3 2.2 x 10"1 8.8xlO"2 1.1
5.1X10"1 a-c. unit-operating hour
1.4xlO+4 dwelling unit-year
dwelling unit-year
dwelling unit-year
3.5xlO+3 dwelling unit-year
dwelling unit-year
7.9^10 dwelling unit-year
Note: A 1600 square foot dwelling unit is assumed.
'S' represents the sulfur percentage of oil, by weight.
-------
Air Conditioning
Central
Electricity
Room
Electricity
Process
Hot Water
Electricity
Gas
Oil
Cooking
Llectricity
Gas
Miscellaneous
TABLE 4-3
MOBILE HOME RESIDENTIAL LAND USE BASED EMISSION FACTORS
Activity
Space
Heating
Electricity
Gas
Oil
pound of
PM S0x
1.7 x 10~4 9.9 x 10~6
1.4 x 10"3 2.0 x 10"2S
pollutant
CO
3.3 x
6.9 x
ID'4
io-4
(or kilowatt-hours) per
HC NOV
J\
1.3 x IO"4
4.2 x IO"4
1.7 x 10
1.7 x 10
measure
kWh
2.32
-3
-3
dwell
dwell
dwell
Measure
ing unit*
ing unit-
ing unit-
ht.
ht.
ht.
d.i
d.i
d.i
3.4
5.1x10
-1
dwelling unit-op.hr.
a.c. unit-op.hr.
3.0 x IO"1 1.8 x 10"2 6.0 x IO"1 2.4 x IO"1 3.0
2.5
3.6 x 10+1S 1.2 7.5 x IO"1 3.0
l.lxlO"1 6.6 x IO"3 2.2 x IO"1 8.8 x IO"2 1.1
1.3 x 10 dwelling unit-year
dwelling unit-year
dwelling unit.year
3.5 x 10 dwelling unit.year
dwelling unit.year
7.9 x 10+3 dwelling unit.year
Note: A 720 square feet per dwelling unit is assumed.
'S1 represents the sulfur percentage of oil, by weight.
-------
TABLE 4-4
LOW RISE MULTIFAMILY RESIDENTIAL LAND USE BASED EMISSION FACTORS
Activity
PM
pound of pollutant (kilowatt-hours) per measure
SO.
CO
HC
NO.
kWh
Measure
ro
Space Heating
Electricity
Gas
Oil
Air Conditioning
Central •
Electricity
Gas
Oil
Room
Electricity
Process
Hot Water
Electricity
1.3
1.2 x 10
1.1 x 10
-4
-3
7.3 x 10
-6
2.4 x 10
-4
1.7 x 10~2S 5.7 x 10"4
9.7 x 10
3.4 x 10
-5
-4
1.2 x 10
1.4 x 10
-3
-3
1.5
6.2 x 10
4,5 x 10
-5
-4
3.7 x 10
6.4 x 10
-6
-3
1.2 x 10
2.2 x 10
-4
-4
5.0 x 10
1.3 x 10
-5
-4
6.2 x 10
5.3 x 10
-4
-4
5.1x10
-1
dwelling unit«ht.d.d.
dwelling unit«ht.d.d.
dwelling unit-ht.d.d.
dwelling unit« op.hr.
dwelling unit.op.hr.
dwelling unit«op.hr.
a.c. ur.it-op.hr.
1.1 x 10 dwelling unit.year
Gas
Oil
Cooking & Dryer
Electricity
Gas
Miscellaneous
2
2
_
1
-
.4 x 10~'
.0
.2 x 10'1
1
2
_
7
-
.4 x 10^
.9 x 10+1S
.2 x 10~3
4
1
_
2
-
.8 x 10"'
.0
.4 x 10"1
1.9
6.0
_
9.6
-
x 10"'
x 10"1
x 10"2
2.4
2.4
_
1.2
-
dwelling
dwelling
3.8 x 10+3 dwelling
dwelling
4.4 x 10 dwelling
unit.year
unit.year
unit.year
unit- year
unit-year
Note: A 900 square foot dwelling unit is assumed.
'S1 represents the sulfur percentage of oil, by weight.
-------
•fe.
!""•
to
TABLE 4-5
HIGH RISE MULTIFAMILY RESIDENTIAL LAND USE BASED EMISSION FACTORS
pound of pollutant (or kilowatt- hours) per measure
Activity PM SO CO HC NOV kWh
X X
Space Heating
Electricity - - - - - 1.5
Gas l.OxlO"4 6.2 x 10"6 2.1 x 10"4 8.3 xlO"5 l.OxlO"3 -
Oil l.OxlO"3 1.5xlO"2S 5.2 x 10"4 3.1 x 10"4 1.3xlO"3 -
Air Conditioning
Central
Electricity - - - - - 1.5
Room
Electricity - - - _ _ .51
Process
Hot Water
Electricity - - - - - 6.2 x 10+3
Gas 1.4 x 10"1 8.4 x 10"3 2.8 x 10"1 l.lxlO"1 1.4
Oil 1.1 1.6 x 10+1S 5.7 x 10"1 3.4 x 10"1 1.4
Cooking & Dryer
Electricity - 3.8 x 10+3
Gas 1.2.x TO"1 7.2xlO"3 2.4 x 10"1 9.6 x 10"2 1.2
Miscellaneous ----- 5.9 x 10+3
Measure
dwelling unit-ht.d.d.
dwelling unit-ht.d.d.
dwelling un1t*ht.d.d.
dwelling unit»op.hr.
dwelling unit*op.hr.
dwelling unit -year
dwelling unit-year
dwelling unit-year
dwelling unit -year
dwelling unit*year
dwell ing unit*year
Note: A 900 square foot dwelling unit is assumed.
'S' represents the sulfur percentage of oil, by weight.
-------
TABLE 4-6
RETAIL ESTABLISHMENTS, WAREHOUSES, WHOLESALING ESTABLISHMENTS, LAND USE BASED EMISSION FACTORS
Activity
Space Heating
Electricity
Gas
Oil
Air Conditioning
Electricity
Process
Hot Water
Electricity
Gas
Oil
Lighting
Auxiliary
Equipment
Appliances
Refrigeration
pound of pollutant (or kilowatt-hours) per measure
PM SOV CO HC NO- kWh
A A .
1.3 x 10"3
9.8 x 10"8 5.9 xlO"9 2.0 x 10"7 7.8 x 10"8 9.8 x 10"7 -
1.7xlO"6 1.2xlO~5S 2.9 x 10"7 3.3 x 10"5 4.4 x 10"6 -
5.2 x 10"3
5.0 x 10"1
2.4 x 10"5 1.4 x 10"6 4.8 x 10"5 1.9xlO"5 2.4 x 10"4 -
5.2 x 10"4 3.6 x 10" 3S 9.1 x 10"5 l.OxlO"2 1 .4 x 10"3 -
8.0
. , - - - - 3.6
- - - - - 2.0
8.9
Measure
sq.ft.* ht.d.d.
sq. ft. •• ht.d.d.
sq.ft. •ht.d.d.
sq.ft.* cl.d.d.
sq.ft.^year
sq.ft.* year
sq.ft.* year
sq.ft. -year
sq.ft.* year
sq.ft. -year
sq.ft.* year
Note: 'S' represents the sulfur percentage of oil, by weight.
-------
TABLE 4-7
OFFICE BUILDING LAND USE BASED EMISSION FACTORS
Activity
PM
pound of pollutant (or kilowatt-hours) per measure
SO CO HC NO kWh
X "
Measure
Space Heating
Electricity
Gas
Oil
Air Conditioning
Electricity
Gas
Oil
Process
9.4 x 10"8 5.6 x 10"9
1.9 x 10
-7
1.7 x 10"6 1.2 x 10"5S 2.9 x 10"7
7.4 x 10"8 4.4 x 10"9
1.5 x 10
-7
1.3 x 10"6 9.1 x 10"6S 2.3 x 10"7
7.5 x 10"8 9.4 x -10"7
3.3 x 10"5 4.4 x 10"6
5.9 x 10"8 7.4 x 10"7
2.6 x 10"5 3.4 x 10
-6
1.9 x 10"3 sq.ft.-ht.d.d.
sq.ft.*nt.d.d.
sq.ft.'ht.d.d.
1.5 x 10
2.8 x 10
-3
sq.ft.- cl .d.d.
sq.ft.- cl .d.d.
sq. ft., cl .d.d.
sq.ft.' year
Note: 'S' represents the sulfur percentage of oil, by weight.
-------
TABLE 4-8
NONHOUSEKEEPING* RESIDENTIAL LAND USE BASED EMISSION FACTORS
Activity
PM
pound of pollutant (or kilowatt-hours) per measure
SO.
CO
HC
NO.
kWh
Measure
Space Heating
Electricity
Gas 9.4 x 10
Oil 1.4 x TO
Air Conditioning
Electricity
Gas
Oil 4.1 x 10
Process
-8
-6
5.6 x 10
-9
1.9 x 10
-7
9.9.x 10"6S 2.5 x 10"7
7.5 x 10
2.8 x 10
-8
-5
9.4 x 10
2.8 x 10
-7
-5
2;'3 x 10"8 1.4 x 10"9
4.6 x 10
-8
-7
2.8 x 10"6S 7.1 x 10"8
1.8 x 10
8.0 x 10
-8
-6
2.3 x 10
•1.1 x 10
-7
-6
1.7 x 10"3 sq.ft.-ht.d.d.
sq.ft.-ht.d.d.
sq.ft.-ht.d.d.
4.7 x 10"4 sq.ft.-ci.d.d.
sq.ft.-cl ,-d.d.
sq.ft.tcl.d.d.
1.2 x 10+1 sq.ft.-year
* Hotels, Motels, Dormatories, etc.
Note: 'S1 represents the sulfur percentage of oil, by weight.
-------
TADLE 4-9
HOSPITAL LAND USE BASED EMISSION FACTORS
Activity
PM
pound of pollutant (or kilowatt-hours) per measure
SO
CO
HC
NO
kWh
Measure
Space Heating
Electricity
Gas
Oil
Air Conditioning
Electricity
Process
2.2 x 10
"3
1.8 x 10"7 1.1 x 10
"8
3.3 x 10
"6
2.3 x 10"5S
3.7 x 10"7 1.5 x 10"7 1.8 x 10"6
5.8 x 10"7 6.6 x 10"5 8.7 x 10"6
5.9 x 10
"3
sq.ft.-ht.d.d.
sq.ft.-lit.d.d.
sq.ft.«ht.d.d.
sq.ft.«cl.d.d.
1
H-»
-xl
Lighting
Auxiliary
Equipment
Appliances
Hot Water
Electricity -
Gas 2.4 x 10"4 1.4 x 10"5
Oil 5.2 x. 10"3 3.6 x 10"2S
1.5 x 10+1
1.7 x 10+1
5.9
5.0
4.8 x 10"4 1.9 x 10"4 2.4 x 10"3 -
9.1 x 10"4 1.0 x 10"1 1.4 x 10"2 -
sq. ft« year
sq. ft- year
sq.ft.* year
sq.ft. *year
sq.ft. 'year
sq.ft. 'year
Note: 'S' represents the sulfur percentage of oil, by weight.
-------
TABLE 4-10
CULTURAL BUILDING LAND USE BASED EMISSION FACTORS
pound of pollutant (or kilowatt-hours) per measure
Activity PM SOV CO HC N0¥ kWh Measure
/\ /\
Space Heating
Electricity - - -
Gas
Oil
9.0 x
1.6 x
10"8
10"6
5.
1.
4 x 10
1 x 10"5S
1.8 x
2.8 x
IO"7 7.2
10"7 3.2
x 10"8
x IO"5
9.0 x
4.2 x
1.8 x IO"3 sq.ft.
lO"7 -
lO"6 -
sq . ft .
sq.ft.
•ht.d.d.
•ht.d.d.
•ht.d.d.
Air Conditioning
------------ ---^
Electricity - -
Gas
Oil
Process
2.9 x
5.1 x
-
ID'8
10"7
1.7 x
3.6 x
-
lO"9
10"6S
5.7 x
8.9 x
-
io-8
io-8
2.3 x
1.0 x
-
10"8 2.9 x
10"5 1.3 x
-
5.Q x IO"4
lO"7 -
io-6 -
1.2 x 10+1
sq.ft.*cl .d
sq.ft.'cl.d
sq.ft. ••cl.d
sq.ft. -year
.d.
.d.
.d.
Note: 'S' represents the sulfur percentage of oil, by weight.
-------
TABLE 4-11
CHURCH BUILDING LAND USE BASED EMISSION FACTORS
Activity
Space Heating
Electricity
Gas
Oil
Air Conditioning
Electricity
Gas
Oil
Process
PM
_
1.4 x
2.6 x
_
1.8 x
3.3 x
-
pound of pollutant (or kilowatt-hours)
SO CO HC
/\
ID'7
io-6
ID'7
io-6
_
R.6 x
1.8 x
_
1.1 X
2.3 x
-
io-9
10"6S
io-8
10'6S
_
2.9 x
4.5 x
_
3.7 x
5.7 x
-
ID'7
1C'7
io-7
io-7
_
1.1 x IO"7
5.0 x IO"5
_
1.5 x IO"7
6.4 x IO"5
-
per measure
NOV kWh
/\
_
1.4 x
6.7 x
_
1.8 x
8,6 x
-
2.9 x IO"3
10"6 -
10"6 -
3.8 x IO"3
10"6 -
lO"6 -
4.2
Measure
sq.ft
sq.ft
sq.ft
sq.ft
sq.ft
sq.ft
sq.ft
.-ht.d
.•ht.d
.•ht.d
.•cl.d
.•cl.d
.•cl.d
.'year
.d.
,d.
.d.
.d.
.d.
.d.
Note: '$'• represents the sulfur percentage of oil, by weight.
-------
TABLE 4-12
SCHOOL BUILDING LAND USE BASED EMISSION FACTORS
Activity
PM
pound of pollutant (or kilowatt-hours) per measure
SO.
CO
HC
NO.
kWh
Measure
ro
o
Space Heating
Electricity
Gas
Oil
Air Conditioning
Electricity
Gas
Oil
Process
8.0 x 10"8 4.8 x 10"9 1.6 x 10"7 6.4 x 10"8 8.0
1.2 x 10"6 8.5 x 10~6 2.1 x 10"7 2.4 x 10"5 3.2
2.3 x 10"8 1.4 x 10"9 4.6 x 10"8
4.1 x 10"7 2.8 x 10"6
7.1 x 10
-8
1.8 x 10
8.0 x 10
-8
-6
x 10
-7
3.2 x 10"6
2.3 x 10
-7
1.1 x 10
-6
1.7 x 10"3 sq.ft.'ht.d.d.
sq.ft.-ht.d.d.
sq.ft.-ht.d.d.
4.7 x 10"4 sq.ft.* cl.d.d.
sq.ft.* cl .d.d.
sq.ft.* cl .d.d.
7.1 sq.ft.* year
Note: 'S1 represents the sulfur percentage of oil, by weight.
-------
TABLE 4-13
ESTIMATED NATIONAL INDUSTRIAL LAND USE BASED EMISSION
FACTORS BY TWO DIGIT 1967 STANDARD INDUSTRIAL CLASSIFICATION CODE
SIC
Code
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
19 &
pounds of pollutant
PM SOX
.64
1.22
.58
.06
.06
.11
3.12
.01
.10
1.06
.51
.17
4.03
3.06
.14
.22
.22
.68
.95
39 .08
.50
1.02
.54
.04
.07
.08
3.09
.02
.46
2.78
.38
.17
2.67
2.38
.12
.18
.20
.48
.70
.13
(or kWh
-------
V. THE TRAFFIC MODEL
A. INTRODUCTION
This chapter discusses the vehicle miles traveled (VMT) submodel of
the GEMLUP methodologies for predicting air pollutant emissions from the con-
struction and operation of a major land use project and its induced land uses.
The VMT submodel predicts the vehicle miles traveled from either an absolute
amount of land use in the 10,000 acre area of influence or from an amount of
change in land use in the area of influence. The input to the VMT submodel
is therefore, either,
• The absolute amount of development in the 10,000 acre area of
influence, obtained by suming the output of the predictive
equations of the land use submodel and the projected size of
the major project, or
• The change in development in the area of influence, either between
two major project configurations or between the projected land use
at the end of ten years and the base case.
The elements of the VMT model include vehicle trip generation rates for the
several categories of land use, the corresponding trip lengths, the VMT
traveled by vehicles in several categories, the VMT traveled in several speed
ranges, and the VMT occurring only within the study area.
The precedents for this kind of model were found to be limited.
Generally, this is because most land use transportation studies are based on
the interaction of relatively large segments of regional or metropolitan
areas. In the current study, the character and quantity of the several land
uses are likely to be unique for each study area. Nevertheless, the location
of a given land area, in a metropolitan area is a geographic specification of
distance relationships that are also principal determinants of several factors
affecting VMT. Among these are distances to the metropolitan core and other
major centers of development, and orientation to principal transportation
elements such as expressways, transit terminals, and airports. As a result,
the regional transportation planning agency is a valuable potential source
for some of the model's variables. In this vein, it is considered desirable
to use local sources of information when available.
5-1
-------
B. SUMMARY
Vehicle miles traveled are computed for an "impact area" which is
defined to include both the VMT occurring within the 10,000 acre area of
influence as well as VMT outside the area of influence but occurring because
of the presence of the major project and its induced land uses.
Total vehicle miles traveled in the impact area (denoted VMT ), is
computed as a product of vehicle trips and trip lengths, corrected for a
duplication of trips, viz.,
T 2 17 6
VMT1 = [Z Z LU. L. T. .] - L (Z LU, T2,,)F
i=l j=l J 1 lij r j=l J ^'J
and VMT in the area of influence is computed as follows,
A 2 17 6
VMTA = L [Z Z LU. T. .] - L (Z LU. T2 ,)f
r i=l j=l J 1>J r j=l J ^'J
LU. = Amount of land use (number of dwelling units or thousands of
J
square feet of building floor area) where j is defined as,
1 = Residential Single family detached
2 = Residential Single family attached
3 = Residential Multifamily lowrise
4 = Residential Multifamily highrise
5 = Residential Mobile home
6 = Residential, Nonhousekeeping (Hotels, Motels, etc.)
7 = Commercial <50,000 ft2 (COMM1)
8 = Commercial 50,000-100,000 ft2 (COMM2)
9 = Commercial >50,000 ft2 (COMM3)
10 = Office (OFFICE)
11 = Manufacturing (MANF)
12 = Wholesale & Warehousing (WHOLE)
5-2
-------
13 = Cultural (CULTUR)
14 = Church (CHURCH)
15 = Hospitals (HOSPTL)
16 = Educational (EDUC)
17 = Recreation (REC)
L. = Trip length (miles), where i is defined as
L. = Work oriented tmp length
L2 = "Other trip" length
L = Study area radius
T. = Trip generation rate (vehicles per day), where i is defined as:
T, = Work-trip generation rate
T2 = "Other" trip generation rate
F = Fraction of home based "Other" trips with an Lp < L
The total VMT in both cases (i.e., impact area and area of influence)
are then divided into four vehicle classes. These classes are gasoline,
automobiles, light duty gasoline trucks, heavy-duty gasoline trucks, and
heavy-duty diesel vehicles. The fractionation is computed as a function of
the mix of land use categories in the area of influence.
The total VMT in both cases are also divided into three facility
classes, viz., local streets, arterial streets, and expressways. .The pro-
portion of the VMT occurring on each facility type is a function of the
estimated trip length, L. and L2> for work and other trips. Finally, the
average route speed of VMT on each facility type is estimated to permit the
calculation of air pollutant emissions from motor vehicles.
The significance and characteristics of the factors in the model
are presented in the following section. Procedures for applying the model
are presented in Chapter VII. An example of the application of the model
is shown'in Chapter VIII.
5-3
-------
C. PRINCIPAL ELEMENTS OF THE VMT MODEL
1. Introduction
This study is concerned, in .part, with the air quality impacts
of changes in mobile sources, occurring from predicted changes in mobile
sources, accruing from predicted changes in land use within a 10,000-acre
study area. VMT, categorized in appropriate vehicle classes and route
speeds, is the appropriate data base for calculation of air pollutant emis-
sions from mobile sources. The VMT should be conceived of as an "overlay"
to the regional condition. Thus, whereas vehicular travel through the area
is not evaluated, all the principal elements of VMT generated by the land
uses within the study area are calculated.
The time at which the VMT estimate is to be made is important.
Factors such as trip generation and trip length may be affected by time.
These factors will be discussed individually in following paragraphs. In
respect to time, however, care should be taken in viewing the study area in
the perspective of comparison with similar land uses and related factors for
which current data exists.
The output of the predictive land use model is the primary input
for the VMT model. These data are to be abstracted in the form of the
number of housing units and the thousands of square feet or acres of the
land use categories. The types of residence are measured in numbers of
dwelling units; all other land uses are measured in thousands of square feet
of building except for land devoted to active recreation, which is measured
in acres.
2. Vehicle Trip Generation Rates
A trip in the traffic submodel is defined as a one-way vehicle
movement with either the origin or the destination inside the area of influ-
ence. A trip generation rate is the 24-hour estimate of vehicle trips to
and from a unit of land use (e.g., trips per dwelling unit, trips per thousand
5-4
-------
square feet of floor area, and trips per acre of land use). Vehicle trips
are defined in terms of trip purpose. In the GEMLUP traffic model, only
two categories are used, "work" trips and non-work related "other" trips.
Trip rates will vary with such factors as geographical loca-
tion within a metropolitan area, distance from the core, auto ownership,
density, etc. For example, it is generally believed that auto ownership and
income can be used to estimate trip generation rates for residence area.
For example, Table 5-1 shows characteristic relationships for a large regional
area. The variability of trip rates for other land uses is less defined.
For example, Levinson [151 tabulates the following ranges and typical values
for retail-commercial land use vehicle trip generation rates:
Land Use Vehicle Trip Generation Rate
Retail Commercial l(r Square Feet
Range Typi ca1
Neighborhood Retail 70-240 130
Corminity Retail 60-140 80
Regional Retail 30-50 40
Central Area Retail 10-50 40
As default values for use in the model, a compilation of trip
generation rates for land use categories have been prepared based on references
[15,16,17]. These rates are tabulated in Table 5-2. However, it is believed
that superior results will be obtained by considering locally available data
before selecting trip generation rates. Regional transportation planning
projects in many areas have produced and tabulated similar data that may be
more appropriate.
The values contained in Table 5-2, though they correspond
generally to those contained in one or more of the references, have been
adapted so as to reflect the categories of land use that are aggregated in
5-5
-------
TABLE 5-1
EFFECT OF CAR OWNERSHIP ON AVERAGE NUMBER OF TRIPS PER HOUSE-
HOLD' BY TRIP PURPOSE,'CINCINNATI URBANIZED AREA
Trip Purpose
Noncar One-car Multicar Ratio Ratio
Households Households Households One/None Multi/One
Home-based Work
Home-based Shopping
Home-based Social -Rec.
Home-based School*
Home-based other
Nonhome-based
All Purposes
.62
.37
.30
• 1.7
.32
.19
1.97
1.66
1.05
1.11
0.44
0.87
1.37
6.50
2.49
1.58
2.10
1.04
1.58
2.86
11.65
2.68
2.84
3.70
2.59
2.71
7.20
3.30
1.50
1.50
1.89
2.36
1.81
2.09
1.79
* Based on trip and household data from households interviewed during school year.
Source: "Urban Transportation Models", Ohio-Kentucky-Indiana Regional Trans-
portation and Development Plan, Wilbur Smith and Associates, 1972.
5-6
-------
TABLE 5-2
DEFAULT VEHICLE TRIP GENERATION RATIOS FOR VARIOUS LAND USE CATEGORIES
Land Use Type Trips per measure
single family detached
single family attached
multi family low rise
multifamily high rise
mobile homes
hotels, motels
commercial 1
commercial 2
commercial 3
office
manufacturing
wholesale/warehousing
cultural
churches
hospitals
educational
recreation
dwelling units
dwelling units
dwelling units
dwelling units
dwelling units
103 sq. feet
sq. meters
103 sq. feet
sq. meters
103 sq. feet
sq. meters
103 sq. feet
sq. meters
103 sq. feet
sq. meters
103 sq. feet
sq. meters
103 sq. feet
sq. meters
103 sq. feet
sq. meters
103 sq. feet
sq. meters
103 sq. feet
sq. meters
103 sq. feet
sq. meters
acres
sq. meters
Work Trips, TI
Range Typical
1.0-2.5
0.8-2.2
0.6-1.8
0.3-0.8
1.5-2.0
0.3-0.5
.003-. 01
0
0
0
6.60
.06-. 65
.5-6
.01-. 06
.5-5.5
.01-. 06
0
0
5-35
.05-. 38
Q
0
1.8
1.5
1.2
0.8
1.8
0.4
.004
0
0
0
16
.17
5
.05
4
.04
0
0
16
.17
0
0
Other Trips, T~
Range Typical
6-13
5-11
4-8
2-7
4-7
4-12
.04-. 13
70-240
.75-2.58-
60-140
.65-1.51
30-50
.32-. 54
0
0
0
1-4
.01-. 04
1-4
.01-. 04
0
1-5
.01-. 05
8-30
32375-121407
9
7
6
4
5
10
.11
130
1.4
80
.86
40
.43
0
0
0
2
.02
2
.02
0
4
.04
10
40469
5-7
-------
the individual classifications in the output of the GEMLUP land use model.
In most instances, ranges and typical values have been adjusted based on our
judgment. ,
The vehicle trip generation rates in Table 5-2 assume a generally
low level of mass transit usage. In general, this is an acceptable assump-
tion. For example, the Washington Council of Governments 118] reported the
following distribution of vehicle trips by vehicle type:
automobile 90.9%
truck . 7.6%
public transit 0.6%
taxi 0.9%
In situations where the use of public transit is more predominent, the vehicle
trip generation rates would require adjustment downward.
The accuracy of the traffic sub-model is more closely tied to
estimation of numbers of vehicle trips than to any of the other variables.
Considerations of the nature of the study area, its location with respect to
the regional core, its mix of land uses, its population, etc., should precede
selection of trip generation rates. No fixed logic pattern can be cited. Any
unusual situations would require the use of local data and possibly the
services of a transportation specialist.
3. Trip Lengths
The location of the study area within the metropolitan area
affects vehicular trip lengths greatly. Core-oriented work trips are obviously
affected, but so are other trips. For example, suburban residents drive
farther as a rule for shopping, social and recreational, and other purposes
than do residents in denser areas nearer the core.
5-8
-------
Only two categories of trip lengths are used in the model.
These are average lengths for work trips, L^, and other trip purposes, L2-
Even though significant variations exist in lengths of other trip purposes,
these two groups are consistent with the level of accuracy obtainable with
the model. Again, it is desirable to make use of data available from metro-
politan area planning processes. There are two alternatives for estimating
the mean trip lengths, viz.,
• Obtain values from the regional transportation planning
agency data. It is desirable to consider generalized
location (distance from central city) and land use
characteristics.
• Average vehicle trip lengths, LI and 1-2, both generally
have a high correlation with the distance from the metro-
politan core. Table 5-3 is an example of data that might
be available from a regional transportation planning
agency. The ring system is defined on a map of the region
shown in Figure 5-1. To select trip lengths, one would
first locate the study area, and then would determine the
ring designation. Depending on the year of development
and the year upon which the data is based, the ring used
for the estimate may be one ring closer to the core than
the actual location otherwise would indicate. Thus, for
a Ring 6 location and a 1985 estimate, an appropriate
selection might be LI - 9.2 miles and Lg = 5.5 miles,
based on the given data.
• Alternatively one can solve the following equations:
L1 = 0.003 * p°-20 * S^'49
L2 = 1/2 (0.003 * p°'18 * S^'40 + 0.003 * p°'26 * S^'25)
where P = The SMSA population and
S, £ = The average network vehicle travel speed in
' miles per hour
These regression equations are presented in Reference 6.
These formulas resulted from regression analyses of data
from a number of U.S. cities ranging in population from
33,000 to 6,489,000 [19].
The value for the population that should be used in that
of the Standard Metropolitan Statistical Area (SMSA). The
value should be a prediction of the SMSA population for the
year the traffic submodel is being used for. The value for
$1,2 may be tnat available from area transportation planning
data or that selected from the following relationships.
5-9
-------
01
I
TABLE 5-3
AVERAGE TRIP DISTANCES AND AUTOMOBILE
TRAVEL BY RESIDENCE LOCATION - 1968
Residence Average Auto Trip Distance
Location Home to Home to Non-Home
(Ring) Work Non-Work Based
0 6.1
1 4.5
2 4.8
3 6.3
4 7.5
5 9.2
6 10.1
7 14.5
ALL 8.0
3.3
4.2
3.8
4.2
4.4
5.4
6.3
7.1
4.9
2.0
4.4
4.0
4.5
4.8
5.6
5.7
5.8
5.0
Average Daily Miles Percent of Average Daily Miles
Per Resident.. Households Per Household^
Automobile Owning Cars Work Non-Work TOTAL
9.0
8.6
9.6
13.0
15.6
18.2
19.0
24.2
15.9
26
43
59
72
92
97
97
96
81
1.5
1.8
3.4
6.3
10.5
15.1
15.4
22.0
9.6
2.2
4.3
5.5
9.0
16.8
21.9
24.4
23.7
14.4
3.7
6.1
8.9
15.3
27.3
37.0
39.8
45.7
24.0
*Abstracted from Information Report No. 60, Table IV, WCOG, 1973.
^Averages include both car-owning and non-car owning households.
-------
01
I
FIGURE 5-1 EXISTING ANALYSIS RINGS, WASHINGTON METROPOLITAN REGION
-------
MPH Road Networks Description
S, « = 20 Dense urban network or networks with
I ,&
poor arterial spacing (>1 mile) and
few or remote expressways.
S, 9 = 25 Intermediate network between suburbs and
i ȣ
core, fair arterials, some expressway
service with fair access.
S., 9 = 30 Suburban networks; medium to good arterials
I ,£
good expressway access.
S, 9 = 35 Open network with good to excellent
I ,&
arterial and/or expressways with good
access.
The value for L , required for trip lengths for VMT only within
the study area, is usually 2.23 miles. If the study area is non-circular
because of its location and the local geography, a local effective l_r will
need to be estimated. >
4. Duplicated Trips
Because of the size of the study area and the nature of trip
generation rates, there is a duplication of estimated trips within the area
of influence. The source of this duplication is the double counting that
occurs when both ends of a trip are within the area of influence. The dwel-
ling unit (shopping) trip and the commercial (shopping) trip is an example.
The procedure for obtaining a correction for duplicated trips
is based on the premise that the majority of such trips are home to work or
home based other trips. Duplicated trips between dwelling units or between
non-residential trip generators are assumed to be negligible. The correction
factor is, therefore, the proportion of home to work and home based other
trips that have a trip length less than the radius of the area of influence
Lp.
5-12
-------
If trip length distribution curves are available from regional
transportation data, this fraction may be selected directly. A typical
curve of the type that may be obtained is shown in Figure 5-2.
If such curves or specific values are not available, the factor
may be assumed to be 0.40. A number of locations, such as that shown in
Figure 5-2, were determined to have slightly higher values. The value of
0.40 was selected as a conservative correction factor (i.e., the higher a
porportion that is used, the greater will be the number of trips that are
subtracted as duplicated trips).
5. Speed Ranges
The relationship between vehicle average route speed and air
pollutant emission rates (i.e., in grams per mile) make it critical to con-
sider the average route speeds of the estimated VMT in calculating emissions.
The emission factors presented in the EPA Compilation of Emission Factors
[8 ], are estimated from a typical driving cycle that includes acceleration,
deceleration, and constant operating speeds. Therefore, it is appropriate
to only consider the average route speed.
The first step in determining the average speeds of the estimated
VMT is to estimate the proportion of travel on various facilities. Most
trips include distances and periods of time that are traveled on local streets,
arterials, and expressways. It is recommended that through the use of a local
highway network map and the consideration of the major trip generators in the
study area, the proportion of time spent on each type of facility be esti-
mated for a specific application of the model. In lieu of the procedure (e.g.,
where the highway network is unknown) the proportion of time spent on each
type of facility can be estimated using a theoretical distribution. These
graphs show the theoretical use of each facility given a freeway access ramp
spacing (viz., two or four miles). The local streets and arterial streets
are assumed to be distributed, respectively, every 0.125 and 1.0 mile. The
freeways are spaced every two or four miles. Figures 5-3 and 5-4 show the
5-13
-------
or
School
Personal Busi-
ness & Shopping
Total Non Work
Work
Sodo-Recreation
1
10
11 12
23456789
• Distance (miles)
FIGURE 5-2 PLOT OF AVERAGE TRIP LENGTH FREQUENCY DISTRIBUTION BY TRIP TYPE
5-14
-------
14
o
.£ 8
a
§2
-------
relative proportion of travel on each of the three types of facilities [20].
Using the previously calculated average trip lengths for work and other
trips, the relative proportion of trips on each facility can be estimated.
The airline trip distance is ignored. If it is not certain which figure is
most applicable, it is suggested that the mean of the two figures be employed.
For example, a hypothetical eight mile work trip length and five mile other
trip length would result in the following calculation:
Distance Traveled on Each Facility Type
TjCS miles) T2 (5 miles)
2 mile 4 mile
spacing spacing mean
local
arteri al
expressway
0.5
2.5
5.0
0.5
3.5
4.0
0.5
3,0
4.5
8
2 mile
proportion spacing
.06
.38
.56
1.00
0.5
2.75
1.75
4 mile
spacing
0.5
3.0
1.5
mean proportion
0.5
2.875
1.625
.10
.575
.325
1.00
The proportion of distance traveled on each facility type (i.e., .06, .38,
.56, .10, .575, .375) would then be employed (i.e., they should be entered on
work sheet number (VMT-3)).
The calculation of VMT occurring in the area of influence had an
assumed trip length of L (usually 2.23 miles). However, many of these trips
are longer than l_r; all that the calculation attempts to estimate is that
portion of the trip occurring within the area of influence. Therefore, rather
than entering Figures 5-3 and 5-4 and calculating a new set of proportions of
travel on each facility type, it is appropriate to use the proportions calcu-
lated in the preceeding paragraph.
While it is conceivable that the average route speed of all
travel on each facility type could now be estimated, it is more accurate to
consider the slower average route speed of VMT that occur during the peak
5-16
-------
14
12
10
8
2 -
Total Distance
Traveled
Distance
Traveled
on
Expressways
Distance Traveled
on Arterials
1 Local Streets
8
10
Airline Trip Length in Miles
FIGURE 5-4 THEORETICAL RELATIVE USE OF LOCAL STREETS,
ARTERIAL STREETS, AND EXPRESSWAYS WITH ;A
FOUR MILE RAMP SPACING
SOURCE G2Q]
5-17
-------
hour (i.e., rush hour). Approximately 10 percent of the average annual
daily traffic will occur in the thirtieth highest peak hour [21]. The average
route speeds of the 10% of VMT occurring during the peak hour* are then cal-
culated separately from the 90% of VMT that is off peak.t
The average operating speed of each facility type is a function
of the highway design speed and the ratio between the volume demand and the
capacity. These relationships are depicted in Figures 5-5 and 5-6 for,
respectively, expressways and urban arterials. If the typical volume-capacity
ratio and design speed for the network under consideration is known, the
average operating speed can be estimated. In lieu of obtaining this informa-
tion, it is suggested that "C" level of service for off peak hour and "E"
level of service for peak hour be used as conservative approximations.
Accordingly, the following estimates can be made:
expressway
arterial 35 mph speed
local 25 mph speed
off peak
45
28
18
peak
37
20
15
These values should be employed unless better estimates, based on local volume-
capacity ratios and design speeds, are available.
6. Vehicle Class
In view of available emission factors, four classes of vehicles
are used:
AG = gasoline powered automobiles
LOG = light-duty, gasoline powered trucks, <. 6,000 pounds,
gross
HDG, = heavy-duty, gasoline powered vehicles, > 6,000 pounds,
gross vehicle weight
HDD = Diesel-powered vehicles (predominantly trucks and buses)
* Work trips are assumed to be of equal length, L-, (other trips are also assumed
to be of equal length, L2).
t Vehicle speed during the afternoon peak hour is typically not substantially
reduced.
5-18
-------
70
60
T
0
t!
40
APPROX.,
.5530
.
o
20
10
• 0
NO. OF
I !•/
BASIC
INDEPENDENT
LEVEL
OF
SERVICK
RANGES
0.1
0.2
0.3
0-4
0.6
0.7
0.8
0.9
1.0 PHF
JES
4
6
8
-i A »i B v'|'Dv||u c ,L D Bt, E T
p - C t-i D »!-• C »
1 fe^D-l0. ^ E— 1)
A B r^-FTtir^^T-6 — !— r
C| D | E •
1.00
0.91
0.83
0.77
0,91
0.83
0.77
0.91
0.33
0.77
FIGURE 5-5 RELATIONSHIPS BETWEEN V/C RATIO AND OPERATING SPEED, IN ONE
DIRECTION OF TRAVEL, ON FREEWAYS AND EXPRESSWAYS, UNDER UNIN-
TERRUPTED FLOW CONDITIONS
SOURCE: REFERENCE J22J.
5-19
-------
s
a"
>
<
oc
B -j
-------
Analysis of trip generation for the land use categories and vehicle
classes, indicate that the probable fraction of HG and DL vehicles generated
by residences and commercial land uses is so small as to be far less than the
probable error of the trip generation rate for these land uses.
From the average distribution of trucks cited in reference [8 ],
the following gross vehicle classification factors have been selected:
AG automobile, gasoline .804
LOT light duty truck, gasoline .118
HDG heavy duty vehicle, gasoline .046
HDD heavy duty vehicle, diesel .032
TOTAL 1.00
These factors should produce a reasonable and conservative estimate of truck
distribution. They are generally applicable, without correction, where the
study area is predominantly residential, or where there is a relatively uni-
form distribution of land uses, or where the amount of development likely
to generate about average numbers of HG and DL trips is not significantly
large.
However, some study areas are likely to have large, if not domi-
nant truck generating land uses such as manufacturing and wholesaling/ware-
housing. A method to adjust the gross factors in these study areas has been
developed since these land uses have been found to generate above-average
truck trips. The method is based on a ratio of the number of trips generated
by manufacturing, wholesaling and warehousing to the sum of all trips in
study area. The factor, Fj, describes the significance of above average
truck-generating land uses. The logic of this approach is based on the pre-
mise that normal distribution of HHG and HDD trucks is about .07. If all the
trips generated by large truck generators exceed .07 of all "other" study
area trips, this is an indication of abnormal distribution. Therefore, the
excess (over .07) would represent a reasonable correction amount to be
distributed among GH and DL trucks, and subtracted from automobile. Light
duty truck is assumed to remain constand. If Fj were less than .07, condi-
tions would be considered "average" and no correction would be required.
5-21
-------
The correction should be applied as follows:
Assume F- = 0.17
then .17 - .07 = .10 which is the required amount of the
correction.
The adjusted truck factors are:
A6 = .804 - .10 = .704
LOT = .118
HD6 = .046 + .8(.10) = .04 + .08 = .126
HDD = .032 + .2(.10) = .01 + .02 = .052
5-22
-------
VI. MOTOR VEHICLE EMISSION FACTORS
Emissions of the five criteria pollutants resulting from motor vehicle
operation can be calculated using the EPA publication, Compilation of Air
Pollution Emission Factors. Second Edition [ 8], This publication is com-
monly referred to as "AP-42" and is regularly updated as research of the
emission characteristics of all sources continues. The latest update to
this publication is Supplement 5, issued in December 1975. This supplement
prescribes a number of changes to the methodology for computing motor vehicle
emissions presented in earlier versions of AP-42. It is expected that further
updates to mobile source emission factors will be included in Supplement 7
which should be issued in the beginning of 1977. The methodology presented
in this workbook is based on Supplement 5 and is expected to be consistent
with future updates; only the emission factors are expected to change.
The Clean Air Act originally required emissions of three motor vehicle-
related pollutants to be reduced 90 percent from the 1971 model year emis-
sions before 1977. Amendments to this Act have subsequently relaxed these
standards and it is expected that further relaxations will be enacted during
1976. The result of the Act and its amendments is that vehicles of each
model year after 1967 have a different new vehicle emission rate, react
differently to changes in speed, ambient temperature, and operating temper-
ature, and deteriorate with age at different rates.
To account for this variation, AP-42 presents a tabulation of the aver-
age emission rate for each model year for the calendar years 1971 through
1980, 1985, and 1990. Also included are equations which describe the vari-
ation in emissions with ambient temperature, operating temperature, and
speed, for each model year. These tables represent emissions projections
based on the present schedule for implementation of emission controls. Any
changes in this timetable will be reflected in future Supplements to AP-42,
however, close attention should be paid to changes in the Act as the delay
time in issuing a supplement is relatively long. AP-42 presents a table
of several emissions standards and the subsequent deterioration factors
which can be substituted for the existing tables on an interim basis until
Issuance of a future supplement by EPA.
6-1
-------
The basic equation used to calculate a composite emission factor for a
given calendar year and pollutant* is based on the Federal Test Procedure
(FTP) methodology.** The equation is:
enpstwx = *=_ (cipnminvipsziPtriPtwx + min (fi + ei »
where: e
nDStwx = Composite emission factor in grams per mile for
p calendar year (n), pollutant (p), average speed (s),
ambient temperature (t), percentage cold operation (w),
and percentage hot start operation (x)
cion = ™e FTP mean emission factor for the ith model year
M vehicles during calendar year (n) and for pollutant
(P)
m. = The fraction of annual travel by the ith model year
vehicles during calendar year (n)
v. = The speed correction factor for the ith model year
P vehicles for pollutant (p), and average speed (s)
z. t = The temperature correction for the ith model year
P vehicles for pollutant (p) and ambient temperature
(t)
r. . = The hot/cold vehicle operation correction factor for
ipwx ^Q .^ model year vehicles for pollutant (p), ambient
temperature (t), percentage cold operation (w), and
percentage hot start operation (x)
f. = Crankcase hydrocarbon emission rate in grams per mile
from vehicles of model year (i)
e. = The evaporative hydrocarbon emission factors in grams
per mile for each model year (i). This factor includes
both diurnal losses and "hot soak" emissions.
This equation represents the methodology to compute emissions for all
regions, vehicle types and calendar years using the Federal Test Procedure
(FTP). AP-42 presents emissions for calendar years 1971 and 1972 using the
*Carbon monoxide, nitrogen oxides, and hydrocarbons. Emissions of sulfur
oxides and particulates are computed by employing relatively simple
methodology discussed in Chapter VIII.
**This procedure is described in Section 3.1.2 of AP-42, Supplement 5 [ 8].
6-2
-------
results of actual vehicle testing. Emissions from vehicles after 1972 have
been estimated based on extrapolation of previous tests. These emission
factors and correction factors are presented in Appendix D of AP-42, Supple-
ment 5 [8]. This appendix provides a more general methodology for comput-
ing emissions and is used in development of the worksheets presented in
Chapter VII.
A. DESCRIPTION OF EMISSION FACTOR COMPONENTS
This section describes each of the variables presented in Equation
6-1. A general description of each component for each vehicle class (auto-
mobile, light-duty truck (LOT), heavy-duty gasoline-powered vehicle (HDG),
and heavy-duty diesel-powered vehicle (HDD)) is presented including reference
to the specific tables in AP-42 [ 8]. The impact of several vehicle classes
(motorcycles, construction equipment, etc.) have not been included in this
analysis. Also the impact of transportation control strategies which affect
individual vehicle emissions such as inspection/maintenance plans and retro-
fit devices are not addressed. The user is referred to AP-42 for further
information if these subjects will have a significant impact.
1. Mean Emission Factor (C. )
This factor represents the emissions (including deterioration)
of a given pollutant (p) from the ith model year during calendar year (n).
This factor is presented for three regions (low altitude, high altitude,
and California) and three pollutants in Tables Dl.l through Dl.20 for auto-
mobiles. Emissions of the three pollutants for LOT and HDG are presented
in Tables D2.1 through D2.10 and D4.1 through D4.10, respectively, for low
altitude only. Section D4.5 presents a methodology for estimating high
altitude and California emission factors for these vehicle classes. Emis-
sion factors for HDD are presented in Table D5.1 for three pollutants and
all years. These factors are applicable for both low and high altitude
operation. Section D5.4 presents the methodology to estimate HDD emissions
in California.
6-3
-------
2. Weighted Annual Travel On, }
This variable reflects the relative miles traveled of the ith
model year during calendar year (n). This variable is calculated for the
12 years prior to year (n) representing a total of 13 values. The value of
m. which represents the oldest model year (year (n-12)) also includes
vehicles older than 12 years 1n order to include all vehicles in the weight
ing. The equation used to compute m. is:
1n
j=n-12
where: a. = Number of vehicles of model year (i) in calendar
year (n)
b- = Average number of miles driven by vehicles of model
year (i) in calendar year (n)
If the data required to compute these values is not available,
national averages for min which can usually be applied for any calendar year
(n) are presented for automobiles, LOT, HDG, and HDD in Tables D1.22, D2.ll,
D4.ll, and D5.2, respectively, of AP-42.
3. Speed Correction Factor (v. )
This factor adjusts the emissions of pollutant (p) from a
vehicle of model year (i) traveling at a speed (s). The mean emission
factors (c. ) presented above are calculated for a single speed for each
vehicle type (18 mph for HDD, 19.6 mph for all other classes). Vehicle
speed, however, has a significant effect on emission rates. Equations for
(v. s) are presented for each region and year in Table D1.23 for automobiles
and LOT (Table D2.12 is identical to Table D1.23). Table D4.12 presents
speed factors for HDG and Equations D5.2 and D5.3 present an adjustment
factor for HDD. For the vehicle classes other than HDD, the effective range
of these equations is 15 to 45 miles per hour. At speeds lower than 15 mph,
6-4
-------
Tables D1.24, D2.13, and D4.13 should be applied to the respective classes.
Guidance in estimating emissions at speeds greater than 45 mph is not pro-
vided in AP-42. An assumption which can be made is that the speed correc-
tion factor for 45 mph is applicable at all speeds greater than 45 mph.
4. Ambient Temperature Correction Factor (z,-Dt)
This correction factor adjusts the pollutant (p) emission rate
• f
(C4««) °f vehicles of the 1th model to account for ambient temperature (t°F).
i pn ' . • .
This factor is applied only if (t) is outside the FTP range of 68-86°F.
Equations for z. . are presented in Table D1.25 for automobiles and light-
duty trucks, providing separate equations for vehicles equipped with cata-
lytic converters and those without. (Table D1.25 is identical to Table
D2.14.) HDG and HDD emissions are assumed not to vary with ambient tempera-
ture (zipt = 1.0).
5. Operating Temperature Correction Factor (<"• t )
This factor adjusts the pollutant (p) emission factor of
vehicles of the ith model year as a function of ambient temperature (t),
percent of vehicles operating from cold start (w), and percent of vehicles
operating from hot start (x). This factor is applied only to automobiles
and LOT; heavy-duty vehicles are assumed to be operated only in the warmed-
up state (r,-D4.wx = 1.0). This variable is applied if the mix of light-duty
vehicles among cold start, hot start, and warmed-up varies significantly
from the FTP standard of 20 percent, 27 percent, and 53 percent, respectively.
Equations Dl-2 and Dl-3 present equations to calculate (n) for pre-1975 and
post-1974 model years, respectively. Equations for f(t) and g(t) are
identical for automobiles and LOT and are presented in Table D1.25.
6. Crankcase Hydrocarbon Emission Factor (f. )
This quantity is the amount of hydrocarbons emitted from the
crankcase of model year (1) vehicles. This factor has no effect when com-
puting carbon monoxide and nitrogen oxide emissions (f. = 0.0). Tables
6-5
-------
D1.26, D2.15, and D4.14 present crankcase emissions for automobiles, LOT,
and HD6, respectively. Crankcase emissions from HDD are negligible. Crank-
case hydrocarbon emissions have been eliminated in all post-1967 automobiles
and trucks (f.. =0.0).
7. Evaporative Hydrocarbon Emissions (e.)
This quantity is an estimate of hydrocarbon losses from the
carburetor and fuel systems from vehilces of the ith model year. This
factor has no effect on the calculation of carbon monoxide and nitrogen
oxide emissions (e.. = 0.0). This factor is calculated by:
e. = (g. + k.d)/t .
where: g. = diurnal evaporative loss (grams/day)
k. = hot soak evaporative emissions (grams/trip)
d = average number of trips per day
t = average number of miles traveled per day
Table D1.27 presents values of g. and k. for automobiles. Conversion to
grams/mile was achieved using an assumption of 3.3 trips per day and a total
of 29.4 miles traveled per day. Tables D2.15 and D4.14 present values of
6| for LOT and HD6, respectively. Evaporative losses from diesels are neg-
ligible.
6-6
-------
VII. COMPUTATION WORKSHEETS AND INSTRUCTIONS
A detailed step-by-step procedure is presented in the following sections
for the computation of land use development, traffic generation, and the
resulting air pollutant emissions associated with a planned Major Project.
A series of computation worksheets (summarized in Table 7-1) are the vehicle
for this procedure.
A. LAND USE MODEL
Predictive equations have been developed for total land use devel-
opment in an area of influence ten years after the initiation of a Major
Project (see Section III). These equations have been incorporated into a,
set of worksheets, presented below, that can be used to project future land
use development. These equations are applicable to areas where a large
Residential, Industrial, or Office Major Project will be, or already has
been, built. They do not constitute a general land use model. Specifically,
a proposed Residential projects final size should be in the range of 1,100-
5,300 total dwelling units, and the final size of an Industrial or Office
project should be in the range of 3,600-9,100 employees. The use of the
following worksheets should be limited to situations where the Major Project
is in this size range.
Unless otherwise noted, the variables defined in Worksheets RLUM-1,
RLUM-2, IOLUM-1, IOLUM-2, and IOLUM-3 correspond to the year of project initi-
ation, which could be now, in the past, or the near future. They are listed
under the geographical area they correspond to, e.g., Item 2 on Worksheet
RLUM-1 is the number of dwelling units in the area of influence. The size of
the area of influence (Item 1 on Worksheets RLUM-1 and IOLUM-1) in all cases
2
is either 10,000 acres or 40,470,000 m , depending upon whether English or
Metric units are used. The land use quantities predicted refer to the area
of influence ten years after the initiation of the Major Project. It is
assumed that the Major Project will be completed by the end of this ten year
interval.
7-1
-------
TABLE 7-1
SUMMARY OF COMPUTATION WORKSHEETS
RlLUM-1 through RLUM-7
IOLUM-1 through IOLUM-7
LUM-1, LUM-2
LUM-3
VMT-1
VMT-2
VMT-3
VMT-4
VEM-1
VEM-2
Calculation of Estimated Land Use
Residential Model
Calculation of Estimated Land Use
Industrial-Office Model
Calculation of Land Use Model Confidence
Intervals
Summary of Land Use Predictions
Calculation of Vehicle Trips
Calculation of Gross VMT
Calculation of VMT by Facility Type
Calculation of Vehicle Classification
Proporticns
Calculation of Motor Vehicle Emission
Rate for a Specific Vehicle Category
Calculation of Composite Motor Vehicle
Emission Factor
Calculation of Stationary Source Emissions
Calculation of Motor Vehicle Emissions
Summary of Emissions
7-2
-------
Separate complete worksheets are provided for Residential (Section
VII.A.I) or Industrial/Office (Section VII.A.2) projects. In both cases,
the worksheets have been designed to allow the computation of land use in
either English or Metric units. The convention used throughout is that items
or numbers referring to Metric units are placed immediately following corres-
ponding English unit items in parentheses. A second convention used in the
worksheets is that some numbers are expressed in exponential notation due
q
to their extremely small or large size. For example, 3.35E9 = 3.35 x 10 =
3,350,000,000 and 4.02E-5 = 4.02 x 10"5 = 0.0006402. Finally, after comput-
ing the projected total land uses, confidence intervals can be obtained for
each projection using Worksheets LUM-1 and LUM-2, discussed in Section III.
A.3.
1. Residential Land Use Model
The following step-by-step procedure should be followed:
a. Worksheets RLUM-1 and RLUM-2
Enter the values of the variables listed in Items 1 through
39 and perform the arithmetic operations indicated. All items must be com-
pleted.
b. Worksheets RLUM-3, RLUM-4, and RLUM-5
Enter the values of the variables called for in the far
left hand column from the two previous worksheets. For each variable entered,
perform the indicated multiplications between the variable and the constants
that appear in the Land Use Category columns, and record the result in the
blank space below each constant. For example, if the value of item 3 is 0.80
(English units), then the quantity 0.80 x 8910 = 7128 is recorded in the
first blank space in Column L,. Proceed and fill in all these worksheets.
Next, sum up the recorded quantities in each of the 12 Land
Use Category columns and enter the result on Worksheet RLUM-5 on the now
labeled "Z Columns". Add in the appropriate constants given in the next row
7-3
-------
and enter the total predicted land uses 1n the final row. Revise any nega-
tive total land use quantities up to the value of 0.
c. Worksheet RLUM-6 $r,d RLUM-7
Enter projected percentages for the disaggregation of Resi-
dential and Commercial land use in the Area of Influence at project comple-
tion in Items 40 through 47. These percentages should take account of the
local factors which will most likely effect the density of future develop-
ment in the Area of Influence. In the absence of such data, default values
are provided which refer to an average Major Project and the development
patterns of the period 1960-1970.
Perform the multiplications indicated in the second half
of the worksheet and record the values of final projected land use in Items
48 through 65.
2. Industrial/Office Land Use Model
The following step-by-step procedure should be followed:
a. Worksheets IOLUM-1, IOLUM-2, and IOLUM-3
Enter the values of the variables listed in Items 1 through
44 and perform the arithmetic operations indicated. All items must be com-
pleted.
b. Worksheets IOLUM-4, IOLUM-5, and IOLUM-6
Enter the values of the variables called for in the far left
hand column from the two previous worksheets. For each variable entered,
perform the indicated multiplications between the variable and the constants
that appear in the Land Use Category columns, and record the result in the
blank space below each constant. For example, if the value of item 3 is
0.80 (English units), then the quantity 0.80 x 2480 = 1984 is recorded in
the first blank space in column L,. Proceed and fill in all three worksheets.
7-4
-------
Next, sum up the recorded quantities in each of the 12 Land
Use Category columns and enter the result on Worksheet IOLUM-5 on the now
labeled "Z Column". Add in the appropriate constants given in the next row
and enter the total predicted land uses in the final row. Revise any nega-
tive total land use quantities up to the value of 0.
c. Worksheet IOLUM-7
Enter projected percentages for the disaggregation of Resi-
dential and Commercial land use in the Area of Influence at project comple-
tion in Items 45 through 52. These percentages should take account of the
local factors which will most likely effect the density of future develop-
ment in the Area of Influence. In the absence of such data, default values
are provided which refer to an average Major Project and the development
patterns of the period 1960-1970.
Perform the multiplications indicated in the second half
of the worksheet and record the values of final projected land use in Items
53 through 70.
3. Computing Confidence Intervals
The preceding worksheets are an application of predictive equa-
tions that have been developed for land use in various categories. These
equations are of the form:
n
Y = bn + Z b.X.
0 .=1 i i
where: Y is the land use being predicted
X-,, Xp, ... X are the independent variables used in the
prediction
bjj, b-|, ... bn are the model coefficients
The variables X-j, ... Xn are those which are recorded in the far left column
of Worksheets RLUM-3,4,5 and IOLUM-4,5,6. The model coefficients bQ, b1, ...
bn are the constants listed in the center of these worksheets.
7-5
-------
For a given land use equation, confidence Intervals can be
specified 1n the form:
Y + tV(Y)1/2
where t Is the t-statistic for the regression of the model equation and
V(Y) is the variance of Y. V(Y) can be expressed as:
n n
V(Y) = Z z X.X, Covariance (b.b.)
1=1 j=l ] J ] J
All of the necessary data for computing confidence intervals are contained
on the previous worksheets and in Appendix A to this report, which summarizes
the statistical output for the predictive land use equations. To compute
confidence intervals for each of the 18 categories of final projected land
use listed on Worksheets RLUM-6 and IOLUM-6, fill out 18 sets of Worksheets
LUM-1 and LUM-2.
a. Worksheet LUM-1
Enter the final projected land use category on the first
line, find the corresponding dependent variable name and general land use
category from Table 7-2 and place it on the second line. From the appropri-
ate Land Use Worksheets (RLUM-3,4,5 or IOLUM-4,5,6), find which predictor
variables are used under the column corresponding to the general land use
category, find the prediction variable names from Table 7-3 (Residential)
or 7-4 (Industrial-Office) and record these names along with the corres-
ponding variable values from the far left column of the worksheets under
items 1 through 6. For example, for the final projected land use category
"Single Family Attached", the dependent variable name is "RES" and the gene-
ral land use category is "L,". Assuming the Residential model from Worksheets
RLUM-3,4,5 and Table 7-3, the predictor variable names are found to be
"DUACRE, DELP2, DISCBD, HWYINT, MPR70, and SEWER". The corresponding values
would be entered for items 3, 20, 4, 5, 33, and 30, respectively.
7-6
-------
TABLE 7-2
LIST OF DEPENDENT VARIABLE NAMES FOR EACH
FINAL PROJECTED LAND USE CATEGORY
Final Land Use Category Dependent Variable
Name
Single Family Attached
Single Family Detached
Mobile Homes
Multi family Low Rise
Multi family High Rise
Commercial <50K
Commercial 50-100K
Commercial >100K
Office
Manufacturing
Non-Expressway Highway Lane Distances
Wholesale-Warehousing
Hotels, Motels
Hospitals
Cultural Facilities
Churches
Educational Facilities
Recreational Facilities
RES
RES
RES
RES
RES
COMM
COMM
COMM
OFFICE
MANF
HWLMNX
WHOLE
HOTEL
HOSPTL
CULTUR
CHURCH
EDUC
REC
General
Land Use
Category
h
h
Li
LI
h
4
4
4
4
L4
4
4
4
4
4
ho
Ln
L,o
7-7
-------
TABLE 7-3
LIST OF PREDICTOR VARIABLE NAMES
FOR THE RESIDENTIAL MODEL
? Predictor Variable
Item Number
(Worksheets RLUM-3,4,5)
3
20
4
5
7
21
9
23
11
26
16
32
33
28
14
27
35
36
30
38
39
Predictor Variable
Name
DUACRE
DELP2
DISCED
HWYINT
OFFACR
EMP60
VACACR
DELEMP
MANACR
AUT02
MINC
MPR68
MPR70
OFFVAC
ZRES
AUTO
INCMPL
UNIV
SEWER
MPKIDS
MPACRE
7-8
-------
TABLE 7-4
LIST OF PREDICTOR VARIABLE NAMES
FOR THE INDUSTRIAL/OFFICE MODEL
Predictor Variable
Item Number
(Worksheets IOLUM-3,4,5)
3
5
36
7
8
12
16
42
17
13
29
19
32
27
20
22
14
40
24
39
30
43 '
44
41
35
Predictor Variable
Name
DUACRE
VACACR
OFFVAC
VACHS6
DISCBD
ZCOMM
OFFACR
MPE70
RRMI
ZOFF
MINC
MANACR
DELEMP
MINCC
HWYINT
WWEA
ZIND
SEWER
NONHSE
ENERGY
EMP60
MPACRE
PVTSCH
MPET2
AUTO
7-9
-------
Locate the statistical output in Appendix A corresponding
to the appropriate Major Project type (Residential or Industrial/Office)
and dependent variable name. For each pair of predictor variables listed
on this Worksheet and LUM-2, find and record the covariance from the
"variance-covariance matrix" in Appendix A. Continuing the previous example,
we would record "0.677537E7" under the covariance for the pair 1 and 1
(DUACRE and DUACRE), "-353084" for the pair 1 and 2 (DUACRE and DELP2), etc.
Compute items 7 through 42 by performing the indicated
multiplications. Note that equations with less than 6 predictor variables
will have less covariances to record and multiplications to perform.
b. Worksheet LUM-2
Compute the variance of the dependent variable by summing
up Items 7 through 42. Record the t-statistic of the predictive equation
from Appendix A, and perform the equations noted in items 44 through 48 to
obtain the confidence interval. Note that final projected land use cate-
gories that are disaggregations of total Residential or total commercial
land use will require calculation of Items 1 through 46 only once in the
general land use category.
c. Worksheet LUM-3
Enter the projected final size of the Major Project in
Items 49 through 56. Copy the final projected land uses/excluding Major
Project from Items 48 through 65 of Worksheet RLUM-6 or from Items 53
through 70 of Worksheet IOLUM-6, depending upon the Major Project type, and
perform the indicated additions to obtain total projected land use (includ-
ing Major Project).
7-10
-------
B. CALCULATION OF VEHICULAR TRAFFIC
This section contains procedures and guidelines that are designed
to facilitate the computation of VMT. A step by step procedure is provided.
The virtue of this model is its relative simplicity. Nevertheless, the appli-
cation of the model requires sound judgment in either modifying or accepting
the default trip rates and trip lengths as representative of a particular
study area. Familiarity with these concepts and their typical values is
desirable.
1. Worksheet No. 1 (VMT-1)
Compute the effective radius of the area of influence and
enter on line no. 1. In most instances the area will be
a circle, so 2.23 miles (3589 meters), should be entered.
However, in some areas a circle with a 2.23 mile radius
will include a coastline or an impenetrable geographical
barrier. In order to maintain the same area (i.e.,
10,000 acres (4.05 x 107 square meters) the radius must
be adjusted, i.e.,
R, miles =
15.625
IT - i (9 - sin 9)
R. meters - ^466351
ir - 1 (9 - sin 9)
2"
N.B. 9 is in radians
Enter the total amount of land use for each category in
column 2. The first five categories (i.e., residential)
are in units of dwelling units. The next ten categories
are in units of 10^ square feet (or square meters) and the
last category, recreation, should be entered in acres.
The definition of each category is shown in Table 2-1.
Enter the work trip rate, T,, in column 3
(i.e., non-work) trip rate, T«, in column 5
rates are shown in Table 5-2.
and the other
Default trip
7-11
-------
Compute the work, trips and other trips by taking the product
of, respectively, columns 2 and 3 and columns 2 and 5. Enter
the results in columns 4 and 6.
Compute the total work trips and total other trips by summing
respectively, column 4 and 6. Enter the results on lines
23 and 24.
2. Worksheet No. 2 CMT-2)
• Copy the values from lines 23 and 24 on VMT-1 to the lines
with the circled numbers.
- Determine the work trip length and the other trip length in
miles (or meters) according to the instructions on page
Enter the results in column 25 as well as Lp from line 1.
• Compute the total uncorrected VMT by multiplying the total
work trips and total other trips by the work trip length
and the other trip length.
• Compute the total residential work trips and total residen-
tial other trips by adding the first six entries in columns
4 and 6 on worksheet no. 1. Enter the sums on worksheet no.
2 on, respectively, lines 30 and 34.
• Determine a value for the proportion of work and other trip
lengths less than Lr» through the use of either local infor-
mation or the default value 0.40.
• Compute the VMT correction by multiplying the duplicated
trips by the radius of the area of influence. After sub-
traction, enter the net VMT on lines 38 through 41.
• Enter the proportion of VMT occurring in the peak hour on
line 42.
• The difference between 1 and this proportion are the VMT
occurring during the off peak hour. Enter this number on
line 43.
• From local information or through the uses of Figures 5-3
and 5-4, determine the proportion of distance traveled on
each facility type (i.e., local streets, arterials, and
expressways) for the average work trip and other trip.
Enter these values on lines 44 through 49.
3. Worksheet No. 3 (VMT-3)
• As indicated, the VMT on each facility type, for the peak
and off peak time interval, for both the impact area and area
of influence.
7-12
-------
4. Worksheet No. 4 (VMT-4)
• Compute the number of trips that are manufacturing or ware-
house related by adding lines 19, 20, 21, and 22 on VMT-1;
enter the sum on line 62.
• Compute the proportion of trips that are warehousing related.
• If the proportion, line 63, is greater than 0.05, determine
the adjustment factor (i.e., the excess above 0.05} and
calculate the vehicle classification factors.
• Determine the average route speeds from local information
or through the use of Figures 5-19and 5-20. Otherwise use
the default values indicated below:
69_. 50 mph 72_. 37 mph
70_. 28 mph 73,. 20 mph
71. 18 mph 74. 15 mph
7-13
-------
C. CALCULATION OF MOTOR VEHICLE EMISSIONS
This section contains guidelines and worksheets that can be employed
to calculate motor vehicular emissions. The methodology for computing
vehicular emissions is derived from the EPA publication Compilation of Air
Pollutant Emission Factors, Second Edition [3]. This publication is
generally known as "AP-42" and has been updated by means of five supplements
at this writing. Chapter VI presents further information on expected updates
to these emission factors. The methodology to compute vehicle emissions
which is presented in this section relies heavily on Appendix D of AP-42.
It is essential that this document (including all updates) be available
before beginning these calculations.
The motor vehicular traffic data required to estimate emissions
are calculated in the previous section. For either the 10,000 acres of
influence or the larger impact area, the following is required:
• Vehicle miles traveled (VMT) on each facility type for the
peak hour and off peak.
• Average route speed of the VMT on each facility type for the
peak hour and the off peak.
• Vehicle class distribution. This is assumed to be constant
between facility types and peak/off peak time intervals.
In addition, as discussed in Chapter VI, several decisions must be made
regarding the applicability of the default assumptions in the emission
equation (i.e., the use of a national vehicle age distribution, and the
assumption of the relative number of cold starts).
1. Motor Vehicle Emissions Worksheet No. 1 (VEM-1)
This worksheet should be completed for each vehicle type, pol-
lutant (CO, HC, N0¥), and speed (i.e.,, facility type and peak-off peak time
A
interval) of interest. In the case of the GEMLUP traffic sub-model, this
would require 72 (four vehicle types, three pollutants, and six average
route speeds) different sheets. It is essential that givens9i.e., vehicle
7-14
-------
type, speed, design, year, etc, lines 1 through 8 be specified before
beginning the computation. It is often helpful to note the model year
(column 9) which corresponds to each vehicle age so that model year depen-
dent factors are computed correctly.
The AP-42 (Appendix D) tables which will be referred to in
this discussion have been described in detail in Chapter VI. For this
reason, the specific table or tables to be used in computing each factor
will not be included in this discussion. The reader is referred to Chapter
VI for both a further description of each of the components and on enumera-
tion of the tables to be used to compute each component. With this infor-
mation, the following procedures can be used to compute emissions.
• Fill in the model year corresponding to each vehicle age.
The design year (line 4) is Age=l, Age=2 is the design
year minus 1, etc. Finally, Age >_ 13 is the model year
twelve years prior to the design year.
• Locate the table in AP-42 corresponding to the correct
vehicle type (line 1), pollutant (line 2), design year
(line 4), and region (line 5). Fill in column 11 with
the emission rate (C^ ) which corresponds to each model
year. p
• Fill in column 12 with the vehicle age weighting factors
(M. ) corresponding to the correct vehicle type (line 1)
ana vehicle age (column 10).
• Calculate the speed correction factor (column 13) for the
correct vehicle type (line 1) and each model year (column
9) using the design speed (line 3). See Chapter VI for the
correct application of this factor.
• Fill in the ambient temperature correction factor (column 15)
for model year using the ambient temperature (line 8). This
factor is identical for automobiles and light-duty trucks
and is negligible for heavy-duty vehicles [z. .=1.0).
• Insert in column 15 for each model year, the appropriate opera-
ting temperature correction factor (rjn^s) based on the percent
cold starts (line 6) and hot starts (Tine 7). Again, this
factor is not applicable to heavy-duty vehicles.
• Calculate the emission contribution (column 16) of each model
' year by multiplying columns 11 through 15 and writing the
results in column 16.
7-15
-------
• If the pollutant for which the emission rate being calcu-
lated is not hydrocarbons, then add the emission contribu-
tion (column 16) of each model year. The sum is inserted on
line 22, the final average emission factor (en).
• For hydrocarbons, non-exhaust emissions must be considered.
In column 17, fill in the crankcase emission factor appropriate
for the vehicle type and model year.
• In column 18, fill in the evaporative emission factor in
grams per mile for the vehicle type and model year.
• Complete column 19 identically to column 12.
• Add columns 17 and 18 and multiply the sum by column 19 for
each model year. Insert the result in column 20.
• Add columns 16 and 20 for each model year and place in
column 21. This is the model year weighted hydrocarbon
emission factor (eT ).
'i
• The summation of column 21 should be inserted on line 22.
This is the final hydrocarbon emission factor (en).
• This procedure must be carried but for each vehicle class.
2. Worksheet No. 2 (VEM-2)
Worksheet No. 2 is employed to calculate the composite emission
rate (i.e., the weighted average of the emission rates of the four vehicle
classes, AG, LOT, HDG, HDD. The weighting factors, the proportion of VMT
in each vehicle class, are obtained from the VMT-4 worksheet. The emission
rates for each vehicle class are calculated on the VEM-1 worksheet in the
case of carbon monoxide, hydrocarbons, and oxides of nitorgen.
Emissions of particulates and sulfur oxides are relatively
invariant with mode of operation. Only the change to using unleaded gaso-
line in catalytic-equipped vehicles has resulted in a different particulate
emission rate. Table D.l-21, D.2-16, D.4-15, and D.544 of AP-42 present
emission factors of particulates and sulfur oxides for automobiles, LOT,
HDG, and HDD, respectively. The following equation is used to calculate the
particulate emission factor for each vehicle class.
epn
7-16
-------
where,
e = particulate emission factor in year (n) including
p exhaust, tire, and brake wear emissions
P = percent vehicles equipped with catalytic converters in
year (n)
e = total particulate emission rate for vehicles equipped
nc
with catalytic converters
e = total particulate emission rate for vehicles not equipped
with catalytic converters.
The sulfur oxide emission factor can be established directly from AP-42.
• Insert the correct emission rate for the pollutant (line 2)
in column 23. In the case of carbon monoxide, hydrocarbons,
and oxides of nitrogen, this value is line 22 on worksheet
VEM-1 for each vehicle class. In the case of particulate
matter, the above equation is employed. The sulfur oxide
emission rate is obtained from AP-42.
• Insert in column 24 the proportion of VMT by each vehicle
class from worksheet VMT-4, lines 65 through 68.
• Compute the product of columns 23 and 24 and enter in
column 25.
• Sum the values in column 25 and insert the answer on line 26.
This is the composite emission factor for the speed in line
3.
A comprehensive example of the calculations for computing emissions of
carbon monixide, hydrocarbons, and nitrogen oxides is presented in Chapter
IX.
7-17
-------
D. CALCULATION OF EMISSIONS
1. Worksheet No. 1 (EMI-1)
This worksheet is employed to calculate stationary source
emissions in the area of influence. It should be filled out for each fuel
type (i.e., gas, oil) and pollutant, as well as electricity consumption.
• Enter the amount of land use in units of dwelling units,
103 square feet (or square meters) of floor area, or acres
(or square metersi) in Column 1. these values are obtained
from worksheet LUM-3, the product of lines on the worksheet
and the proportion of land use in that category using a
particular fuel type. The proportion of a land use cate-
gory using a particular fuel type should be obtained from
local utulity companies or from national information
[23, 24,].
• Enter the process emission factor in column 2, the space-
heating emission factor in column 4, the space cooling
emission factor in column 6. These values can be found by
employing the tables in Chapter II. The process emission
factor must be adjusted if a time period other than one
year is under consideration. The space heating and space
cooling emission factors must first be multiplied by the
number of degree days or operating hours.
• Total manufacturing land use should be entered on line 17'.
The composite industrial emission factor, obtained from
Chapter II or from Volume II of this report is entered on
line 18.
• Total emissions are calculated and entered on lines 8, 9, 10,
and 19.
2. Worksheet No. 2 (EMI-2)
This worksheet is used to calculate motor vehicle emissions
in the area of influence and in the impact area. It is filled out five
times, once for each pollutant.
• Enter in columns 12 and 13 the speed and amount of VMT in
each category from worksheets VMT-3 and VMT-4.
• Enter in column 14 the emission factor from worksheet VEM-2
for the appropriate speed category.
7-18
-------
• Compute emissions in each, category by taking the product
of columns 13 and 14. Sura the first six lines (line 16)
and the second six lines (line 17).
3. Worksheet No. 3 (EMI-3)
This worksheet summarizes the emissions.
• On the first four lines enter the emissions for each pollu-
tant in columns 21 through 25. These values are obtained
from the EMI-1 worksheets, lines 8, 9, 10, and 19.
• The total stationary source emissions in the area of influence
are summed in line 27.
• Enter the mobile source emissions from lines 15 and 16, work-
sheet EMI-2.
• Take the product of line 26, total electricity consumption,
and the electric utility emission factors in Chapter II.
Enter these products in line 29.
• Total emissions in the area of influence are computed by
adding lines 27 and 15.
• Total emissions are computed by adding lines 27, 29 and 16.
7-19
-------
EXHIBIT 7-1
WORKSHEET RLUM-1
RESIDENTIAL LAND USE MODEL VARIABLE DEFINITIONS
(For year t+0 unless otherwise noted)
AREA OF INFLUENCE
2
Size of Area of Influence in acres (or m ) 1.
Number of dwelling units, excluding major project 2.
Dwelling units per acre (or m ) = (2) * (T) 3.
Distance from center of major project to nearest
Central Business District in miles (or km) 4.
Projected number of limited-access highway
interchanges in 5 years 5.
Number of office employees = 6.
' f\ ~~~™^^^~
Office employment per acre (or m ) = (6) * (Y) 7.
Projected percent developable land area in 10 2
years, excluding major project, in acres (or m ) 9.
Number of manufacturing employees 10.
2
Manufacturing employment per acre (or m ) =
(5) * CD • 11.
Median income level of families and individuals 12.
2
Projected number of acres (or m ) zoned for
residential use in 5 years 13.
Projected percent residential zoned area =
© * © I4-.
COUNTY
Median income level of families and individuals 15.
Median income index - area of influence relative
to county = (fg) * ($5) 16.
Total population 17.
Projected total population in 10 years 18.
2
Area of county in acres (or m ) 19.
2
Projected population growth per acre (or m ) =
( (SI) - . (R)) * (J?) 20.
7-20
-------
EXHIBIT 7-2
WORKSHEET RLUM-2
RESIDENTIAL LAND USE MODEL VARIABLE DEFINITIONS
(For base year, t+0, unless otherwise noted)
Total employment 21.
Projected total employment in 10 years 22.
2
Projected employment growth per acre (or m ) =
( @ - @ ) -i @ 23.
Total licensed automobile drivers 24.
Number of dwelling units 25.
Auto drivers per dwelling unit = (24) * (25) 26.
Auto drivers per acre (or m ) = (24) * (Ti) 27.
METROPOLITAN AREA
Percent vacant office buildings 28.
Projected median income level in 2 years 29.
NEAREST MUNICIPALITY
Projected percent of land which will have public
sewerage available in 5 years 30.
MAJOR PROJECT
Projected total dwelling units in 2 years 31.
Projected total dwelling units 2 years before
project completion 32.
Projected total dwelling units at completion
(ten years from now) 33.
Projected median income level in 2 years 34.
Set equal to 1 if @ •* (29) < 0.85, otherwise = 0 35.
Set equal to 1 if a university exists within 3.23
miles (or 5.20 km) of the center of the major
project, otherwise = 0 36.
Projected number of school age children in 2 years 37._
School age children per dwelling unit = @ * (3l) 38.
O
Area of major project in acres (or m ) 39.
7-21
-------
EXHIBIT 7-3
WORKSHEET RLUM-3
Land Use Categories
Variables
3
20 x
4 x
5 x
V1 7 x
ro
21 x
9 x
23 x
11 x
26 x
h
8910
3.35E9)
6790
2.55E8)
-351
-20300)
-1360
-126000)
L2
6,56
(2.47E8)
-73.7
(-4260)
-200
(-18600)
791
(2.97E8) '
0.0032F
(0.304)
0.0647
(0.00149)
L3
-400
(-1.50E8)
-14.1
(-814)
85.8
(7900)
845
(3.18E8)
601
(2.26E8)
L4
1050
( 3.95E8)
L5
-
40.5
(264000)
-2.79
(-2.79)
-135
(-217)
16
-269
(-1.01E8)-
-11.4
(-658)
97.1
(9020)
L7
L8
191
(7.18E7)
h»
0.00175
(4.02E-5)
ho
-4.42
(-255)
(
hi
-25.5
(-1470)
0.0408
(9.37E-4)
L12
(Metric coefficients are 1n parenthesis)
-------
EXHIBIT 7-4
WORKSHEET RLUM-4
Land Use Categories
Variables
16 x
32 x
33 x
j
> 28 x
14 x
27 x
35 x
36 x
30 x
h
-0.682
(-63.4)
_
41.2
(3830)
L2
L3
L4
L5
46.6
(75.0)
0.00595
(0.00957)
knr-tr ^
L6
-0.0736
(-6.84)
15.0
(1390)
L7
-0.968
(-89.9)
230
(Q CCC7\
-150
f-1396fi)
L8
0.0196
(1.82)
L9
60.2
(5590)
ho
202
(18800)
hi
=•
2.46
(229)
L12
-------
EXHIBIT 7-5
WORKSHEET RLUM-5
Land Use Categories
Variables
38 x
39 x
£ Columns
+ Constant
TOTAL LAND USE
h
7200
(669000)
L2
1380
(128000)
L3
355
(33000)
L4
761
(70700)
L5
78.8
(127)
L6
488
(45300)
L7
81.7
(7590)
L8
-23.9
(-2200)
L9
2.54
(236)
ho
-14.7
(-1370)
hi
184
(17100)
244
(22700)
L12
0.103
(417)
-33.5
(-136000)
(Metric coefficients are 1n parenthesis)
-------
EXHIBIT 7-6
WORKSHEET RLUM-6
— •. "' '• .'• '
RESIDENTIAL LAND USE MODEL
PROJECTED DISAGGREGATION OF LAND USE IN AREA OF INFLUENCE
AT PROJECT COMPLETION
Residential Proportions
Single family detached homes
Single family attached homes
Mobile homes
Multifamily low rise structures
Multifamily high rise structures
(Note@ +
must = 1
Commercial
<50,000 ft2
50-100,000 ft2
>100,000 ft2
(Note @5) + © + @ must = 1)
FINAL PROJECTED LAND USE
In Dwelling Units
Single Family Attached =
Single Family Detached =
Mobile Homes
Multifamily Low Rise
Multifamily High Rise =
In 1,000 ft2 (m2)
Commercial <50K =
Commercial 50-1OOK
Commercial >100K
Office
40.
1L
42.
43.
44.
45.
46.
47.
48.
4SL
50.
5JL
52.
54,
55.
56.
Default Values
,61
,03
,06
.29
.01
.51
.20
.29
7-25
-------
EXHIBIT 7-7
WORKSHEET RLUM- 7
RESIDENTIAL LAND USE MODEL
Manufacturing =(C) 57.
Wholesale-Warehousing =(tfi) 58.
Hotels, Motels =U 59^
Hospitals
Cultural Facilities =& 61.
Churches = £7^ 62.
Educational Facilities = K\ 63.
Recreational Facilities = /CT^ 64.
In miles (km)
Non-expressway highway =(Q] 65.
lane distances
7-26
-------
EXHIBIT 7-8
WORKSHEET IOLUM-1
INDUSTRIAL/OFFICE LAND USE MODEL VARIABLE DEFINITIONS
AREA OF INFLUENCE
2
Size of Area of Influence in acres (or m ) l._
Number of dwelling units 2._
Dwelling units per acre (or m ) = (F) * (T) 3._
O
Number of developable acres (or m ) 4._
Percent developable area = (7) * (V) 5._
Vacant dwelling units 6._
Percent vacant housing = (if) * (F) 7._
Distance from center of Major: Project to nearest
Central Business District in miles (or km) 8.
2
Projected number of acres (or m ) zoned for
commercial use in 5 years 9.
2
Projected number of acres (or m ) zoned for office
use in 5 years 10.
2
Projected number of acres (or m ) zoned for
industrial use in 5 years 11._
Projected percent commercial zoned area = (V) * (T) 12._
Projected percent office zoned area = (KJ) * (T) 13._
Projected percent industrial zoned area = (Tl) * (l)l4._
Number of office employees 15.
Office employment per acre (or m ) = Q5) * (T) 16._
Railroad mileage (or km) 17..
Number of manufacturing employees 18.
2 .'
Manufacturing employment per acre (or m ) =
19.
Projected number of limited access highway
interchanges in 5 years '20.
Number of wholesale and warehouse employees 21.
2
Wholesale-warehouse employment per acre (or m ) =
(2l) •» (T) 22.
2
Nonhousehold population per acre (or m ) =
Nonhousehold population 23.
24..
7-27
-------
EXHIBIT 7- 9
WORKSHEET IOLUM-2
INDUSTRIAL/OFFICE LAND USE MODEL
VARIABLE DEFINITIONS
Median income level of families and individuals 25.
Current U.S. average income level for families
and individuals 26.
Median income index-relative to U.S. average =
(25) * @ 27.
COUNTY
Median income level of families and individuals 28.
Median income index-area of influence relative
to county = (£§)* (|8) 29.
Total employment 30.
Projected total employment in 10 years 31.
Projected employment growth rate = ((3(3)-(|o)) *(30)32.
Total licensed automobile drivers 33.
2
Area of county in acres (or m ) 34.
1 ^_< ^_i^
Auto drivers per acre (or m ) = (33) * (34) 35.
METROPOLITAN AREA
Percent vacant office buildings 36.
Cost of 1500 kWh of electricity (commercial rate) 37.
Average U.S. cost of 1500 kWh of electricity
(commercial rate) 38.
Energy cost factor = (3^) * (3§) 39.
NEAREST MUNICIPALITY
Projected percent of land which will have public
sewerage available in 5 years 40.
7-28
-------
EXHIBIT 7-10
WORKSHEET IOLUM-3
INDUSTRIAL/OFFICE LAND USE MODEL VARIABLE DEFINITIONS
MAJOR PROJECT
Projected total employees 2 years after project
Initiation 41..
Projected total employees at project completion 42.
Land area of completed major project in acres
(or m2) 43..
Set equal to I if a private school exists within
3.23 miles (or 5.20 km) of the center of the major
project 44._
Size of floor area of completed major project in
= 43.56 x (43) (English units)
= $3) (Metric units) 45.
1,000 ft2 (or m2) = 43.56 x fos) (English units)
7-29
-------
EXHIBIT 7-11
WORKSHEET IOLUH-3
Land Use Cateqories
Variables
3 x
5 x
36 x
7 x
0 x
l* x
ifi
4? x
17 x
n x
h
2480
(9.32E8)
2.05
(0.0471)
563
(52300)
-128000
(-1.19E7)
-406
(-23400)
L2
669
(3.27E8)
119
(11100)
-2090
(-7.86E8J
0.0553
(b.14)
L3
-0.0273
(-2.54)
11.5
(664)
68.9
(6400)
L4
0.252
(0.00579)
L5
19.6
(128000)
0.00385
(1.53E-6)
1.70
(2.74)
L6
726
(2.73E8)
L7
-11.5
(-664)
-249
(9.36E7)
L8
-283
(1.06E8)
L9
ho
0.0314
(721E-4)
8.65
(499)
Lll
0.0974
(0.00224)
34.8
(2010)
L12
17.6
(71200)
14.8
(37200)
I
CO
o
(Metric coefficients are in parenthesis)
-------
EXHIBIT 7- 12
WORKSHEET IOLUM-4
Land Use Categories
Variables
29 x
19 x
32 x
27 x
£ 20 x
~^
22 x
'14 x
40 x
24 x
39 X
Ll
L2
L3
507
(47100)
254
(9.55E7)
L4
10100
(3.80E9)
-2620
(-253E5)
L5
-6.08
(-9.78)
-
L6
7470
(2.81E9)
90.8
(8440)
11.4
(1060)
L7
1180
(4.44E8)
L8
443
(1.67E8)
478
(1.8E8)
L9
-34.6
(-3210)
L10
-1.07
(-99.4)
hi
L12
1.88
(3.08E9)
1440
(2.36E10)
387
(1.57E6)
(Metric coefficients are in parenthesis)
-------
EXHIBIT 7-13
WORKSHEET IOLUM-5
Land Use Categories
Variables
30 x
43
44
41
35 x
£ Columns
+ Constant
TOTAL LAND USE
h
-9530
(-885000)
L2
-838
(77900)
L3
-326
(-30300)
L4
1120
(104000)
:L5
-25.7
(-41.4)
L6
-1650
(-153000)
L7
145
(13500)
L8
-20.5
(-1900)
L9
0.00004
(0.00372)
0.0411
(9.44E-4)
12.6
(1170)
-
23.6
(2190)
L10
-0.0146
(-1.36)
-148
(-13700)
hi
0.0802
(7.45)
-925
(-85900)
L12
-615
(1.01E10)
-604
(-2.44E6)
I
OJ
no
(Metric coefficients are in parenthesis)
-------
EXHIBIT 7-14
WORKSHEET IOLUM-7
INDUSTRIAL/OFFICE LAND USE MODEL
PROJECTED DISAGGREGATION OF LAND USE IN AREA OF
INFLUENCE AT PROJECT COMPLETION
Residential
Single family detached homes
Single family attached homes
Mobile homes
Multifamily low rise structures
Multifamily high rise structures
(Note© +(46) +@ +(4§) +@)
must = 1 )
Commercial
<50,000 ft2
5(5-100,000 ft2
> 100, 000 ft2
(5
must = 1)
45_.
46.
47.
ii:
49.
50.
51_.
52.
Default Values
.68
.02
.06
.23
.01
.66
.14
.20
FINAL PROJECTED LAND USE
In Dwelling Units
Single Family Attached =
Single Family Detached =
Mobile Homes =
Multifamily Low Rise
Multifamily High Rise =
In 1.000 ft2 (m2)
Commercial <50K =
Commercial 50-100K =
Commercial >100K =
Office
Manufacturing =
Wholesale-Warehousing =
Hotels, Motels
Hospitals =
Cultural Facilities
Churches
Educational Facilities =
Recreational Facilities1
In miles (km)
Non-expressway highway =
lane distances
58.
54.
55.
56_.
57.
58.
60.
62.
63.
65.
66.
67.
68.
69.
70.
7-33
-------
EXHIBIT 7-15
WORKSHEET LUM-1
CONFIDENCE INTERVALS FOR PREDICTED LAND USE
Final Projected Land Use Category
corresponds to Dependent Var1
PREDICTOR VARIABLE
Name
able Name
1.
2.
3.
4.
5.
6.
and General Category
COMPUTING VARIANCE OF DEPENDENT VARIABLE
§xQ) x Covarlance = 7.
x £2) x
x 0 x
8x © x
x © x
© x © x
© x © x
1 C, I A V y "
© x © x
0 x © x
0 x © x
(2) x © x
V 1 \^ J
§X © X
x O x
X Up X
© x © x
(4) x (Dx
Covariance
Covariance
Covarlance
Covariance
Covariance
Covariance
Covarlance
Covarlance
Covarlance
Covariance
Covariance
Covariance
Covarlance
Covarlance
Covariance
Covariance
Covariance
Covariance
Covariance
Covariance
Covarlance
Covariance
Covarlance
= 8.
= 9.
= 10
= 11
= 12
= 13
= 14
= 15
« 16
= 17
= IB
= 19
= 20
= 21
= 22
= 23
= 24
= 25
= 26
= 27
= 28
= 29
= 30
.
.
.
.
.
.
.
.
.
.
.
.
t
t
7-34
-------
EXHIBIT 7- 16
WORKSHEET LUM-2
CONFIDENCE INTERVALS FOR PREDICTED LAND USE
COMPUTING VARIANCE OF DEPENDENT VARIABLE
fs) x (T) x Covariance _ 31.
x (2) x Covariance =32.
x 13J x Covariance =33.
x (4) x Covariance =34.
I5)x (5J x Covariance = 35.
x (6j x Covariance = 36.
r6) x (1 ) x Covariance = 37.
16) x (2) x Covariance = 38.
(6) x (3) x Covariance = 39.
16) x (4) x Covariance = 40.
(6} x (5) x Covariance = 41.
(6)x (6) x Covariance = 42.
Sum(7)through @ = 43.
Standard Deviation of Dependent Variable
F - statistic of predictive equation 45.
t - statistic of predictive equation =
'(IB) 46.
If Final Projected Land Use is Residen-
tial or Commercial, set equal to the
disaggregation percentage 47. IQQ
^Confidence Interval = (46) x ^4) x(47)
T 100 48.
7-35
-------
EXHIBIT 7-17
WORKSHEET LUM-3
FINAL LAND USE MODEL CALCULATIONS
MAJOR PROJECT
If Residential, total projected dwelling units at
project completion, by the following types:
Single Family Detached
Single Family Attached
Mobile Homes
Multifamily Low Rise
Multifamily High Rise
If Industrial/Office, total projected land area in
1,000 ft'2 (or m2), by the following types:
Office
Manufacturing
Wholesale-warehousing
TOTAL PROJECTED LAND USE (Including Major Project)
In Dwelling Units
Single Family Detached =
Single Family Attached =
Mobile Homes =
Multi family Low Rise =
Multifamily High Rise =
In 1.000 ft2 (or m2)
Commercial <50K =
Commercial 50-TOOK
Commercial >100K =
Office
Manufacturing
Wholesale-warehousing =
Hotels, Motels
Hospi tals
Cultural Facilities =
Churches =
Educational Facilities =
Recreational Facilities2
In Miles (or km)
Nonexpressway highway =
49.
50.
51.
52.
53.
54.
55.
56.
RLUM-6*
vO)
'§)
(57)
(60)
(g)
or
or
or
or
or
or
or
or
or
or
or
or
or
or
or
or
or
or
IOLUM-6*
@
I
@
(58)
(§)
.
§'
©
From Above
+ (49) 57.
+ (50) 58.
+ (51) 59.
+ (52) 60.
+ (53) 61.
62.
63.
64.
+ (54) 65.
+ (£5) 66.
+ ^6) 67.
68.
69.
70.
71.
72.
73.
or
(70))
74.
Use of one the two depends on project type. If the major project is
residential, obtain values from RLUM-6. If the major project is
industrial or office, obtain values from IOLUM-6.
7-36
-------
EXHIBIT 7-18
WORKSHEET VMT-1
VEHICLE TRIPS CALCULATION
COMPUTATION SHEET (Numbers tn circles Indicate previous numbered data
entries or computed values.)
Radius
R, Effective radius of study area 1.
Trip Calculation
Single Family Detached
Single Family Attached
Multi family Low Rise
Multifamily High Rise
Mobile Home
Hotel, Motel
Commercial, < 50, 000 sq.ft.
Commercial, 50,000-100,000
sq.ft.
Commercial >100,000 sq.ft.
Office
Manufacturing
Whol esal i ng-Warehousing
Cultural
Churches
•lospitals
Educational Facilities
Recreation
2
Amount of
Land Use
Lt
: 3
: Work
Trip Rate
; T!
4
Work Trips
7.
8.
9.
10.
n.
12.
19.
20.
. 5
Other
Trip Rate
T2
6
Other Trips
(2) x (5)
13.
T4.
15.
16.
17.
18.
21.
22.
Total Trips
24.
7-37
-------
UNCORRECTED VMT
EXHIBIT 7-19
WORKSHEET VMT-2
uncorrected VMT
x.
x Lr CD
r ^
X L_4.u««
o tner
x L_ CD
= 26.
= 27.
= 28.
= 29.
VMT1,
VMTA,
VMT1,
VMTA,
work
work
other
other
VMT CORRECTION
Total residential work trips =(?)+®+©+©
^—+
Proportion of residential work trips less thanQl^
Work trips correction =(5(J)x(3l)
Work VMT correction =@x®
Total residential other trips -
f "
Proportion of residential other trips less thanvj.
Other trips correction =@)x(|5)
Other VMT correction = ^|) x(l
= 38.
VMT1, work
= 39, VMTA, work
= 40.
= 41.
VMT1, other
VMTA, other
Peak Hour Proportion
proportion of VMt in peak hour
proportion of VMT in off peak hours
Facility Classification
30.
31.
32.
33.
35.
36^
37.
= 42.
= 43.
(Default = .10)
(Default = .90)
Enter the proportion of distance on each facility for work and other
trips.
local streets
arterials
expressways
44.
45.
46.
47.
48.
49.
7-38
-------
IMPACT AREA
(38
©
x©
x©
x
x
x(4)
AREA OF INFLUENCE
39)
x(42)
x©
x©
x©
x@
x@
x
EXHIBIT 7-20
WORKSHEET VMT-3
x(42) x@
X&6)
x
x
x
x
.x
x
x
x
X©
X©
x(47)
50.
51.
52.
53.
54.
55.
56.
58.
VNTT1 off peak, local streets
VMT off peak, arterials
VMT off peak, expressways
VMT , peak, local streets
VMT , peak, arterials
VMT , peak, expressways
VMTM, off peak, local streets
57. VMT , off peak, arterials
59.
VMT , off peak, expressways
VMT , peak, local streets
7-39
-------
EXHIBIT 7-21
WORKSHEET VMT-4
(39)
(4*T)
x(41)
x@
x@
x(4?)
x (4
x(49)
60.
61.
VMT , peak, arterials
VMT , peak, expressways
Vehicle Classification Proportions
Total Manufacturing and Warehousing Trips = (19) + (20) + (21) + (22) = 62_.
= 63,
If(63)> 0.07, Then (63;- .05 = 64.
If(63)< 0.07, Then
= 65.
= 66.
0.804 -(64)
0.118
0.046 + .8 x@
0.062 + .2 x(64)= 68.
= 67.
Average Route Speeds:
local streets
arterials
expressways
peak
69.
70.
71.
64 = 0
automobile
light duty truck, gasoline
heavy duty vehicle, gasoline
heavy duty vehicle, diesel
offpeak
72.
73.
74.
7-40
-------
EXHIBIT 7-22
WORKSHEET VEM-1
Vehicle Type = 1.
Pollutant = 2.
Speed = 3_.
Design Year
Region = 5.
4.
Cold Starts = 6.
(Gasoline Automobile, Light-Duty Trucks, Heavy-Duty Gasoline Vehicles, Heavy-Duty Diesel Vehicles)
(Carbon Monoxide, Nitrogen Oxides, Hydrocarbons)
mph
Ambient Temperature = 8.
(Low Alt., High Alt., Calif)
%, Hot Starts = _7_.
°F
10
Vehicle
Age
(years)
1
2
3
4
5
6
7
S
9
10
11
12
>13
9
Model
Year
11
Base
Emission
Rate
Cipn
12
Fraction
of
Annual
Travel
min
13
Speed
Correction
Factor
Vips
14
Ambient
Temperature
Correction
Factor
Zipt
15
Operating
Temperature
Correction
Factor
riptws
16
Model year
Emission
Contribution
eipntwx
17
Hydro-
carbon
Crankcase
Emissions
fi
18
Hydro-
carbon
Evaporative
Emissions
ei
19
Fraction
of
Annual
Travel
min
20
Model
Year
Hydro-
carbon
Emissions
eHCi
21
Model
Year
Total
Emissions
6Ti
Note: When calculating carbon monoxide and nitrogen oxide emissions, e
HCi
= 0.
Average emission factor 32
-------
EXHIBIT 7- 23
WORKSHEET VEM-2
CALCULATION OF THE COMPOSITE EMISSION FACTOR
Pollutant = 2_._
Speed = 3.
Design Year = 4.
Region = 5.
Cold Starts = 6.
mph
Ambient Temperature = 8.
(low alt., high alt., Calif.)
%, Hot starts = ]_.
°F
Vehicle Class
Automobiles
Light-duty Trucks
Heavy-duty Gasoline Vehicles
Heavy-duty Diesel Vehicles
24
23 Vehicle Class
Emission Rate Weighting
25
Product
•
!
Composite emission factor = 26
7-42
-------
EXHIBIT 7-24
WORKSHEET EMI-1.
I
-p>
CO
LAND USE CATEGORY
residential single family attached
residential single family detached
residential mobile homes
residential multifamily low rise...
residential multifamily high rise..
commerci al <50K
commercial 50-100K
commerci al >1 OOK
office
wholesale-warehousing
hotel s , motel s
hospi tal s
cultural facilities
churches
educati onal f a ci 1 i ti es
recreati onal f aci 1 i ti es
TOTAL EMISSIONS
1
amount
2
process
emission
factor
3
process
emissions
8.
process
4
space
heating
emission
factor
5
space
heating
emissions
9.
space-
heating
6
space
cooling
emission
factor
7
space
cooling
emissions
TO.
space-
cooling
manufacturing
land use
x 18.
= 19.
(See page 7-18)
-------
EXHIBIT 7-25
WORKSHEET EMI-2
MOTOR VEHICLE EMISSIONS
12
SPEED
12
VMT
14
EMISSION
FACTOR
15
EMISSIONS
VMT , off peak, local streets
VMTA, off peak, arterials
a
VMT , off peak, expressways
VMTA, peak, local streets
VMr\ peak, arterials
a
VMT , peak, expressways
VMT , off peak, local streets
VMT , off peak, arterials
VMT , off peak, expressways
VMT , peak, local streets
VMT , peak, arterials
VMT , peak, expressways
AREA OF INFLUENCE TOTAL 16
IMPACT AREA TOTAL 17
7-44
-------
8 process emissions
20
PM
EXHIBIT 7-26
WORKSHEET EMI-3
EMISSIONS SUMMARY
21
SOX
22
CO
23
HC
24
NOX
25
Electricity
9 space heating emissions
10 space cooling emissions
19 industrial emissions
•P»
cn
27 STATIONARY SOURCE TOTAL
AREA OF INFLUENCE
area of influence
15 motor vehicle emissions
26.
28 TOTAL, AREA OF INFLUENCE
26 x emission factor =
29 electric, utility emissions.
16 motor vehicle, impact
area emissions
30 TOTAL*
* Area of influence stationary sources, secondary sources (i.e., electrical generation), and motor
vehicle emissions in impact area.
-------
VIII. EXAMPLES OF USE OF COMPUTATION WORKSHEETS
This chapter presents examples of the use of the computation worksheets
from the previous chapter. The first section illustrates the use of the
land use model worksheets to compute the predicted land use for a residential
4
major project. The second section illustrates the estimation of VMT for
the same example case study. The third section illustrates the computation
of a motor vehicle emission factor for one pollutant, hydrocarbons. The
final section illustrates the emissions computation in the final three
worksheets.
A. EXAMPLE OF LAND USE MODEL CALCULATIONS
This section presents an example of the use of the land use model
worksheets using data from one of the Residential major projects in this
study, namely, Northglenn, Colorado. This planned residential development
was initiated in 1960 and completed in 1970 with a final size of 7,000
dwelling units. The worksheets are filled in English units for this example.
The computation of confidence intervals is shown for only one of
the projected land uses.
8-1
-------
COUNTY
Median income level of families and individuals 15.
Median income index - area^ of influence rela-
tive to county =© ^ 16.
EXHIBIT 8-1
WORKSHEET kLUM -1
RESIDENTIAL LAND USE MODEL VARIABLE DEFINITIONS
AREA OF INFLUENCE
Sfze of Area of Influence in acres (m2) 1. \OjOOO
Number of dwelling units, excluding major project L 3SO
Dwelling units per acre (m2) =(2^© 3- Q
Distance from center of major project to
nearest Central Business District in miles
f(km) 4.
Projected number of limited-access highway
,interchanges in 5 years jr.
Number of office employees = JL
Office employment per acre (m2) =(6)KD Zj
Projected developable land
area in 10 years, excluding major q «
project, in acres (m2) — P)
Number of manufacturing employees JOj
Manufacturing employment per acre (m2) =(9}-r(T) 11.
Median income level of families and individuals 12.
Projected number of acres (m2) zoned for resi-
dential use in 5 years 13. ~7f 600
Projected percent residential zoned area =(Tj) rCv) 14.
Total population 17. /£ O^ ff
Projected .total population in 10 years 18. /££* t 7
Area of county in acres (m2) 19. 7^6 ?CO
' ?
Projected population growth per acre (m )
= (flS- n7)) v (fa 20. £-0
8-2
-------
EXHIBIT 8-2
RESIDENTIAL LAND USE MODEL VARIABLE DEFINITIONS
WORKSHEET RLUM-2
Total employment 21.
Projected total employment in 10 years 22.
Projected employment growth per acre (m2) ?3
Set equal to 1 if (34) ^\£&)< 0.85 35_.
Set equal to 1 if a university exists
within 3.23 miles (5.20 km) of the
center of the major project 36.
Projected mumber of school age chil-
dren in 2 years 37.
SchopJage children per dwelling unit
= (37) - (3?) 38.
Area of maj cr project in acres (m ) 39.
8-3
Total licensed automobile drivers 24. 35^005"
Number of dwelling units 25. S \ ^Sl
Auto drivers per dwelling unit £7A
per acre (I"2) 27. a,M3?
METROPOLITAN AREA
Percent vacant office buildings 28. "7* ^
Projected median income level in 2 years 29. «!&
NEAREST MUNICIPALITY
Projected percent of land which will
have public sewerage available in
5 years 30. _ I OQ
MAJOR PROJECT
Projected total dwelling units in 2
years 31. 3
Projected total dwelling units 2 years ^
before project completion 32. _ fe>; OOP
Projected total dwelling units of com-
pletion 33.
Projected median income level in 2 years 34.
-------
EXHIBIT 8-3
WORKSHEET RLUM-3
Land Use Categories
Variables
(7) 0,035.
(20) O.C823*
(T) 12-2 x
(T) a x
(7) 0 x
(21) 40j*>3£ x
r^ ^7x
@ 0,$3fcOx
(TT) Ox
N^
@ 0,tf£0 x
Ll
8910
(3.35E9)
Sl^
6790
(2.55E8)
5^7
-351 '
(-20300)
'^80
-1360
(-126000)
-9aao
L2
656
(2.47E8)
&3.t>
-73.7
(-4260)
-8?^
-200
(-18600)
-4 00
791
(2.97E8)
O
0.0032,'"
(0.304)
ISO
0.0647
(0.00149)
530
L3
-400
(-1.50E8)
-14
-14.1
(-814)
-173s
85.8
(7900)
C7Z
845
(3.18E8)
O
601
(2.26E8)
ai.6
L4
1050
( 3.95E8)
O
L5
40.5
(264000)
3^33
-2.79
(-2.79)
-34,0
-135
(-217)
-qi.fc
L6
-269
(-1.01E8)
-8
191
(7.18E7)
0
^
0.00175
(4.02E-5)
14.3
L10
-4.42
(-255)
-523
hi
-25.5
(-1470)
-311
0.0408
(9.37E-4)
334
L12
c»
-p.
-------
EXHIBIT 8-4
WORKSHEET RLUM-4
Land Use Categories
Variables
x
Li . !
-0.682
(-63.4)
— 477O
•
41.2
(3830)
4IAS
L2
L3
"•A i
L5
46.6
(75.0)
4%.^
0.00595
(0.00957)
35.?
L6
-0.0736
(-6.84)
~5lS
15.0
(1390)
io9>
*
L7
-0.968
(-89.9)
-73. C,
230
in ecn\
lO,ODt/J
/br >
-150
\ " 1 ijjljj/
o
L8
0.0196
(1.82)
IS?
\
h>
60.2
(5590)
ti
L10
202
(18800)
309
hi
2.46
(229)
24fe
L12
00
en
-------
EXHIBIT 8-5
WORKSHEET RLUM-5
1
Variables
@ $.3 x
fi) I,ft03 x
• C Columns
+ Constant
TOTAL LAND USE
h
-fe77?
7200
(669000)
4*1
! Land Use Categories
L2
-fc\t
1380
(128000)
764
L3
7,0,
355
(33000)
3Q>3
L4
0
761
(70700)
7£l
L5
-38,5
78.8
(127)
40.3.
L6
-3fe)
488
(45300)
131?
L7
-63,5
81.7
(7590)
l^.3k
L8
\3f
-23.9
(-2200)
U3
L9
14.3
2.54
(236)
|fc,S
L10
ISS
-14.7
(-1370)
140
hi
184
(17100)
4a3
fe?a
244
(22700)
?36
L12
0.103
(417)
(86
Ifife
-33.5
(-136000)
/53
00.
-*
-------
EXHIBIT 8-6
WORKSHEET RLUM-6
RESIDENTIAL LAND USE MODEL
PROJECTED DISAGGREGATION OF LAND USE IN AREA OF INFLUENCE
AT PROJECT COMPLETION
Residential
Single family detached homes
Single family attached homes
Mobile homes
Multifamily low rise structures
Multifamily high rise structures
@ + © + @ + © +
must = 100)
Commercial
<50,000 ft2
50-100,000 ft2
>100,000 ft2
(Note 6$) + $6) +tf
must = 100)
FINAL PROJECTED LAND USE
In Dwelling Units
Single Family Attached =
Single Family Detached =
Mobile Homes =
Multifamily Low Rise =
Multifamily High Rise =
In 1.000 ft2 (m2)
Commercial <50K =
Commercial 50-1OOK
Commercial >100K =
Office
40.
41.
42.
43.
44.
45.
46_.
47.
Default Values
61.
3.
6.
29.
1.
51.
20.
29.
43. 3fc*5
49.
4
50. 4
51.
52.
30
O
53. afS
54. 61
55. 4O&~
56.
36^
8-7
-------
8-7
WORKSHEET RLUM-7.
RESIDENTIAL LAND USE MODEL
Manufacturing =(Q 57. 7£> I
Wholesale-Warehousing '- 58. \3t~p
lane distances
Hotels, Motels = (E) 59. 1%.
Hospitals =^S 60. U 3
Cultural Facilities =?Q 61. 16*
Churches = r^ 62.
Educational Facilities = r 63.
Recreational Facilities = /cT 64.
In miles (km)
Non-expressway highway =(Q} 65.
8-8
-------
EXHIBIT 8-8
WORKSHEET LUM-1
CONFIDENCE INTERVALS FOR PREDICTED LAND USE
Final Projected Land Use Category Coww\g.CC'oH >\QO>OQO
corresponds to Dependent Variable Name CO MM and General Category .
PREDICTOR VARIABLE
Name
i. O
i sir
2.
3.
4.
5.
6.
13.3
O.O35"
COMPUTING VARIANCE OF DEPENDENT VARIABLE
CD xj) x Covarlance £32.97.3 - 7.
i x 2) x Covarlance "7339
I1?-
8.
9.
10.
x \£) x CovaHance ^
x 0 x CovaHance *Zl 3 3d
x (§) x CovaHance —0. Dfe4fe - 11.
x (?) x Covarlance .
x (T) x CovaHance _
x (?) x CovaHance
12.
13.
O
0
to'
0
x © x CovaHance _ -%b4.7 • 15. -"
x 0 x CovaHance I3»fefc « 16. "3~Zj
x © x CovaHance • ~fl.QQg.Sfe7 • 17. - Id3 I
x (O x Covarlance Q.*»,7/d • 1'8. 371 101
19.
^) x ([) x CovaHance ~_
2) x @ x Covarlance -854.?
fa5 x CD x Covarlance 1^0777 - 21.
20.
x (50 x
x (6J x CovaHance
(4) x Q3 x CovaHance
f?) x ^) x CovaHance
Covarlance - O.D4I»I =
25.
26.
x CovaHance ~i
x CovaHance
4 x (& x CovaHance
x (&}x Covarlance
-365"
x Covarlance -11758 • 22. -
* 24, -
27. - I , 66 3
28.
29.
30.
8-9
-------
EXHIBIT 8-9
WORKSHEET LUM-2
CONFIDENCE INTERVALS FOR PREDICTED LAND USE
COMPUTING VARIANCE OF DEPENDENT VARIABLE
x © x Covariance -0> Ofe4'85 _ 3K 0
x \z) x Covariance -D^Ota^? = 32. -1,431
x x Covariance -0. Ofelfrl = 33.
x x Covariance 0»003"7lo = 33. 37, 10 I
© x © x Covariance - ? ^ = 39. - 513
© x © x Covariance D.?,4o = 40. _3
x x Covariance 0^00060 74^ = 41.
x x Covariance 0^00^55 = 42. I 39, 077
Sum©through@ = 43. 43 / 5M
Standard Deviation of Dependent Variable
* " 44.
F - statistic of predictive equation 45. /Q.O636
t - statistic of predictive equation =
46. 3J7P
If Final Projected Land Use is Residen-
tial or Commercial, set equal to the
disaggregation percentage • 47. -IQQ 53
+ Confidence Interval = (£E) x(44) x(47) ,
~T 100 48. IJO^l
8-10
-------
EXHIBIT "8-10
WORKSHEET LUM-3
FINAL LAND USE MODEL CALCULATIONS
NUOR PROJECT
If Residential, total projected dwelling
units at project completion, by the
following types:
Single-family Detached
Single-family Attached
Mobile Homes
Multlfamily Low Rise
Multlfamily High Rise
If Industrial/Office, total projected
land area in 1,000 ft2 (m2), by
the following types-
Office
Manufacturing
Whol esal e-warehousi ng
TOTAL PROJECTED LAND USE (including Major Project)
49.
50.
51.
52.
53.
7 coo
0
0
0
0
In dwelling units
RLUM-6 or IOLUH-6
Single Family Detached = (
Single Family Attached =
Mobile Homes
MJHi family Low Rise =
Hilt1fam11y High Rise -
In 1.000 ft2 (m2)
Commercial <50K
Commercial 50-1OOK
Commercial >100K
Office
Manufacturing
Wholesale-Warehousing
Hotels, Motels
Hospitals
Cultural Facilities
Churches
Educational Facilities
Recreational Facilities
In Miles (km)
Non-expressway highway
lane distances
((65) or
54.
55.
56.
57,
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
=1
O
/a?
8-n
-------
B. EXAMPLE OF TRAFFIC MODEL CALCULATIONS
Using the projected land uses shown on Worksheet LUM-3, page 8-11,
the VMT in both the area of influence and impact area are computed. If any
development existed in the area of influence in the base year, it would
first be subtracted from the values on LUM-3 before they are entered on
VMT-1.
1. Worksheet VMT-1
The amount of land use is copied from Worksheet LUM-3 and
then multiplied by trip generation rates obtained from Table 5-2.
2. Worksheet VMT-2
The work and other trip lengths are computed via the equations
on pages 5-9.
T = 0.003 * p°'20 * S^'
= 0.003 * 12297980'20 * 301 AB
L1 = 0.003 * 16.52 * 158.82
Li = 7.87
L2 = .5(0.003 * p°'18 * S21>4° +.0.003 * p°'26 * S21>25)
L2 = .5(0.003 * 12297980'18 * 301 AQ + 0.003 * 12297980'26
* 301'25)
L2= .5(0.003 * 16.52 * 116.94 + 0.003 * 16.52 * 70.21)
L2= .5(5.80 + 3.48)
L2= 4.64
8-12
-------
The calculation of the corrected VMT is self-explanatory.
The default peak hour proportion is used. The facility classifications
are calculated using Tables 5-3 and 5-4 as discussed on page 5-13. Rounding
off the trip lengths calculated previously to eight and five miles respec-
tively, the example on page 5-16 is appropriate.
3. Worksheet VMT-3 and VMT-4
The calculations, though tedious, are self explanatory. The
default average route speeds discussed on page 5-18 are employed.
8-13
-------
EXHIBIT 8-11
WORKSHEET VMT-1
VEHICLE TRIPS CALCULATION
COMPUTATION SHEET (Numbers tn circles Indicate previous numbered data
entries or computed values.)
Radius
R, Effective radius of study area 1
Trip Calculation
Single Family Detached
i
Single Family Attached
i
Multi family Low Rise
!
Multi family High Rise
Mobile Home
Hotel , Motel
Commercial, <50,000 sq.ft.
Commercial, 50,000-100,000
sq.ft.
Commercial >100,000 sq.ft.
Office
Manufacturing
Wholesaling-Warehousing
Cul tural
Churches
I
Hospitals
Educational Facilities
Recreation
2
Amount of
Land Use
4
-7813
y
30
o
i
18.2*
Z18
a
Z
O.2>
1,8
o.l
o
&
o
It
?
V
o
0
H*
0
O .
4
Work Trips
7. /Y09?
8. (f>
9. Jfe
10. 0
n. "7
12. 1
0
0
O
5868
19. 38of
20. **8
£>
O
/808
5
Other
Trip Rate
T2
7 .
7
6
y
5"
10
)3o
fca
io
o
0
0
ol
A
o
V
10
6
Other Trips
(2) x (5)
13. 70^1
14. l&
15. /^O
16. • 0
17. 10
18. I8t
38W
i8&)
/&>2oi>
0
21. 0
22. 0
3V
ZSO
o
S'HM
\$3V
Total Trips
23.
8-14
-------
:EXHIBIT,6-12 .
WORKSHEET VMT-2
UNCORRECTED VMT
uncorrected VMT
VMT1. work
x L
iothcr v.
27. 56/67 m*> work
28.^1^22 VMT1, other
29.303930 VMT*, other
VMT CORRECTION
Total residential work trips =(?)+(8)+(D+ © + @ + ©
Proportion of residential work trips less than(T)
Work trips correction = C
Work VMT correction =
Total residential other trips = < + + + +
Proportion of residential other trips less than(
Other trips correction = (^ x i|5)
Other VMT correction =
= 38.
* 39.
40.
-
. work
VMTA. work
other
VMTA. other
Peak Hour Proportion
30.
31.
32.
33.
34. 70 W 7
35.
37.
proportion of VMT in peak hour
proportion of VMT in off peak hours
Facility Classification
= 42.
= 43.
(Default = .10)
(Default = .90)
Enter the proportion of distance on each'facility for work and other
triPs- . Work / '•-' .Other
local streets 44.
arterials 46.
expressways 48.
45.
47. .5S
49.
8-15
-------
IMPACT AREA
AREA OF INFLUENCE
x() v
x() .
x .
EXHIBIT,,8-13
WORKSHEET VMT-3
.06,
VMT1 off peak, local streets
VMT1 off peak, arterial s
VMT1 off peak, expressways
53. 6g^8 VMT, peak, local streets
54. H0^^7 VMT1, peak, arterial s
55.2.qOt>£ VMT1, peak, expressways
.10
VMT, off peak, local streets
57. /V/2?7 VMTA, off peak, arterial s
58. ^aagS VMl^, off peak, expressways
.to
59. a<>80 VMTA, peak, local streets
8-16
-------
EXHIBIT 8-14
WORKSHEET VMT-4
x4
60. <£&?£ VMTA, peak, arterial s
JO
61. /6£S3 VMTA, peak, expressways
Vehicle Classification Proportions
Total Manufacturing and Warehousing Trips. = 19 + 20 + 21 + 22 = 62. V3/3
62 T 23 + 24 = 63. .03
If 63 > 0.07, Then .......... . ...... 63 - .05 = 64.
If 63 <^0707, Then ..................... 64 = 0
0.804 - 64 = 65. tgO automobile
0.118 = 66. .Hfe light duty truck, gasoline
0.046 + .8 x 64 = 67. .QJt> heavy duty vehicle, gasoline
0.062 + .2 x 64 = 68. .Ot>3> heavy duty vehicle, diesel
Average Route Speeds: peak offpeak
local streets 69. 3? 72.
arterial s 70. Q 73.
expressways 71. t+ 74.
8-17
-------
C. EXAMPLE OF CALCULATION OF MOTOR VEHICLE EMISSION FACTORS
An example of calculation of a composite motor vehicle emission
factor 1s presented in this section. An effort has been made to present
an example which represents a typical situation but requires explicit
calculation of each of the factors.
The composite emission factor to be evaluated is that for hydro-
carbons during the design year 1985,for vehicles operating in Boston,
Massachusetts. The emission factor will be-calculated for a speed of 30
miles per hour, 30 percent of the vehicles are operating from cold start,
and 40 percent are operating from hot start.* The mean ambient temperature
during this period is 50°F. The first step is to calculate the emission
factor for each vehicle type using Worksheet VEM-1 four times (i.e., one
for each vehicle type).
1. Automobiles
• Fill in lines 1 through 8 using the information stated
in the problem, placing automobiles (GA) on line 1.
• Fill in column 9 with the model year corresponding to
each vehicle age, 1985 is put next to Age=l, 1984 is
inserted next to Age=2, down to 1973 next to Age>13.
• Locate the correct table of base emission rates (Cjpn)
in Appendix D of AP-42 corresponding to the information
on lines 1, 4 and 5. The correct table is D1.17. Fill
in column 11 with the appropriate hydrocarbon base
emission rate (in grams/mile).
• Put in column 12 the proportion of VMT traveled by each
vehicle model year in 1985. For this example, the
national mix presented in Table D1.22 has been used.
• Calculate the speed correction factor to represent 30 mph
using Table D1.23. Examination of the model years (column
9) indicates that all vehicles being considered were pro-
duced after 1970. Thus, a single speed correction factor
is appropriate. This is:
*That is, 30 percent of the currently running vehicles have just started
cold, 40 percent hot, and 30 percent have not recently started.
8-18
-------
V = exp [0.942-0.0592(30) + 0.000567(30)2]
= 0.72
This factor 1s Inserted 1n column 13 for all model years.
The effects of operating the vehicle at an ambient
temperature of 50°F instead of the FTP range of 68°-86°
is calculated using the equation in Table D1.25. It is
being assumed that all post-1974 automobiles are to be
equipped with catalytic converters. Thus, separate
correction factors must be calculated for 1973-1974 and
for 1975-1985. These factors are:
Zpre-1975 = -0-0113(5°) + !-81
= 1.24
Zpost-1974 = -0-0304(50) + 3.25
= 1.73
These values are inserted with the proper model years in
column 14.
Calculation of the operating temperature correction
factor involves use of equations Dl-2 and Dl-3 plus Table
D1.25. Again, 1t 1s necessary to segregate by catalyst
and noncatalyst cars.
Noncatalyst cars (pre-1975)
- _ 30 + 70
20 + 80 f(tj
f(t) = 0.0079 (50) + 0.03
= 0.425
thus,
rpre-1975 = 1J1
Catilyst cars (post-1974)
U • 30 + 40 f(t) + 30 g(t
r " 20 + 27 f(t) + 53 g(t
f(t)catalyst = 0.0050 (50) - 0.0409
= 0.2091
G(t)cata]yst = 0.0018 (50) + 0.0095
= 0.0995
'catalyst =1'34
8-19
-------
These values are Inserted with the appropriate model years
1n column 15.
• For each model year, multiply columns 11, 12, 13, 14, and
15 together and put the result 1n column 16.
• Fill in crankcase hydrocarbon emissions (column 17) from
Table D1.26. As all model years being considered are post-
1967, this factor is 0.0 for all model years.
• Using Table D1.27, evaporative hydrocarbon emissions can be
determined for each model year. These are:
Model Year Etirission (g/mi)
Post-1979 0.5
1973-1979 1.76
These values are inserted in column 18.
• Column 12 is copied to column 19.
• Columns 17 and 18 are added and the result is multiplied by
column 19. The result is placed in column 20. For model
year 1985 this is:
(col 20) = (0.0 + 0.5) (0.112)
= 0.056
• Columns 16 and 20 are added and the result put in column 21.
For model year 1975:
(col 21) = 0.05 + 0.056
= 0.106
• The contributions of each model year to the total emissions
(column 21) is summed and the result, 2.1:5 gr/mi, is placed
on line 22. This completes calculation of the automobile
emission factor.
2. Light-Duty Trucks
• Using a new VEM-1, put light-duty trucks (LOT) on line 1.
• Fill in lines 2 through 8 and column 9 identically to the
automobile worksheet
• Use Table D2.9 to fill in column 11 and Table D2.ll to fill
1n column 12.
• Speed, ambient temperature and operating temperature correc-
tion factors for LOT are identical to those for automobiles.
8-20
-------
Complete columns 13, 14, and 15 similarily to the res-
pective columns on the automobile worksheet.
• Compute column 16 by multiplying columns 11 through 15
together for each model year.
• Fill in crankcase and evaporative emissions (columns 17
and 18) by model year using Table D2.15.
. Copy column 12 to column 19; add columns 17 and 18 and
multiply the sum by column 19. Put the result in column
20.
• Add columns 16 and 20 for each model year; insert the result
in column 21. Add column 21 and put the sum on line 22.
This is the LOT hydrocarbon emission factor.
3. Heavy-Duty Gasoline Vehicles (HDG)
• On a new VEM-1, write HDV on line 1 and again copy lines 2
through 8 and column 9 from the automobile worksheet.
• Fill in the base emission factors (column 11) from Table
D4.10. Fill in the model year mix (column 12) from Table
D4.ll.
• Determine the speed correction factor, v- s, using Table
D4.12. This is determined as follows:
v = exp (1.07 - 0.0663 (30) + 0.000598 (30)2)
v = 0.68
This value should be placed in column 13.
• The ambient and operating temperature, correction factors for
HDG is 1.0. This value is placed in columns 14 and 15 for
each model year.
. Multiply columns 11 through 15; insert the result in column
16 for each model year.
. Fill in columns 17 and 18 using Table D4.14. Note that a
50 percent reduction in evaporative emissions is assumed
after 1978.
• Copy column 12 to 19. Calculate the HDG hydrocarbon
emission rate (line 22) similarly to the LOT calculation.
4. Heavy-Duty Diesel (HDD)
• Put HDD on line 1, compute lines 2 through 8 and column 9
as was done on the other worksheets.
8-21
-------
• Fill 1n the hydrocarbon base emission rate using,Table
Table D5.1. This number 1s constant for all model years.
• Use Table D5.2 to complete column 12.
Use equation D5.3 and Table D5.3 to compute v., .
„ _ 18 [(60-30) (1.38) + (30-18H2.25)]
V 42(30) 1.38
= 0.71
This value should be Inserted in column 13.
• Columns 14 and 15 are 1.0 for HDD.
• Calculate column 16 for each model year by multiplying
columns 11 through 15.
• Columns 17 and 18 are 0.0 for HDD.
• Line 22 thus becomes the summation of column 16.
5. Calculation of Composite Emission Rate
• To calculate the composite emission factor, worksheet VEM-2
should be used.
• Fill in lines 2 through 8 as it was done on VEM-1
• In column 23, write the final emission rate (line 22 of
VEM-1) for each vehicle type.
• Insert in column 24, the mix of vehicle types by class.
In this example, a nationwide mix presented in Chapter
D7.1 of AP-42 has been used.
• Multiply columns 23 by 24 and insert the product in column
25 for each vehicle class.
• Line 26 is the sum of column 25. This is the final com-
posite hydrocarbon emission rate.
8-22
-------
EXHIBIT 8-15
WORKSHEET VEM-1
00
ro
co
Vehicle Type = 1. OTT (6A, LDT, HD6, HDD)
Pol 1 utant
Speed = 3.
Design Yea
Region = 5
-- 2. ftC- (CO, NOX. HC)
&U nph
r = 4. J7 g9
•
./13
9
Model
Year
I1ZS
mi
tft3
Htz.
/ff/
/?&>
mi
117*
1117
I17t
1175
11 W
117$
Base
Emission
Rate
ipn
4 2. 7
0.72
0,3%
JF\ £/ y
f^ £f&
O,S*f
0,$1
0,££
2,(>0
9,w
3.Ot>
6,20
6,20
12
Fraction
of
Annual
Travel
m1n
0.IIZ,
AW3
0,13®
Oi /2/
0,10$
o.ow
0,011
O.W3
0,0*7
0,032
0,011
0,013
0,031
8. GO
°F
13
Speed
Correction
Factor
Vips
0.73
0,72
0.72.
O.72
O.72.
0,7Z
0.72
0,72
0,73.
O,72
O.72
O.72.
6,72
14
Ambi ent
Temperature
Correction
Factor
A 73
1,73
1,73
1,73
1,73
1.73
1.73
173
1.73
A 73
1,73
I, -2,4
/
1,71,
1,7k
/>7t>
}.71*
l>7(*
19
Fraction
of
Annual
Travel
min
0,112.
0./H3
0,130
0,\2J
fii/0%
0,oif
0,071
O.Ote
O.W7
0,0*2
0.0tf
0,0 IZ
0.031
20
Model
Year
Hydro-
carbon
Emissions
eHd
o.et
0,07
0.0-7
0,0 (c
0,05
0.05
O.li
0.11
0,0*
O.oL
0,o3
0,02.
e,t>7
21
Model
Year
Total
Emissions
0,/t
0,15
Otl5
0*1 &
O.rf
^J3
0-22
(>.]<*
0.2%
0>2\
0./3L
0,10
0.3\
Note: When calculating carbon monoxide and nitrogen oxide emissions,
0.
Average emission factor 22:
-------
EXHibIT 8-lb
WORKSHEET VEM-1
Vehicle Type =• 1 . J-0T (6A
Pol 1 utant
Speed = 3.
, LOT, HOG, HDD)
= 2. H& (CO, NOX, HC)
30 mph
Design Year = 4. 13% S
Region = 5
. Led A-l"h (Low Alt., High Alt., Calif)
Cold Starts = 6. 3O %, Hot Starts = 7. */D %
Ambient Temperature =
CO
ro
10
Vehicle
Age
(years)
1
2
3
4
5
6
7
3' • '
9
10
11
12
>13
9
Model
Year
fits
/?*y
m3
ii*z
mi
11*0
197s?
/77s
(177
ff?t>
/?7&
l?7i
&13
11
Base
Emission
Rate
Cipn
/.o
1,2
h¥
•l.t
1,%
3,t>
3,2
2*7*
C/
^y
5", 7
7.^
7;l
12
Fraction
of
Annual
Travel
"in
O.otf
0,/V/
Ot\32
0,11-3
0,0
0,OS7
O.wy
0.033
0,023
Q.oKe
0.6*1
8. S"O
°F
13
Speed
Correction
Factor
vips
0.73
0,71
0,73.
0,7 Z
0,71
0,71
0,71
0.71
a 72
0.71
0,72
0, 71
0.73
14
Ambient
Temperature
Correction
Factor
Z1pt
A 73
1,73
1,73
1,73
l<73
/,73
J.73
J.73
1,73
1,73
A 73
/.zy
/, If
15
Operating
Temperature
Correction
Factor
riptws
/• 3$
i*3i
1.34
)>3¥
1,3?
l,3i
/<&
/.sy
i-ty
J-sy
1*31
i.i/
jji
16
Model year
Emission
Contribution
eipntwx
&.II.
0,3*
0.3)
0.33
o.w
0,3-%
0,2%
0,Z3
0f37
0,*l&~
0,t>l
17
Hydro-
carbon
Crankcase
Emissions
fi
0,V
0,0
0.0
0,0
0,0
o.v
0,0
0,0
0,0
o.o
0,0
0,0
O(D
18
Hydro-
carbon
Evaporative
Emissions
ei
£>.£
0,$
07
0,0$
0,15
21
Model
Year
Total
Emissions
0.30
0,3S
0,3*
0,31
0,34
0,31
o.zi
o.tl
0.GI
0,3^
0*1%
0,17
0.
-------
EXHIBIT 8-17
WORKSHEET VEM-1
Vehicle Type = ]_.
Pollutant = 2.
Speed = 3.
Design Year = 4.
Region = 5. l*O»>
_ (GA, LOT. HDG. HDD)
(CO, NOX. HC)
"30 mph
Cold Starts = 6. "SO
Ambient Temperature = 8.
(Low AH., High AH., Calif)
%, Hot Starts = 7.
°F
10
Vehicle
Age
(years)
9
Model
Year
11
Base
Emission
Rate
C1pn
12
Fraction
of
Annual
Travel
min
13
Speed
Correction
Factor
vips
14
Ambi ent
Temperature
Correction
Factor
Zipt
15
Operating
Temperature
Correction
Factor
riptws
16
Model year
Emission
Contribution
ipntwx
17
Hydro-
carbon
Crankcase
Emissions
fi
18
Hydro-
carbon
Evaporative
Emissions
ei
19
Fracti on
of
Annual
Travel
min
20
Model
Year
Hydro-
carbon
Emissions
CHC1
21
Model
Year
Total
Emissions
ff.6%
0,017
CO
ro
01
mi
J.O
0.0
0,
0.3L
AO
o.in
0.13
/•v
0.0
o.no
0.32
om
6,6%
0,0
0.013
427
0.0*0
/D
0,2%
o.ou
117*
6,3
J.O
o.v
6,33
0.'$')
1177
LL3.
0,0*-)
0,6$
/,*>
/.v
0,0
0,37
10
11,0
J.V
/,#
0.0
0,23
11
1173
/v
0.5D
0,0
0.03J
12
/.o
0,20
0,0
O.Olj
0,32
1173
/,&
t.v
3.31
Note: When calculating carbon monoxide and nitrogen oxide emissions, e^ = 0.
Average emission factor#2
-------
EXHIBIT 8-18
WORKSHEET VEM-1
Vehicle Type = 1. MQ (6A, LOT. HD6, HDD)
Pollutant = 2. H& (CO, NOX, HC)
Speed = 3. 3 t? mph
Design Year = 4_._
Region = 5^_
Cold Starts = 6.
Ambient Temperature = 8.
(Low Alt., High Alt., Calif)
3D %, Hot Starts = 7.
"F
10
Vehicle
Age
(years)
1
2
3
4
5
6
7
3
9
10
11
12
>13
9
Model
Year
Mr*
m*
m*
m*
mi
/w>
im
HI*
1171
MIL
in*
I17{
1173
II
Base
Emission
Rate
Cipn
«*
•41
U
**>
f4
4L
til
$6
tit,
U
¥d
Vt
V<1
12
Fraction
of
Annual
Travel
min
0,491
0>/t>1
*Jit
a/ 1 if
AIM
aozo
aon
aotz
ff. .039
0,01%
pfo/l
0,067
0.oy)
13
Speed
Correction
Factor
V
A7/
0m
0n\
0,11
on)
onl
oni
oni
CM
0,71
0.11
0,7)
0.71
14
Ambient
Temperature
Correction
Factor
Z1Pt
/*
/^
/.o
/.*>
/,D
S,2>
/.o
At)
/.O
/.O
/,t>
to
/,V
15
Operating
Temperature
Correction
Factor
riptws
/^
/.O
/.o
J.t>
J.v
J,0
40
J.V
/.z>
J,z>
/.O
/.O
/,t>
16
Model year
Emission
Contribution
ipntwx
A3J
0,4*
0££
0,53
0.3t>
£>.2t>
0,22
fi.tt
0.JJ
AM
0.0?
0.02
0.05
17
Hydro-
carbon
Crankcase
Emissions
fi
0.0
0.0
0.0
0.0
0.0
O.D
O,Q
O,0
O.O
O.D
0,V
o.o
oto
18
Hydro-
carbon
Evaporative
Emissions
ei
0,0
0.V
0.O
0.0
0,0
o.o
0.V
0.0
0,D
oto
0,0
o.o
0.0
19
Fraction
of
Annual
Travel
min
20
Model
Year
Hydro-
carbon
Emi ssi ons
CHC1
0,O
Ao
0.0
0.0
0.0
O.e
ao
0.0
0.O
00
o.t>
0.0
O.D
21
Model
Year
Total
Emissions
CTi
0,31
0,£$
0,5$
0,53
0(34>
o.u
0.2*
o,lt>
o.tl
6,06>
0.t>¥
0.03.
0.07
c»
IN3
Note: When calculating carbon monoxide and nitrogen oxide emissions,
0.
Average emission factor 22
-------
EXHIBIT 8-19
WORKSHEET VEM-2
CALCULATION OF THE COMPOSITE EMISSION FACTOR
Pollutant = 2
:Ht
Speed = 3. 30 nph
Design Year = 4. tftf&
Region = 5. h&aft- (low alt., high alt., Calif.)
Cold Starts = 6. 3& %. Hot starts - 7.
Ambient Temperature = 8.
Vehicle Class
Automobiles
Light-duty Trucks
Heavy-duty Gasoline Vehicles
Heavy-duty Diesel Vehicles
23
Emission Rate
24
Vehicle Class
Weighting
0,\\
0, 0 3 Si
• 25
Product
A73
to
O./O
Composite emission factor = 26
8-27
-------
D. CALCULATION OF EMISSIONS
This section continues with the Northglenn, Colorado, case as an
example. Only the emissions of nitrogen oxides are computed, calculations
of the emissions of the other criteria pollutants would be analogous. Assur
all stationary source energy demands are met by gas or electricity.
1. Worksheet No. EMI-1
The worksheet is filled out twice, once for gas combustion and
once for electricity consumption.
• The amount of land use is copies from worksheet LUM-3,
• The process emission factors are copied from the tables
in Chapter 4,
• The spaceheating emission factor is copied from the same
tables. Before entry, it is multiplied by the number of
degree days in Northglenn as estimated from Figure 4-1.
(e.g., in the case of residential single family detached,
this would be
2'6 * 10"3 g£Tft,9 unU-ht.d.d. * 675° too™ days
= pounds/dwelling unit),
• The space cooling emission factor is computed in an
analogous manner,
• The industrial floor area is assumed to be entirely r *•-;
composed of SIC36.
2. Worksheet No. EMI-2
The motor vehicle traffic values are copied from worksheet
VMT-3 and VMT-4.
For the purposes:of this example, a set of emission factors
were computed from AP-42 [8] using the 1972 national mix NO emission rate
y\
for low altitude and 19.6 mph. A speed correction was applied using the
light duty speed correction equations and extrapolation to high altitude
was made using a ratio derived from Table 3.1.2r3 [8].
8-28
-------
3. Worksheet No. EMI-3
Emissions on worksheets EMI-1 and EMI-2 are copied On worksheet
EMI-3. Emissions from electrical consumption, line 26, is computed by
multiplying line 26 by the appropriate emission factor from Table 4-1.
8-29
-------
00
I
EXHIBIT 8-20
WORKSHEET EMI-1
Pollutant
LAND USE CATEGORY
residential single family attached
residential single family detached
residential mobile homes
residential multifamily low rise...
residential multifamily high rise..
commercial <50K..
commerci al 50- 100K
commercial >100K
office
wholesale-warehousing
hotel s , motel s
hospi tal s .•
cul tural fadl 1 11 es
churches
educati onal fad 1 1 11 es
recreational facilities
TOTAL EMISSIONS
Heating
1
amount
7383
V
i
36
o
2<18
(o\
Kf
?&3
tel
/A
111
n
/yo
13 1>
'$3
degree da
2
process
emission
factor
o
O
O
O
0
O
0
n
0
o
O
o
0
6>
ys
3
process
emissions
i. 0
Cool ing
4
space
heating
emission
factor
77.55"
/7.S5*
//«ye
& , IG
(on?
-30
•>-*£>
,30
o
*/>
. >I6
. 10
, 10
\o
..-^/t>
degree days
5
space
heating
emissions
MS*)/
70
Vb
^^
0
&W>
/8V>
tL'S°
0
38/6
IBO
t/io
no
Wot
<>
Ib&too
6
space '
cooling
emission
factor
O
O
O
O
0
O
o
o
o
o
o
o
0
o
/fe
Operati ng
Hours
7
space
cooling
emissions
in 0
space-
heating
space-
cooling
manufacturing
land use
x 18.
19.
-------
EXHIBIT 8-21
WORKSHEET EMI-1
00
CO
LAND USE CATEGORY
residential single family attached
residential single family detached
residential mobile homes
residential multifamlly low rise...
residential multi family high rise..
commercial <50K
commerci al 50-100K
commercial ->100K
office
wholesale-warehousing
hotel s , motel s
hospi tal s
cultural facilities
churches
educati onal faci 1 1 11 es
recreati onal f aci 1 1 11 es
TOTAL EMISSIONS
1
amount
-7503
V
V
3o
o
2?8
6/
yo^
263
/&
n
HO
?fo
/5"3
2
process
emission
factor
^,^^/ofc
k /,Yx/o*
> ^.Sx/^
> /iO PfO'
2-,^ fto
27 f ii\(>
•I, f/0
H,1*io*
a,ox/ofc
S.^y/o^
4Av(ov
j. 2.$ X 10*
process
4
space
heating
emission
factor
o
o
c?
O
O
, O
6
6
izj&zgl:
,0
o
o
0
-&
o
—
kwji
5
space
leatlng
emissions
Y.6^/o6
9>>/o*
space-
heating
6
space
cooling
emission
factor
/33G
/eso
l^kO
GOO
4oo
5 i'2>o
%i\Z*-O
3>2Z>
.foo
*^i '2P
^^y<9
?. 2Q6
, 2v©0
—
M
7
space
cooling
emissions
/. Y f /o **
7,5-Jf/ft"*
5.YK/05
/,-8MOy-
<9
f,?K/0<"
/,
-------
EXHIBIT 8-22
WORKSHEET EMI-2
MOTOR VEHICLE EMISSIONS
Pol'lutant
VMr\ off peak, local streets
VMTA, off peak, arterials
VMT , off peak, expressways
VMr\ peak, local streets
VMT\ peak, arterials
VMT\ peak, expressways
VMT , off peak, local streets
VMT1, off peak, arterials
VMT , off peak, expressways
VMT , peak, local streets
VMT , peak, arterials
VMT , peak, expressways
SPEED
60
Z8_
/a
37
12
VMT
14
EMISSION
FACTOR
/02S3
15
EMISSIONS
/,/
x /o
AREA OF INFLUENCE TOTAL 16 /,d */0
/6
3,1
37
8.1
Zo
3,3
/.*>><
IMPACT AREA TOTAL 17
f /0
8-32
-------
8 process emissions
9 space heating emissions
20
PM
EXHIBIT 8-23
WORKSHEET EMI-3
EMISSIONS SUMMARY
21
SOX
22
CO
23
HC
24
NOX
25
Electricity
2,3 X/0
6
10 space cooling emissions
19 Industrial emissions
00
I
co
co
27 STATIONARY SOURCE TOTAL
AREA OF INFLUENCE
area of Influence
15 motor vehicle emissions
- *> *
8
28 TOTAL, AREA OF INFLUENCE
26 x emission factor «
29 electric, utility emissions.
16 motor vehicle, Impact
area emissions
30 TOTAL*
* Area of Influence stationary sources, secondary sources (I.e., electrical generation), and motor
vehicle emissions 1n Impact area.
-------
IX. REFERENCES
1• 40 Federal Register
40 Federal Register
40 Federal Register
40 Federal Register
40 Federal Register
40 Federal Register
40 Federal Register
40 Federal Register
40048 (October 20, 1975),
41941 (September 9, 1975),
25814 (June 19, 1975),
23746 (June 2, 1975),
18726 (April 20, 1975),
16343 (May 8, 1974),
9599 (April 18, 1973), and
6279 (March 8, 1973),
2. U.S. Environmental Protection Agency, Review of Federal Actions
Impacting the Environment, Washington, D.C., EPA, 1975 (Manual TN2/3-1-75),
Office of Federal Activities, U.S. Environmental Protection Agency,
Guidelines for Review of Environmental Impact Statements; Volume I
Highway Projects. Washington D.C..EPA, 1973, (Volume II on Airports
Sf
)li
and Volume III on Steam Channelization will be published shortly)
39 Federal Register 16186, (May 7, 1974).
Office of Air Quality Planning and Standards, U.S. Environmental
Protection Agency, Guidelines for Preparing Environmental Impact State-
ments. Research Triangle Park, N.C., OAQPS, May 1975.
3. 40 Federal Register 28064 (July 3, 1975),
39 Federal Register 45014 (December 30, 1974),
39 Federal Register 25292 (July 9, 1974),
39 Federal Register 7270 (February 24, 1974), and
38 Federal Register 15834 (June 18, 1973).
4.
40
39
39
38
37
Federal
Federal
Federal
Federal
Federal
Register
Reg i ster
Register
Register
Register
25504 (June 12, 1975),
42510 (December 5, 1974),
31000 (August 27, 1974),
18986 (July 16, 1973), and
23836 (November 9, 1972).
5. Benesh, Frank, Guldberg, Peter, and D'Agostino, Ralph, Growth Effects
of Major Land Use Projects: Volume I Specification and Causal Analysis
of Model.Prepared by Walden Research, Division of Abcor for the U.S.
Environmental Protection Agency, Office of Air Quality Planning and
Standards, Research Triangle Park, North Carolina, May 1976. (EPA-450/3
76-012a)
9-1
-------
6. Benesh, Frank and Peter Guldberg, Growth Effects of Major Land Use
Projects. Volume I: Appendices C and D. Prepared by Wai den Research,
Division of Abcor for the U.S. Environmental Protection Agency, Office
of Air Quality Planning and Standards, Research Triangle Park, North
Carolina, May 1976. Available from Air Pollution Technical Information
Center (MD#18), U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina 27711 (APTIC Document #80998), from the Land Use
Planning Office, SASD, OAQPS, U.S. Environmental Protection Agency,
MD-12, Research Triangle Park, North Carolina 27711.
7. Benesh, Frank, Growth Effects of Major Land Use Projects, Volume II:
Compilation of Land Use Based Emission Factors, Prepared by Wai den
Research, Division of Abcor for the U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards, Research Triangle
Park, North Carolina, June 1976, (EPA-450/3-76-012-b).
8. Office of Air Quality Planning and Standards, U.S. Environmental Protec-
tion Agency, Compilation of Air Pollutant Emission Factors, 2nd Edition,
Research Triangle Park, N.C., April 1973, and supplements (AP-42).
9. See, for example, the Keystone Coal Industry Manual. Published by the
Mining Information Services of McGraw Hi 11, New York, N.Y., 1969.
10. Couillard, James, Browns Directory of North American Gas Companies,
Harcourt Brace Jovanovich, Duluth, Minnesota, 1973. '.
11. Environmental Data Service, National Oceanic and Atmospheric Admini-
stration, Heating and Cooling Degree Day Data, Environmental Informa-
tion Summaries C-14.Ashevilie, North Carolina, September 1974.
12. National Coal Association, Steam Electric Plant Factors, Washington,
D.C., 1973.
13. Edison Electric Institute, Statistical Yearbook of the Electric Utility
Industry, New York, N.Y., 1975^
14. Office of Management and Budget, Standard Industrial Classification Code
Manual, Washington D.C., 1972.
15. Levinson, H.S., Transportation and Traffic Engineering Handbook, Insti-
tute of Traffic Engineers, 1976.
16. "Traffic Generation Rates", Traffic Engineering, February 1973.
-------
17. Transportation Planning Office, MaHcopa Association of Governments,
Trip Generation by Land Use, Phoenix, Arizona.
18. National Capitol Region Transportation Planning Board, Metropolitan
Washington Council of Government, Information Report No. 60. 1973.
19. National Research Council, National ^ Cooperative Research Program
Report No. 48: Factors and Trends in Trip Lengths, Washington D.C.,
~
20. Cerighton, Rodger et.al., "Estimating Efficient Spacing for Arterial s
and Expressways", Highway Research Board Bulletin No. 253. Washington,
D.C., 1960.
21. National Research Council, Highway Capacity Manual, 1965, Highway
Research Board Special Report Number 87, Washington, D.C., 1965.
22. Office of Air Quality Planning and Standards, U.S. Environmental Pro-
tection Agency, Guidelines for Air Quality Maintenance Planning and
Analysis Volume 9: Evaluating Indirect Sources. Research Triangle
Park, North Carolina, January 1975. (EPA-450/ 4- 75-001 )
23. Bureau of the Census, 1970 Census of Housing, Washington, D.C.
24. Bureau of the Census, 1972 Census of Manufacturers. Washington, D.C.
-------
APPENDIX A
STATISTICAL OUTPUT FOR THE PREDICTIVE LAND USE EQUATIONS
A-l
-------
EQUATION 1.
ORDINARY LEAST SQUARES
1. RESIDENTIAL MODEL
DEPENDENT VARU&LE:
RES
RIGHT-HAND
VARIABLE
OUACRE
DELP2
OISCBD
HUYINT
SEBEfc
MPR70
. C
ESTIMATED
COEFFICIENT
8905.83
6785.85
-350.874
-1358.69
41.1915
-.681883
7204.98
STANDARD
ERROR
2602.95
2576.30
141 .187
766.921
25.4075
.472902
4174.30
T-
STATISTIC
3.42143
2.63395
-2.48518
-1.77162
1.62123
-1.44191
1 .72603
R-SOUARED = .7217
LOG OF LIKELIHOOD FUNCTION = -189.311
DURBIN-UATSON STATISTIC (ADJ. FOR 0. GAPS) = 2.3634
NUMBER OF OBSERVATIONS =• 20.
SUM OF SQUARED RESIDUALS = .195094+09
STANDARD ERROR OF THE REGRESSION = 3873.91
SUM OF RESIDUALS = .146484-02
MEAN VALUE OF DEPENDENT VARIABLE = 4784.9C
F-STATISTIC< 6»t 13.) = 5.61937
-------
1. RESIDENTIAL MODEL
ESTIM»TE OF VAfclANCE-COVARIANCE MATRIX OF ESTIMATED COEFFICIENTS
OUACRE OELP2 DISCBO HUYINT SEMES
DUACF-E
DELP2
DISCED
HUYINT
SEWER
MPR70
C
HPR70
I*.*.....*.......................*...*..
.677537+07
-353084.
13017.8
-.118030+07
14836.0
17.7991
. -.196117 + 07
-353084.
.663732+07
-69404.6
110192.
16535.0
475.734
-.389661+07
18017.8
-69404.6
19933.7
32323.0
504.143
-5.33587
-407154.
-.118080+07
110192.
32323.0
583167.
-809.247
-8.98286
-863813.
14836.0
16S35.0
504.143
-809.247
645.541
-.951074
-55546.9
17.7991
475.734
-5.33587
-8.98286
-.951074
.223637
-752.193
OUA-CftE
DELP2
OISC6D
HUYINT
SEUEF.
MPR70
C
-.196117 + 07
-.339661+07
-407154.
-863813.
-55546.9
-752.193
.174248+08
LINE
4.
SMf'L
SMPL =
LINE 5.
OLSO
i
4.
6. 9. 13. 13. 18i
-------
1. RESIDENTIAL MODEL
2.
•*»«»«•*»*»«
LtAST SQUARES
DEPENDENT VARIABLE: COMM
KIGHT-HMND
VARIABLE
OFFACK
UISCBO
DUACHt
HrfYlNT
EMP60
VACACH
C
ESTIMATED
COEFFICIENT
790.700
-73.6968
655.610
-200.069
.327286-02
.647417-01
1376.49
STANDARD
ERROR
472.200
24.9239
347.530
114.997
.845695-03
.453333-01
731.164
T-
STAT1STIC
I.b74b0
-2.95687
1.88648
-1.73977
3.87003
1.42813
1.88260
R-SQUARt(> = .8228
LOfa OF UlKtLlHOOU FUNCTION = -149.006
DURBIN-WAT50N STATISTIC (ADJ. FOP 0. GAPS) = 2.3607
NUMBER OF OBSERVATIONS = 20.
SUM OF SUUARED RESIDUALS = .346576*07
STANDARD ERROR OF THE REGRESSION = 516.330
SUM OF RESIDUALS = .188470-03
MEAN VALUE OF DEPENDENT VARIABLE = 1044.60
F-STATISTIC< 6.* 13.) = 10.0636
-------
1. RESIDENTIAL MODEL
ESTIMATE OF VARIANCE-COVARIANCE MATRIX OF ESTIMATED COEFFICIENTS
OFFACR DISC8D DUACRE H*Y1NT
EMPbO
VACACR
OFFAC*
DISCdtl
DUACRt
HWYINT
EMP60
VACACR
C
222973.
7338.67
-20616.9
23835.7
-.6*8491-01
1.91226
-194816.
1
7338.67
621.200
-854.726
1366.15
-.288654-02
.370956
-16068.2
-20616.9
-854.726
120777.
-23758.1
-.612073-01
-1.99969
31067.8
23835.7
1366.15
-23758.1
13224.4
.815642-02
.240051
-39794.8
,648491-01
.2B8654-0*
.612073-01
,815b42-02
.715200-06
.744408-05
.952083-01
1.91226
.370956
•1.99969
.^40051
.744408-05
.205511-02
-21.2551
CJI
OFFACH
DJSCHD
HWYINT
EMP60
VACACK
C
-194818.
-16068.2
31067.8
-39794.8
-.952083-01
-21.2551
534601.
LINE 5.
OLSQ
-------
1. RESIDENTIAL MODEL
tiiUATION 3.
ORDINARY LEAST SQUARES
DEPENDENT VARIABLE: OFFICE
RiGMT-HANU
VARIABLE
OFFACR
DELEMP
OISCBO
OUACRt
HfcYINT
C
ESTIMATED
COEFFICIENT
84*. 615
601.451
-14.0890
-400.102
85.8478
355.212
STANDAKL)
ERROR
186.772
217.530
9.76455
135.701
45.71bS»
225.771
T-
STATISTIC
4.52218
2.76491
-1.44207
-2.94841
i. 87773
i. 57333
H-SQUARti) = .8061
L013 OF LIKELIHOOD FUNCTION = -131.346
DUR6IN-*ATSON STATISTIC (ADJ. FOR 0. bAPS) = 2.8054
NUMBER OF OBSERVATIONS = 20.
SUM OF SQUARED RESIDUALS = 592677.
STANDARD ERROR OF THE REGRESSION = 205.752
SUM OF RESIDUALS = .2*5639-04
MtAN VALUE 0^ DEPENDENT VARIABLE = 305.900
F-STATISTICC 5.. 14.) = 11.6440
ESTIMATE OF VARIANCE-COVARIANCE MATRIX OF ESTIMATED COEFFICIENTS
OFFACR DELEMP DISCBD OUACRE HWYINT
OFFACR
DELEMP
DISCBD
DUACRt
HWYINT
C
34883.7
-6697.79
1144.73
-3418.58
3951.46
-29102.2
-6697.79
47319.3
-736.557
-3203.73
-574.215
7101.15
1144.73
-736.557
95.3465
-81.2103
226.107
-2070.62
-3418.58
-3203.73
-81.2103
18414.8
-3605.95
941.616
3951.46
-574.215
226.107
-3605.95
3090.22
-5980.48
-29102.2
7101.15
-2070.62
941.616
-5980.48
50972.5
-------
1. RESIDENTIAL MODEL
EQUATION 4.
ORDINARY LEAST SUUARES
DEPENDENT VARIABLE:
ESTIMATED STANDARD T-
VAR1ASLE COEFFICIENT ERROR STATISTIC
MANACR 1050.74 597.153 1.75958
C 761.385 334.760 2.27442
R'SQUAKED = .1468
LUG OF LIKELIHOOD FUNCTION = -172.521
DURBIN-WATSON STATISTIC (ADJ. FOR 0. toAPS) = 1.9656
NUMBEK OF OBSERVATIONS =20.
SUM OF SQUARED RESIDUALS = .363972*08
STANDARD ERROR OF THE REGRESSION = 1421.99
SUM OF RESIDUALS = .213623-03
MEAN VALUE OF DEPENDENT VARIABLE = 94b.600
F-STATISTIC( l.» 16.) = 3.09611
ESTIMATE OF VARIANCE-COVARIANCE MATRIX OF ESTIMATED COEFFICIENTS
MANACR C
MANACH . 356592. -62517.7
C . -62517.7 112064.
-------
1. RESIDENTIAL MODEL
EQUATION 5.
«*••*»•*»»*•
ORDINARY LEAST SUUARES
DEPENDENT VARIABLE: H*LMNX
00
RIGHT-HAND
VARIABLE
DISC8D
UELP2
AUT02
MINC
MPK68
C
ESTIMATED
COEFFICIENT
-2.78590
40.4556
-135.053
46.5689
.595194-02
78.7947
STANDnnu
ERROR
.679119
12.3415
63.4479
19.3215
.3497*2-02
48.0493
T-
STAT1STIC
-4.10223
3.27U01
-2.12056
2.41020
1.70157
1.63987
R-SQUAREL) = .6883
LOG OF LIKELIHOOD FUNCTION = -81.3251
DURBIN-rfATSON STATISTIC (ADJ. FOB 0. GAPS) = 1.5076
NUMBER OK OBSERVATIONS = 20.
SUM OF SQUARED RESIDUALS = 3985.28
STANDARD ERROR OF THE REGRESSION = 16.B72I
SUM OF RESIDUALS = .941753-05
MEAN VALUE OF DEPENDENT VARIABLE = 22.9351
F-STATISTIC( 5.» 14.) = 6
ESTIMATE OF VARIANCE-COVARIANCE MATRIX OF ESTIMATED COEFFICIENTS
DISCtfD DELP2 AUT02 MINC
MPR68
DISCbL)
DELP2
AUT02
MINC
MPR68
c
> .461202
-1.22643
19.7129
2.65591
. -.390880-04
-22.3840
-1.22643
152.313
-119.401
102.544
.257138-01
-129.792
19.7129
-119.401
4025.64
-304.294
-.3731*4-01
-2^93.13
2.65591
102.544
-304.294
373.322
.335064-01
-362.150
-.390880-04
.257138-01
-.373194-01
.335064-01
.122354-04
-.494805-01
-22.3840
-129.792
-2293.13
-362.150
-.494605-01
2308.74
-------
1. RESIDENTIAL MODEL
EQUATION 6.
««*»«»»»«»«»
OKIJIMARY LEAST SQUARES
DEPENDENT VARIABLE: WHOLE
RIGHT-HAND
VARIABLE
HKYINT
MPR70
DUACHE
OISCBO
DFFVAC
C
ESTIMATED
COEFFICIENT
97.1018
-.735730-01
-268.809
-11.4364
14.9875
487.849
STANDARD
ERROR
23.7292
.171831-01
77.6841
4.24861
7.43460
105.358
T-
STATISTIC
4.09209
-4.28172
-3.460^3
-2.69160
2.01591
4.63039
R-SGIMREO = .7852
LOG OF LIKELIHOOD FUNCTION = -120.*66
DURBIN-WATSON STATISTIC (AOj. FOR 0. tiAPS) = 2.2186
NUMBER OF OBSERVATIONS = 20.
SUM Oh SQUARED RESIDUALS = 199b6b.
STANDARD ERROR OF Trit REGRESSION = 119.424
SUM OF RESIDUALS = .448227-04
MtAN VALUE OF DEPENDENT VARIABLE = 150.iOO
F-STATISTIC( 5.. 14.) = 10.*338
ESTlMATt OF VARIANCE-COVARIANCE MATRIX OF ESTIMATED COEFFICIENTS
HWYINT MPR70 DUACRE OISCBD OFF VAC
HWYINT
MPR70
DUACHE
DISCHU
OFFVAC
C
563.073
.109960-01
-1093.84
34.9622
-20.6242
-849.661
.109960-01
.295258-03
.777411-01
.988541-02
-.822210-01
-.651632
-1093.64
.777411-01
6034.81
-4.15917
10.7b79
-921.639
34.V622
.9ttd54l-02
-4.15917
18.0507
-5.34734
-379.993
-20.6242
-.822210-01
10.7679
-5.34734
55.2732
32.4194
-B49.661
-.651632
-921.639
-379.993
32.M94
11100.3
-------
1. RESIDENTIAL MODEL
EQUATION 7.
»»«*»»»»««»«
ORDINARY LEAST SQUARES
DEPENDENT VARIABLE: HOTEL
I
o
RIGHT-HAND
VARIABLE
ESTIMATED
COEFflCIENT
-.968159
229.693
-150.165
81.7404
STANDARD
ERROR
2RES
AUTO
INCMPL
C
R-SQUAREO = .6969
LOG OF LIKELIHOOD FUNCTION = -98.7268
DURBIN-WATSON STATISTIC (ADJ. FOR 0. GAPS) = 1.9559
NUMBEk OF OBSERVATIONS = 20.
SUM OF SQUARED RESIDUALS = 22709.5
STANDARD ERROR OF THE REGRESSION = 37.b742
SUM OF RESIDUALS = .319481-04
MEAN VALUE OF DEPENDENT VARIABLE = 69.8000
F-STATISTIC< 3., 16.) = 12.2624
.251362
52.6012
52.1284
21.5351
T-
STATISTIC
-3.85165
4.35014
-2.88007
3.79568
ZRES
AUTO
INCMPL
C
ESTIMATE OF VARIANCE-COVARIANCE MATRIX OF ESTIMATED COEFFICIENTS
ZRES AUTO INCMPL C
.631830-01
.945b72
-2.16171
-4.08181
.945672
2787.97
-1817.89
-650.777
-2.16171
-ltt!7.«9
2717.37
444.536
-4.08181
-650.777
444. 53b
463.760
-------
1. RESIDENTIAL MODEL
EUUATION rt.
«»««««»»»«««
ORDINARY LEAST SUUARES
otPENuENT VARIABLE: HOSPTL
RIGHT-MANU ESTIMATED STANDARD T-
VARlAbLli COEFFICIENT ERROR STATISTIC
OFFACR 190.692 64.53bO
MPR70 .196279-01 .102566-01 i.91369
C -23.9110 42.0090
R-SQUAREO = .3761
LOG OF LIKELIHOOD FUNCTION = -116.930
DURblN-WATSON STATISTIC (ADJ. FOR 0. GAPS) = 1.4952
NUMBER OF OBSERVATIONS = 20.
SUM OF SQUARED RESIDUALS = 140198.
STANDARD EHHOR OF THE REGRESSION = 90.8125
SUM OF KFSIDUALS = .107586-04
MEAN VALUE OF DEPENDENT VARIABLE = 68.4500
F-STAT1STIC< 2.. 17.) = 5.1«-l333
ESTIMATE OF VAR1ANCE-COVARIANCE MATRIX OF ESTIMATED COEFFICIENTS
OFFACR MPR70 C
OFFACR
MPR70
C
4164.77
. .161220
-1152.95
.161220
.105197-03
-.364389
-1152.95
-.364389
1764.76
-------
1. RESIDENTIAL MODEL
EUUATIOM s>.
»*•»««««««««
ORDINAKY LEAST SQUARES
DEPENDENT VARIABLE: CULTUR
RIGHT-HAND
VARIABLE
ESTIMATED
COEFFICIENT
bTAiMDAHO
ERROR
UNIV 60.?198
VACACrt .174577-02
C . 2.54233
R-SUUARED = .4939
LOG OF LIKELIHOOD FUNCTION = -6l.8dl9
DURBIN-WATSON STATISTIC (AOJ, FOR 0. *iAPS) = 2.6911
NUMflEk OF OBSERVATIONS =20.
SUM OF SQUARED RESIDUALS = 4213.bO
STANDAftU ERROR OF THE REGRESSION = 15.7433
SUM OF fiESIDUALS = .160933-05
MtAN VALUE OF DEPENDENT VARIABLE = 16.5500
F-STATISTICf 2.t 17.) = 6.29*14
16.1777
.1210B2-02
b.39803
T-
STAT ISTIC
3.72239
1.441U1
.302730
ESTIMATE OF VARIANCE-COVAR-IANCE MATRIX OF ESTIMATED COEFFICIENTS
UNIV VACACH c
UNIV
VACACK
C
261.719
-.109717-02
-6.17482
-.109717-02
.146608-05
-.918005-02
-6.17482
-.918005-02
70.5270
-------
1. RESIDENTIAL MODEL
EQUATION 10.
LEAST SQUARES
DEPENDENT VARIABLE: CHURCH
WIGHT-HAND ESTIMATEO STANDARD T-
VARIABLE , COEFFICIENT ERROR STATISTIC
MINC 201.827 88.6159 2.27755
UISCbD
C .
R-SQUARtO = .4QB4
LOG OF LIKELIHOOD FUNCTION = -117.272
DUKblN-WATSpN STATISTIC (AOJ. FOR 0. GAPS) = 2.0615
NUMBER OF OBSERVATIONS = 2.0.
SUM OF SUUARED RfcSlDUALS =
STANDARD ERROR OF THE REGRESSION =
SUM OF KESIDUALS = .438690-04
MEAN VALUE OF DEPENDENT VARIABLE = 100.600
F-STATISTICC 2.» 17.) = 5.86664
ESTIMATEO
COEFFICIENT
201.827
-4.42480
-14.6922
STANDARD
ERROR
88.6159
3.2574b
120.951
ESTIMATE OF VARIANCE-COVARIANCE MATRIA OF ESTIMATED COEFFICIENTS
MINC DISCED C
MINC
DISCMO
C
7852.78
121.823
-1010M.3
121.623
10.6112
-323.609
-10103.3
-323.609
le>o20.4
-------
1. RESIDENTIAL MODEL
EQUATION 11.
ORDINARY LEAST SQUARES
DEPENDENT VARIABLE: EOUC
RIGHT-HANI)
VARIAbLt
VACACn
StWtK
DISCriD
MPKIDS
C
ESTIMATED
COEFFICIENT
.407932-01
3.46382
-25.4996
184.094
2*3.617
STANDARD
EfiKOR
.304691-01
1.78201
11.3157
113.419
339. 8i8
T-
STATISTIC
1.33884
1.38261
-if. 25347
1.62313
.716904
R-SUUARtU = .6481
LOQ OF LIKELIHOOD FUNCTION = -138.227
DUHblN-WATSON STATISTIC
-------
1. RESIDENTIAL MODEL
EQUATION ll.
««»»«»«&««»«
LEAST SOUARES
I
en
ESTIMATED
COEFFICIENT
.102870
-33.5125
STANDARD
ERROR
.277401-01
92.8600
DEPENDENT VARIABLE: REC
RIGHT-HAivD
VARIABLE
MPACRE
C
R-SUUARED = .4331
LOG OF LIKELIHOOD FUNCTION = -142.276
OUR6IN-*ATSON STATISTIC (ADJ. F0» 0. OAPS) = 1.5283
NUMBER OF OBSERVATIONS = 20.
SUM OF SQUARED RESIDUALS = ,17b822»07
STANDARD ERROR OF TrtE REGRESSION = 313.424
SUM OF RESIDUALS = ,762940-Ob
MEAN VALUE OF DEPENDENT VARIABLE = 192.400
F-S>TATISTIC< l.t 18.) = 13.7518
T-
STATISTIC
J.70B35
MPACRE
C
ESTIMATE OF VARIANCE-COVARIANCE MATRIX OF ESTIMATED COEFFICIENTS
MPACRE C
.769512-03
•l.h«993
-1.68993
8622.98
-------
EQUATION 1.
*«»»»»»*»»«»
ORDINARY LEAST SQUARES
2. INDUSTRIAL/OFFICE MODEL
DEPENDENT VARIABLE:
RIGHT-HAND
RES
DUACHE
VACACrf
OFFVAC
VACHSlJ
DISCBD
C
R-SQUARtD = .8151
LOG OF LIKELIHOOD FUNCTION = -183.876
DURBIM-WATSON STATISTIC (ADJ. FOR 0. GAPS) = 2.4802
NUMBER OF OBSERVATIONS = 20.
ESTIMATED
COEFFICIENT
3555.22
1.95250
542.303
-128659.
-281.018
-74*6.75
STANDARD
ERROR
1172.41
.416064
128.944
35403.7
125.513
3587.67
T-
STATISTIC
J. 03240
4.69280
4.20494
-J. 63407
-2.23896
-ei. 07565
SUM OF SQUARED RESIDUALS = .113289+09
STANDARD TRROR OF TriE REGRESSION = 2844.66
SUM OF KESIDUALS = .241852-02
MtAN VALUE OF DEPENDENT VARIABLE = 8433.00
F-STATISTIC1 5., 14.) = 12.3472
ESTIMATE OF VARIANCE-COVARIANCE MATRIX OF ESTIMATED COEFFICIENTS
OUACRE VACACH OFFVAC VACHSG
DISCBD
DUACRE
VACACH
OFFVAC
VACHSG
DISCBD
C
.137455+07
-125.077
6511.41
.799428+07
88258.1
-.128288+07
-125.077
.173109
9.25953
-6314.56
-18.1818
-1024.03
6511.41
9.25953
16626.6
-.139222+07
-2429.66
-137716.
.799428+07
-6314.56
-.139222+07
.125342+10
.171521+07
-.755565+07
88258.1
-18.1810
-2429.56
.171521+0'
15753.4
-124939.
-.128288+07
-1024.03
-137716.
-.755565+07
-124939.
.128714+08
-------
2. IfCUSTRIAL/OFFICE MODEL
EQUATION «?.
««»*««»»«««»
OHDINARY LEAST SQUARES
DEPENDENT VARIABLE:
COMM
I
->J
RIGHT-HAND
VARIABLE
OUACRE
ZCOMM
OFFACH
MPE70
C
ESTIMATED
COEFFICIENT
675.453
129.980
-1651.08
.830785-01
-517.259
STANDARD
ERROR
158.607
26.0961
460.372
.411113-01
33S.258
T-
STATISTIC
4.25866
4.98083
-3.58641
2.02082
-1.54287
R-SQUARED = .7753
LUG OF LIKELIHOOD FUNCTION = -148.981
DURBIN-WATSON STATISTIC (ADJ. F09 0. GAPS) = 1.8428
NUMBER OF OBSERVATIONS = 20.
SUM OF SQUARED RESIDUALS = .345713*07
STANDARD ERROR OF THE REGRESSION = 480.078
SUM OF RFSIDUALS = .289440-03
MEAN VALUE OF DEPENDENT VARIABLE = 1022.15
F-STATISTICI 4.t 15.) = 12.V412
ESTIMATE OF VARIANCE-COVARIANCE MATRIX OF ESTIMATED COEFFICIENTS
DUACRE ZCOMM UFFACR MPE70 C
DUACRE
ZCOMM
OFFACK.
MPE70
C
25156.1
-651.261
2415.89
.467425
-18911.9
-651.261
681. '008
-7918.86
.136378
-3057.60
2415.89
-7918.86
211942.
-3.02865
25129.3
.467425
.136378
-3.02865
.169014-02
-11.3469
-18911.9
-3057.60
25129.3
-11.3469
112398.
-------
2. INDUSTRIAL/OFF ICE MODEL
EQUATION 3.
ORDINARY LEAST SQUARES
DEPENDENT VARIABLE: OFFICE
I
00
RIGHT-HAND
VARIABLE
RRMI
ZOFF
MPE70
MINt
MANACR
C
ESTIMATED
COEFFICIENT
13.8650
67.4673
-.316890-01
4B9.752
223.667
-3*9.544
STANDARD
ERROR
5.37706
19.5253
.116614-01
200.045
119.521
220.143
T-
SJATISTIC
£.50227
3.45537
-2.71731
-------
2. INDUSTRIAL/OFFICE MODEL
INFONtT TSP
EUUATION 4.
ORDINARY LEAST SQUARES
ESTIMATED
COEFFICIENT
10113.6
-2623.49
.252297
1121.07
STAiNUAKO
EHHOR
3346.20
1307.60
.1798B7
1737.67
DEPENDENT VARIABLE: HANF
RIGHT-HANU
VARIABLE
OELEMP
MINCC
VACACH
C
R-SQUAREO = .4741
LOG OF LIKELIHOOD FUNCTION = -169.447
DURBlN-wATSON STATISTIC (ADJ. FOP 0. GAPS) = 1.3809
NUMBER OF OBSERVATIONS = 80.
SUM Of SQUARED RESIDUALS = .367641*06
STANDARD ERROR OF THE REGRESSION = 1293.35
SUM OF RESIDUALS = .327^87-03
MEAN VALUE OF DEPENDENT VARIABLE = I54b.70
F-STATISTIC( 3.. 16.) = 4.8076b
T-
STATISTIC
645UU2
DELEMH
MINCC
VACACR
C
ESTIMATE OF VARIANCE-COVARIANCE MATRIX OF ESTIMATED COEFFICIENTS
DELEMP MINCC VACACP C
.111971*08
-998975.
. -144.174
.110844*07
-998975.
.171034*07
-56.8196
-.124458*07
-144.174
-56.6196
.323593-01
-200.569
.1 10844*07
-.124458*07
-200.589
.302016+07
-------
2. INDUSTRIAL/OFF ICE MODEL
EQUATION 5.
**«*•«««««««
ORDINARY LEAST SQUARES
DEPENDENT VARIABLE: HWLMNX
3>
ro
o
MIGHT-HAND
VARIABLE
VACACR
HWYINT
DUACRE
OFFVAC
C
ESTIMATED
COEFFICIENT
.574184-02
-5.60785
1^.2396
.915400
-22.9269
STANDARD
ERROR
.191867-02
1.71961
4.93252
.660151
18.2013
T-
STATISTIC
t. 99230
-3.26111
2.88689
1.38665
-1.25963
R-SOUARED = .6519
LOG OF LIKELIHOOD FUNCTION = -79.4970
DURBIN-WATSON STATISTIC UDj. FOR 0. GAPS) = 1.9565
NUMBER OF OBSERVATIONS = 20.
SUM OF SOUARED RESIDUALS = 3319.44
STANDARD ERROR OF THE REGRESSION .= 14.8760
SUM OF RESIDUALS = .177026-04
MEAN VALUE OF DEPENDENT VARIABLE = 31.9950
F-STATISTICC 4*. 15.) = 7.02229
ESTIMATE OF VARIANCE-COVARIANCE MATRIX OF ESTIMATED COEFFICIENTS
VACACR HWYINT OUACPE OFFVAC C
VACACH
HWYINT
DUACRE
OFFVAC
C
.368207-05
.104191-03
-.849081-03
.418295-04
-.314397-01
.104191-03
2.95708
-.986078
-.266171
-4.65433
-.849081-03
-.986078
24.3298
.592369
-14.1386
.418295-04
-.266171
.592369
.435800
-3.69930
-.314397-01
-4.65433
-14.1386
-3.69930
331.288
CO
CO
LINE 8.
OLSQ
-------
2. INDUSTRIAL/OFFICE MODEL
EQUATION 6.
«»**»«««»»««
ORDINARY LEAST SQUARES
DEPENDENT VARIABLE: WHOLE
ro
HIGHT-HANO
VARIABLE
MMEA
ZINU
SEWER
OUACRE
C
ESTIMATED
COEFFICIENT
7467.78
90.8420
11.4350
735.642
-1650.77
STANDARD
ERROR
5003.46
27.9858
5.57329
364.803
448.080
T-
STATISTIC
1.49237
J. 24600
it. 05175
1.88575
-3.68410
R-SQUAREU = .7744
LOG Of LIKELIHOOD FUNCTION = -160.047
DUR8IN-WATSON STATISTIC (ADJ. FOR 0. GAPS) = 1.9309
NUMBER OF OBSERVATIONS = 20.
SUM OF SQUARED RESIDUALS = .104548*08
STANDARD ERROR OF THE REGRESSION = 834.857
SUM OF RESIDUALS = .877380-04
MEAN VALUE OF DEPENDENT VARIABLE = 882.850
F-STATISTIC( 4., 15.) = 12.8709
ESTIMATE OF VARIANCE-COVARIANCE MATRIX OF ESTIMATED COEFFICIENTS
WMEA ZIND SEWER DUACRE
MMEA
ZIND
SEWER
DUACRE
C
.250396*08
-51180.4
470.985
. -.134099*07
457564.
-51180.4
783.208
-25.4210
2228.98
-7004.19
470.985
-25.4210.
31*0616
-301.431
-1109.66
-.U4099*07
2228.98
-301.431
146073.
-63565. v
457564.
-7004.1S*
-110*. 66
-63565.9
200776.
-------
2. INDUSTRIAL/OFHCE MODEL
fcUUATION 7.
«»•««««»«*»•
ORDINARY LEAST SQUARES
rx>
ro
ESTIMATED
COEFFICIENT
-11.5140
1182.05
-249.174
145.184
STANDAHD
EKROK
3.76318
385.723
109.656
46.6174
DEPENDENT VARIABLES MOTEL
RIGHT-HAND
VARIABLE
DISCBD
DELEMP
OFFACrt
C
R-SQUARED = .4595
LOG OF LIKELIHOOD FUNCTION = -119.487
DUR&IN-MATSON STATISTIC (ADJ. FOR 0. GAPS) = 1.4542
NUMBER OF OBSERVATIONS = 20.
SUM OF SQUARED RESIDUALS - 161054.
STANDARD ERROR OF THE REGRESSION = 106.376
SUM OF RESIDUALS = .371933-04
MEAN VALUE OF DEPENDENT VARIABLE = 102.200
F-STATISTIC( 3.» 16.) = 4.53418
T-
STAT1STIC
-J.05964
J.06451
-2.272J3
J.11437
DISCBD
DELEMP
OFFACR
C
ESTIMATE OF VARIANCE-COVARIANCE MATRIX OF ESTIMATED COEFFICIENTS
DISCBD DELEMP OFFACR C
14.1615
-555.762
99.6724
-108.405
-555.762
148782.
-30440.4
-4130.45
99.6724
-30440.4
12024.3
-148.614
-108.405
-4130.45
-148.614
2173.18
-------
2. INDUSTRIAL/OFFICE MODEL
EUUATION b.
OHOINAHY LEAST SQUARES
ro
CO
ESTIMATED
COEFFICIENT
477.573
442.980
-283.253
-20.5055
STANDARD
ERROR
228.398
159.240
127.150
53.0288
DEPENDENT VARIABLE: HOSPTL
RIGHT-HAND
VARIABLE
NONHSE
MANACH
OFFACH
C
R-SQUARED = .5140
LOG OF LIKELIHOOD FUNCTION = -127.332
DURBIN-WATSON STATISTIC (ADJ. FOR 0. GAPS) a 1.9446
NUMBER OF OBSERVATIONS = 20.
SUM OF SQUARED RESIDUALS = 3*6759.
STANDARD ERROR OF THE REGRESSION = 157.472
SUM OF RESIDUALS = .367165-04
MEAN VALUE OF DEPENDENT VARIABLE = 91.4500
F-STATISTIC( 3.» 16.) = 5.64018
T-
SfATISTlC
-------
2. INDUSTRIAL/OFFICE MODEL
EUUAT10N 9.
»»«•»*•»«»»•
ORDINARY LEAST SQUARES
DEPENDENT VARIABLE: CULTUP
RIGHT-HAND
VARIABLE
ENERGY
EMP60
MPACRE
PVTSCH
C
ESTIMATED
COEFFICIENT
-34.6341
.439357-04
.410940-01
12.5543
23.5859
STANDAKU
ERROR
16.25*7
.206780-04
.209451-01
0. 56493
18.5760
T-
STATISTIC
-*. 13010
*. 12475
1.96198
1.46578
1.26956
R-SQUAREO = .4312
LOG OF LIKELIHOOD FUNCTION = -B2.7249
DURBIN-nATSON STATISTIC (ADJ. FOR 0. GAPS) = 1.4391
NUMBER OF OBSERVATIONS =20.
SUM OF SQUARED RESIDUALS = 45H4.09
STANDARD ERROR OF THE REGRESSION = 17.4816
SUM OF RESIDUALS = .435114-05
MEAN VALUE OF DEPENDENT VARIABLE = 12.9000
F-STATISTIC( 4.. 15.) = 2.84329
ESTIMATE OF VARIANCE-COVARIANCE MATRIX OF ESTIMATED COEFFICIENTS
ENERGY EMP60 MPACRE PVTSCH ' C
ENERGY
EMP60
MPACRE
PVTSCH
C
264.215
-.105855-03
.366510-01
13.1479
-257.839
-.105855-03
.427579-09
.823688-07
.231915-05
.215793-06
.366510-01
.813668-07
.438699-03
.702592-01
-.186684
13.1479
.231915-05
.702592-01
73.3581
-62.9573
-257.839
.215793-06
-.1866*4
-62.9573
345.142
-------
2. INDUSTRIAL/OFFICE MODEL
EQUATION! 10.
ORDINARY LEAST SQUARES
DEPENDENT VARIABLE: CHURCH
j\»
tn
RIGHT-HAND
VARIABLE
HRMI
VACACR
SEMtR
MPET2
C
ESTIMATED
COEFFICIENT
6.65488
.313526-01
-1.06916
-.145485-01
-147.539
STANDARD
ERROR
3.04563
.113178-01
.574616
.805126-02
90.1540
T-
SJAT1STIC
«i.841t>5
d. 77021
-1.86065
-1.80648
-1.63652
R-SaUARED = .4764
L06 OF LIKELIHOOD FUNCTION = -110.120
DURBIN-MATSON STATISTIC (ADJ. FOR 0. GAPS) = 1.8564
NUMBtH OF OBSERVATIONS = 20.
SUM OF SQUARED RESIDUALS = 70959.6
STANDARD ERROR OF THE REGRESSION = 68.7797
SUM OF RESIDUALS = .176430-04
MEAN VALUE OF DEPENDENT VARIABLE ~ 94.4000
F-STATISTICC 4.» 15.) = 3.41175
ESTIMATE OF VARIANCE-COVARIANCE MATRIX OF ESTIMATED COEFFICIENTS
RRMI VACACR SEHER MPET2
RRMI
VACACR
SEWER
MPET2
C
. 9.27706
. .970243-02
-.566348
. -.205b48-02
-123.864
.970243-02
.128092-03
-.396683-02
-.229718-04
-.904203
-.566348
-.396683-02
.330184
.107381-02
18.3747
-.20564B-02
-.229716-04
.107381-02
.648229-04
-.295834-01
-123.864
-.904203
18.3747
-.295834-01
8127.74
-------
2. INDUSTRIAL/OFFICE MODEL
EQUATION 11.
••*«•«•»»•»•
ORDINARY LEAST SQUARES
.
ro
DEPENDENT VARIABLE: EDUC
RIGHT-HAND
VARIABLE
MPET2
VACACR
HRMl
C
R-SOUARED = .4577
LOG OF LIKELIHOOD FUNCTION = -145.365
DUHBIN-bATSON STATISTIC (ADJ. FOR 0. GAPS) = 2.4306
NUMBER OF OBSERVATIONS = 20.
SUM OF SQUARED RESIDUALS = .241)822*07
STANDARD ERROR OF THE REGRESSION = 387.961
SUM OF RESIDUALS = .408173-03
MEAN VALUE OF DEPENDENT VARIABLE = 646.050
F-STATISTIC( 3.. 16.) = 4.50080
ESTIMATED
COEFFICIENT
.110577
.993291-01
25.1254
-730.439
STANDARD
ERKOR
.441740-01
.505BtJ2-01
16.2500
475.462
T-
STATISTIC
<;. 50323
1.9634U
1.54561
-1.53627
MPET2
VACACH
RRMI
C
ESTIMATE OF VARIANCE-COVARIANCE MATRIX OF ESTIMATED COEFFICIENTS
MPET2 VACACR RRMI C
.195134-02
. -.320427-03
. -.682849-02
-2.84254
-.320427-03
.255917-02
.922152-01
-21.7451
-.6828*9-02
.922152-01
264.258
-2938.16
-2.84254
-2K7451
-293M.16
226064.
-------
EQUATION 12.
*«««»**»«*«»
ORDINARY LEAST SQUARES
2. INDUSTRIAL/OFFICE MODEL
DEPENDENT VARIABLE: REC
RIGHT-HAND
VARIABLE
OFFVAC
DELEMP
MINCC
AUTO
DISCBD
MANACR
C
ESTIMATED
COEFFICIENT
17.565*
14*1.24
387.011
-615.075
14.7740
186.228
-603.80*
STANDARD
ERROR
5.95976
508.061
134.366
231.318
6.05910
132.862
155.590
T-
STATISTIC
Z. 94734
2.83675
2.88029
-2.65900
2. 438.il
1.41672
-3.86074
R-SQUARED = .7504
LOG OF LIKELIHOOD FUNCTION = -120.354
DURBIN-WATSON STATISTIC (ADJ. FOR 0. C.APS) = 2.1696
NUMBER OF OBSERVATIONS = 20.
SUM OF SQUARED RESIDUALS = 197451.
STANDARD ERROR OF THE REGRESSION = 123.242
SUM OF RESIDUALS = .431538-04
MEAN VALUE OF DEPENDENT VARIABLE = 140.250
F-STATISTIC< 6.» 13.) = 6.51368
-------
2. INDUSTRIAL/OFFICE MODEL
OFFVAC
OELEMP
MINCC
AUTO
OISC8U)
MANACrt
C
ESTIMATE OF VAKlANCt-COVARIANCE MATRIX OF ESTIMATtO COEFFICItNTS
OFFVAC OELEMP HINCC AUTO
DISCBO
MANACR
35.5187
. 545.693
-63.8581
150.407
-I*. 3112
-270.879
-109.363
545.693
258126.
-2693.07
-84271.5
509.069
-22491.1
-3914.62
-63. asm
-2693.07
18054.1
-6J14.30
.765480
4227.70
-1B029.1
150.407
-84271.5
-6314.30
53508.2
-76*. 195
3770.23
6258.15
-12.3112
509.069
.765480
-764.195
36.7126
249.132
-195.276
-270.879
-22491.1
4227.70
3770.23
249.132
17652.3
-8744.56
ro
oo
OFFVAC
OELEMP
MINCC
AUTO
OISC80
MANACH
C
•109.363
•3914.62
•18029.1
6258.15
•195.276
-8744.56
24208.3
LINE 15.
SMPL
SMPL = ]
LINE 16.
OLSQ
4.
9. 13. 13. 18. 18.
-------
APPENDIX B
GRAPHS OF ACTUAL VERSUS PREDICTED LAND USE
FOR THE CROSS-VALIDATION ANALYSIS
B-l
-------
1. RESIDENTIAL MODEL
RES EQUATION
PLOT OF ACTUAL<«> *NO FITTED<*> VALUES
FLUT OF RESIDUALS(O)
INJ
10
14
21
22
23
27
29
30
35
37
40
ID
U
21
22
23
27
29
30
35
37
40
ACTUAL
•1008*05
877.0
958*0
8234*
3662*
1122*
.2338*05
3348*
0.
5990.
ACTUAL
945.0
533.0
322.0
715.0
388.0
1438.
4355.
1489.
828.0
975.0
FITTED
-1503. + «
.11)73*05 « +
1903* « *
9439. » *
.1079+05 « *
5139. «•:*..
4540. - + '
3618. . +
1804. H *
1906. * •
LINE 11.
STOP
COW EQUATION
P4.0T OF ACTUAL<»> AND FITTEO<*) VALUES
FITTED
696.3 » *
581.2 *
530.1 » *
1094. « *
8bl.6 » *
IbOS. +
2254. *
137s. +«
1344. « *
1464. « +
LINE b6.
ACTf-IT
RESIDUAL
0.0
.116 + 05 • • .0
-.985*04 0. . .
-945. . 0. .
-.121*04 . 0. .
-.71-3+04 .0 . .
-.402*04 . 0 . .
» .188*05 ...
-270. • 0 .
-.180*04 .0. •
.408*04 • • 0 •
PLOT OF RESIDUALS(O)
RESIDUAL
0.0
247. • . 0 .
-48.2 0 .
-208. . 0 . .
-379. . 0 .
-494. . 0 . .
-67.0 . 0 .
» .210+04 ...
114. • .0 .
-516. . 0 .
-4U9. . 0 . .
CO
CO
-------
OFFICE EQUATIOH
PLOT OF ACTUALO AND FITTEDC*) VALUES
PLOT Of RESIOUAL&tO)
10
14
21
22
23
27
29
30
35
37
40
ID
14
21
22
23
27
29
30
35
37
40
ACTUAL
40,00
8.000
13.00
625.0
510.0
742.0
125.0
248.0
200.4
670.0
ACTUAL
875.0
573.0
22.00
334.0
114.0
60.00
117.0
5427.
703.0
4205.
FITTdO
b^O.O ' *
278.0 « *
143.4 * *
321.9 * *
261.1 * »
479.0 * *
632.4 * *
56.91 * «
366.4 » *
618.0 .' * *
LINE 57.
ACTFIT
MANF EQUATION
PLOT OF ACTUALS) AND FITTED(*) VALUES
FITTED
407.9 * ..*
. 407.9 * * . • •
512.4 « *
432.7 «*
397.9 • *
397.9 « *
743.0 * *
769.2 *
460.0 * * '
435.7 * *
LINE 58.
ACTUT
RESIDUAL
-780.
-270.
-130.
303.
2*9.
263.
-507,
191.
-166.
52.0
0.0
0
0
PLOT OF RESIDUALS(0)
RESIDUAL
467. •
165.
-490. •
-98.7
-284.
-338.
-626.
.466*04 .
243.
.377*04 .
0.0
. 0
.0
0 .
0
0.
0.
0 .
.0
-------
HWIHNX EQUATION
PLOT Of ACTUAL<»> ANO FITTt!U<») VALUES
HLOT OF HESIDUALS(tt)
ID
u
21
22
23
27
29
30
35
37
40
10
14
21
22
23
27
29
30
35
37
40
ACTUAL
12.00
8.000
0.
35.40
6.800
16.40
103.2
21.20
20.00
26.30
ACTUAL
242.0
111.0
256.0
423.0
61.00
0.
0.
160.0
113.0
905.0
FITTED
39.73 • »
40.17 « »
8.7*3 « * .
47.29 * *
15.^3 « «•
19.87 • *
28.7* *
17.36 » •
39.4* • + •
34.40 » *
LINE 59.
*CTFIT
. . WHOLE EQUATION
PLOT OF ACTUAL (*) AND FITTED <*> VALUES
FITTEO
554.6 « »
-32.20 * «
-77.35 » •
244.2 * «
-242.1 + •
-114.* + .
226.9 » 4
44. B6 + *
202. t) » +
742.3 »
LINE 60.
ACTFIT
HE SI DUAL
32^2
8*74
11.9
U.43
3.47
74.5
3.84
19.4
8.10
0.0
RESIDUAL
-313.
143.
333.
179.
303.
114.
-227.
115.
-«9.8
163.
0 .
0.
.0
PLOT Of RESIDUALS(O)
0.0
0 .
-------
1IOTEL EQUATION
PLOT OF ACTUALC) ANU HTTED<*) VALUES
PLOT OF RESIOUALS(O)
ID
14
21
22
23
27
29
30
35
37
40
CO
1
(71
ACTUAL
77.00
160.0
4.000
0.
45.00
195.0
117.0
121.0
135.0
168.0
FITTED
70.07
OJ.5D
12.&7 « »
43. Ob «
123.1
Btt.13
62. 2S
68.50
59.52
84.88
LINE 61.
ACTFIT
RESIDUAL
0.0
»* 6.93 • .0
* « 76.5 . .
-8.67 . 0.
+ -43.1 .0 .
« * -78.1 0 .
* » 107.
+ « 54.8 ••
+' « 52.5
* ...••« 75.5 . .
* * 83.1 . .
HOSPTL EQUATION
0
0
PLOT OF ACTUALO AND FITTED(») VALUES
PLOT OF HESIOUALS(O)
10
14
21
22
23
27
29
30
35
37
40
ACTUAL
40.00
87.00
0.
0.
0.
203.0
359.0
155.0
30.00
0.
FITTED
33.86 *
-8.691 » «
16.27 « »
-44.74 * *
19.07 « »
95.93 * *
55.24 »
13.84 «• •
106.0 * *
-42.29 + «
RESIDUAL
6.14
95.7
-16.3
44,7
-19.1
107.
* 304.
141.
-76.0
42.3
0.0
0
. 0 .
. o« •
• .0 .
.0. «
0.
• • . •
.1
• 0 • .
. . 0 .
LINE 62.
ACTFIT
-------
CULTUR EQUATION
PLOT Of ACTUAL<*> ANU FITTElM*) VALUES
PLOT OF RESIDUALS(0)
DO
I
Ot
ID
>
44
21
22
23
27
29
30
35
37
40
ACTUAL
0.
55.00
0.
4.000
21.00
0.
47.00
0.
75.00
15.00
FITTED
11. V9 • *
12.13 * *
10.94 « »
12.51 • *
12.87 * «
7.403 « *
13.48 «• . . . • •
11.70 » *
12.13 »
13.92 **
RESIDUAL
-12.0
42.9
-10.9
-8.51
8.13
-7.40
33.5
-11.7
* *2.9
1.08
0.0
0 .
. .
0 .
0 .
• •
0 .
• ' .
0 .
• •
.0
LINE 63.
ACTFIT
CHURCH EQUATION
.0
PLOT OF ACTUAL(*> AND FITTED(*) VALUES
PLOT OF RESIDUALS(0)
ID
14
21
22
23
27
29
30
35
37
40
ACTUAL
45.00
3.000
63.00
280.0
35.00
23.00
504.0
152.0
135.0
108.0
FITTED
73.64 . * ' *
37.00 » *
73.03 ** ' .
81.1? *. * .
41. b7 **
63.48 * *
121.1 * ' .
75.01 * «
94.15 * «
118.4 **
RESIDUAL
-28.6
-34.0
-10.0 •
199.
-6.87 .
-40.5
« 383.
77.0
40.9 .
-10.4 .
0.0
0.
0.
0
•
0
0.
•
0.
. 0
0
LINE 64.
ACTFIT
-------
EOUC EQUATION
PLOT OF ACTUAL!*) AND FITTtLH*) VALUES
PLOT OF KESIOUALb<0>
14
21
22
23
27
29
30
35
37
40
ACTUAL
336.0
256.0
121.0
1176.
350.0
359.0
1504.
496.0
330.0
1208.
FITTED
454.3 * +
381. S • »
339.9 » *
553.9 *
423.4 « *
162.2 * »
732.6 *
504.4 »+
551.7 * t
785.8 *
HESIDUAL
i
-118. • 0
-125. • 0
-219. . 0
« 622.
-73.4 . 0
177.
» 771.
-8.42
-222. • 0
» 422.
LINE 65.
*CTFIT
0.0
. 0
CD
I
REC EQUATION
PLOT OF ACTUALO AND FITTED!*) VALUES
PLOT OF RESIOUALS(O)
10
14
21
22
23
27
29
30
35
37
40
ACTUAL
20.00
0.
0.
0.
112.0
1250.
31.00
66.00
0.
746.0
FITTED
122.9 « +
54.55 * *
21.08 »*
-11. 74 ««
29.30 * *' '
658.1 *
35.52 *
182.5 * *•
90.66 « *
-21.20 * *
LINE 66.
STOP
RESIDUAL
-103.
-54.5
-21.1
17.7
82.7
« 592.
-4.52
-117.
-90.9
767.
0.0
0 .
0.
0.
0
• •
• •
0
0 .
0 .
• *
-------
2. INDUSTRIAL/OFFICE MODEL
RES EQUATION
PLOT OF ACTUALC) AND HTTED<»> VALUES
10
8
17
18
24
28
31
32
33
38
39
DO
09
10
8
17
18
24
28
31
32
33
38
39
ACTUAL
.1965*05
.2154*05
5674.
3608.
.1054*05
6697.
5452.
5938.
.1084*05
830.0
ACTUAL
4300.
1497.
360.0
731.0
1519.
922.0
408.0
812.0
291.0
107.0
FITTED
.1667*05 «. «
.2486*05 «
.1215*05 « *
5723. « ,
.1185*05 » »
9218. » *
.1061*05 » +
3604. » »
3031. «. «
-7083. * •
LINE 55.
ACTFIT
COW EQUATION
Plot of Actual (*) AND FITTEO<») VALUES
FITTED
2873. *
1788. • +
816.0 • *
2155. « *
995.0 * «
829.5 * •
1327. • »
546.5 * •
890. 6 * *
481.4 « *
LINE 56.
ACTFIT
RESIDUAL
.299*04
PLOT OK RESIDUALS(O)
0.0
* -.332*04 . 0
-.647*04 0 .
-.212*04
-.131*04
-.252*04
-.516*04 0.
.233*04 .
.781*04 .
.791*04
0 .
RESIDUAL
* .143*04
-291.
-456.
-.142*04 0
524.
92.5
-919.
265.
-600.
-374.
PLOT OF RESIDUALS(0)
0.0
0.
-------
OFFICE EQUATION
PLOT OF ACTUAL!*) AND FITT£U<«) VALUED
OF HESIDUALS101
8
17
18
24
28
31
32
33
38
39
ACTUAL
573.0
86.00
0.
80.00
188.0
78.00
192.0
135.0
322.0
6.000
FITTED
-236.1 * '
803.8 «
115.5 • *
346.8 « • ' *
1 81 . 1 *
340.7 » *
256.8 * *
105.8 **
-196.7 * «
108.4 • *
KESIOUAL
» 8090
* -718. 0
-116o
-267.
6.95
-263.
-64.8
29.2
519.
-102o
0.0
• »
e e
o .
. 0
0
. 0
0.
.0
• • •
0.
LINE 57.
ACTFIT
.0
MANF EQUATION
PLOT OF ACTUALC*) AND FITTEO<*» VALUES
PLOT OF RESIDUALS(0)
10
8
17
18
24
28
31
32
33
38
39
ACTUAL
6730.
144.0
403.0
3359.
1091.
500.0
352.0
3484.
142.0
0.
FITTEO
2009. *
1999. • *.
1924. • *
965.4 * *
1522. « *
1746. • *
1838. « *
821.2 * *
-215.2 * •
-1194. * •
WESIOUAL
» .472*04 .
-.185*04 .0
-.152*04 . 0
.239*04 .
-431. •
-.125*04 . 0
-.149*04 . 0
.266*04 .
357.
.119*04 .
LINE 58.
ACTFIT
0.0
.0
-------
HWUMNX EQUATION
PLOT OF ACTUAL<»> AND FITTEDC*) VALUES
PLOT OF KESIDU4LS<0>
CD
I
10
e
17
18
24
28
31
32
33
38
39
10
8
17
18
24
28
31
32
33
38
39
ACTUAL
56.40
81.20
8.000
12.80
23.60
23.40
32.50
47.20
11.20
12.00
ACTUAL
2605.
4046.
29.00
668.0
599.0
0.
152.0
189.0
232.0
7.000
FITTED
SO. 85 * «
99.89 »
35.24 « *
21.44 » >
57.14 * *
29.02 « *
41.17 » *
8.524 •' * •
4.075 * •
-9.669 * *
LINE 59.
ACTFIT
WHOLE EQUATION
PLOT OF ACTUAL (*) AND FITTEDI*) VALUES
FITTED
5780. «
4681. » +
-1180. » •
957.8 • *
738.1 •»
-656.4 * *
2042. • *
40.12 *•
1202. • *
405.8 * »
LINE 60.
ACTFIT
HESIDUAL
5.55
-18.7
-27.2
-8.64
-33.5
-5.62
-8.67
38.7
7.13
21.7
0.0
.0 . .
0 . . .
• 0 . .
0 .
. 0 . .
0 . .
PLOT OF RESIDUALS(0)
RESIDUAL
-.317*04 0
-635.
.121*04
-290.
-139.
656.
-.189*04
149.
-970.
-399.
0.0
0 .
0.
. 0
.0
-------
HOTEL EQUATION
PLOT OF ACTJAL<«> AND FITTEU1*) VALUES
PLOT Of H£SIDUALS<0)
ID
CD
I
8
17
16
24
28
31
32
33
38
39
ACTUAL
583.0
58.00
29.00
1*7.0
41.00
285.0
0.
20.00
35.00
11.00
FITTED
207.3 *
-103.1 * •
50.24 *.*
-117.3 «• «
80.89 • • »
105.7 • * . *
114.7 * *
82.45 * *
6.6U7 * *
69.30 • *
RESIDUAL
« 376,
161,
-21o2
2640
-39.9
179.
-115.
-62.4
26.4
-58.3
LINE 61.
ACTFIT
HOSPTL EQUATION
. 0
0.0
PLOT OF ACTUAL**) AND FmEO(*) VALUES
PLOT OF HESIDUALS(O)
ID
1
8
17
18
24
28
31
32
33
38
39
ACTUAL
585.0
0.
72.00
3.000
0.
125.0
0.
0.
63.00
23.00
FITTED
409.2 * » .
925.6 *
273.7 » *
307.9 * *
328.3 *' «•
137.3 **
-81.38 * •
-87.57 * •
210.2 * *
-92.35 » «
RESIDUAL
0.0
176o • • 0
* -926. 0 .
-202. • 0
-305. .0
-328. 0
-12.3 .0.
81.4 . «0
87.6 .. •"
-147. 0 .
115. • .0
LINE 62.
ACTFIT
-------
CULTUR EQUATION
PLOT OF ACTUALC) AND FITTED!*) VALUES
PLOT OF kfcSlOUALSlO)
DO
I
10
8
17
18
24
28
31
32
33
36
39
ID
^
17
18
24
28
31
32
33
38
39
ACTUAL
45.00
0.
0.
ll.frO
44.00
59.00
56.00
0.
0.
0.
ACTUAL
305.0
75.00
58.00
77.00
138.0
144.0
22.00
18.00
51.00
25.00
FITTED
1,179
-3,813
-.9939
4.287
4.998
15.05
-24.88 *
-7.231
7.338
40.13
LINE 63.
ACTFIT
FITTED
209.2
39. 2«
69.14
123.3
111.7
81.86
49.34
-45.10 *
215.6
15.02
LINE 64.
ACTFIT
RESIDUAL
* * 43.8
* * 3.81
•** .994
* '• 6.71
* * 39.0
* • 43.9
« 80.9
* « 7.23
» •* -7.34
« • -10,1
CHURCH EQUATION
PLOT OF ACTUAL («) AND FITTE0<») VALUES
RESIDUAL
* * 95.8
» * 35.7
* * -11.1
* * -46,3
* • 26.3
* * ' 62.1
' * -27.3
* 63.1
* * -165.
* • 9.98
0,0
.0
0
. 0
. 0
.0
. 0
. 0
0.
0 .
PLOT OF R£SIOUALS(0)
0.0
• . . o
. . o ,
• 0. •
• 0 . .
. 0
»
o .
•
•
.0
0.
0.
-------
EDUC EQUATION
HLOT UF ACTUALJ*) AND FlTTtlM*) VALUES
i^LOT OF KtSlDUALS(O)
DO
I
ID
8
17
18
24
28
31
32
33
38
39
ID
8
17
18
24
28
31
32
33
38
39
ACTUAL
1483.
576.0
216.0
401. 0
736.0
328.0
724.0
353.0
344.0
102.0
ACTUAL
0.
500.0
325.0
450.0
65.00
0.
56.00
150.0
13.00
3.000
FITTED
919.3
459.1
4t>9.4
781.3
859. H
487.5
7^4.4
326. d
337. F,
-387.^ +
LlNt 65.
ACTFIT
FITTED
704.8
1516.
1385.
886. n
47.4?
462.3
-576.5 +
-4V.42
3/8. (i
378.8
LINE 66.
KESIDUAL
0.0
* « 564.
* ° 117. . . 0 .
• * -253. .0
» + -380. 0 .
» * -124. 0 .
« * -159. 0
* -.407 . 0
* 26.8 . 0
* 6.46 . 0 .
* 489.
REC EQUATION
PLOT OF AtTOAL(«> AND FITTEO<») VALUES PLOT OF RESIDUALS(O)
RESIDUAL
0.0
* * -765. 0 .
* * -.102*04 0
* * -.106*04 U
* * -436. .0 .
* 17.6 . 0
» * -482. . 0
» , 632.
* * 199. . .0
» * -365. . 0
9 " -376. . 0 .
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-45Q/3-7fi-n1?r
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Growth Effects of Major Land Use Projects,
Volume III - Summary
5. REPORT DATE
September 1976
fi. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Frank H. Benesh
Peter Guldb,erg, Ralph D'Agostino
8. PERFORMING ORGANIZATION REPORT NO.
C-781 - c
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Walden Research Division of Abcor
850 Main Street
Wilmington, Massachusetts 01887
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-2076
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Protection Agency
Office of Air Quality Planning and Standards
Strategies and Air Standards Division (MD-12)
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Growth Effects of Major Land Use Projects is a research program whose goal is to
formulate a methodology to predict air pollutant emissions resulting from the
construction and operation of two types of major land use projects, large residential
projects and large concentrations of employment (i.e., office parks and industrial
parks). Emissions are quantified from the major project, from land use induced
by the major project, from secondary activity occurring off-site (ie., generation
of electricity by utilities), and from motor vehicle traffic associated with both
the major project and its induced land uses.
This volume provides a summary of the first two volumes, viz. the specification
and causal analysis of the land use model and the development of the land use
based emission factors. It also discusses the development of the predictive
equations in the land use model and the development of the traffic model.
A set of compulation worksheets and step by step instructions for using t.he
GEMLUP model, as well as an example of their use, are provided.
Previous volumes document the development of the land use model and the
development of a set of land use based emission factors.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Land Use
Planning
Industrial Areas
Residential Areas
Secondary Effects
Induced Land Use
Emission Factors
8. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (Thispage)
JInclassified
22. PRICE
EPA Form 2220-1 (9-73)
------- |