&EPA
           United States
           Environmental Protection
           Agency
          Office of Air Quality
          Planning and Standards
          Research Triangle Park NC 27711
EPA-450/3-78-014b
May 1978
           Air
Growth Effects of Major
Land Use Projects
(Wastewater Facilities)
Volume II:  Summary,
Predictive Equations
and Worksheets

-------
                                      EPA-450/3-78-014b
Growth Effects of Major  Land  Use  Projects
              (Wastewater Facilities)
 Volume II:  Summary, Predictive Equations
                  and Worksheets
                            by

                 Peter H. Guldberg and Ralph B. D'Agostino

                    Walden Division of Abcor, Inc.
                         850 Main Street
                      Wilmington, MA 01887
                      Contract No. 68-02-2594
                   EPA Project Officer: Thomas McCurdy
                          Prepared for

                 U.S. ENVIRONMENTAL PROTECTION AGENCY
                    Office of Air, Noise, and Radiation
                 Office of Air Quality Planning and Standards
                 Research Triangle Park, North Carolina 27711

                          May 1978

-------
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 - as  supplies permit - from the
Land Use Planning Office, Office of Air Quality Planning and Standards,
Environmental Protection Agency, Research Triangle Park, North Carolina
27711; or, for a nominal fee, from the National Technical Information
Service, 5285 Port Royal Road, Springfield, Virginia 22151.
This report was furnished to the Environmental Protection Agency by
Wai den Division of Abcor, Inc., Wilmington, Massachusetts 01887, in
fulfillment of Contract No. 68-02-2594.  This report has been reviewed
by the Land Use Planning Office, EPA and approved for publication.  Ap-
proval does not signify that the contents necessarily reflect the views
and policies of the Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or recommendation
for use.
                   Publication No. EPA-450/3-78-014b

-------
                            ACKNOWLEDGEMENTS

     Special appreciation goes to Mr. Thomas McCurdy, the U.S. Environmental
Protection Agency Project Officer for this study, whose extensive assistence
and advice was indispensable in the conduct at the study.  Appreciation is
also extended to all EPA technical committee members for guidance given
throughout the study.  The technical committee included representatives
of the Office of Transportation and Land Use Policy, the Municipal Con-
struction Division  (Facilities Requirements Branch) and the Control Program
Development Division.

     In addition,  we wish to thank the more than one-hundred individuals in
city/county/regional planning agencies and transportation departments
nationwide who cooperated with us during the data collection task and  helped
provide the data on which this study is based.   Their time and cooperation
were invaluable.

     Urban Systems Research & Engineering, Inc.  (USRE)  of Cambridge,  MA were
subcontractors to  Walden on this  study, assisting in the definition of basic
model  concepts,  infrastructure relationships and exogenous variables,  and
in the testing and refinement of  the causal and  predictive models.  The
cooperation of Dr. James F. Hudson,  who guided the USRE technical effort, is
deeply appreciated.

     Finally,  recognition is due  the Walden staff who performed the long and
difficult task of field data collection - Mr.  Mahesh Shah, Ms. Diane  Gilbert,
Mr.  Michael Geraghty and Mr.  Keith Kennedy.
                                        m

-------
                            TABLE OF CONTENTS
Section
                   Title
Page
   II
  III
INTRODUCTION 	   1-1

A.  Study Overview	   1-1
B.  General Approach 	   1-3
C.  Summary of Study Results and Conclusions ...   1-6
    1.  Volume I  - Model Specification and
        Causal Analysis  	   1-6
    2.  Volume II - Predictive Equations and
        Worksheets	   1-7


PHASE IV '- DEVELOPMENT OF PREDICTIVE MODEL ....   2-1

A.  Land Use Predictive Equations	   2-1
    1.  Approach	   2-1
    2.  Discussion of Results	   2-4
B.  Model Validation 	   2-9
    1.  Cross-Validation 	   2-9
    2.  Weight Validity Index	   2-10
C.  Coefficient Stability Analysis	   2-12
D.  Emission Projection Worksheets 	   2-17
    1.  Definitions	   2-19
    2.  Input Data Requirements	   2-19
    3.  Instructions	   2-25
    4.  English Unit Worksheets	   2-41
    5.  Metric Unit Worksheets	   2-59
    6.  Example on Using Worksheets  	   2-76

REFERENCES	   3-1

APPENDIX A - Complete Statistical Output of the
             Predictive Equations	   A-l


APPENDIX B - Glossary of Terms	   B-l

APPENDIX C - Definition of  Model Variables ....   C-l
              APPENDIX D - Graphs of Actual  Versus Predicted
                           Land  Use for  the  Cross Validation
                           Analysis  	
              APPENDIX E -  Supplementary Information
                                                     D-l

                                                     E-l
                                       IV

-------
                             LIST OF TABLES

Table Number                      Title                           Page

  2-1 a        Summary Statistics of the Land Use
              Predictive Equations ...............   2-5
  2-lb        Summary Statistics of the Disaggregation
              Equations  ....................   2-5
  2-2         Coefficients of Validity Between Actual and
              Predicted Land Use from the Cross-Validation
              Analysis .....................   2-11
  2-3         Weight Validity Indices for the Land Use
              Predictive Equations ...............   2-13
  2-%         Summary Statistics of the Predictive Equations
              Coefficient Stability Analysis ..........   2-14
-

              Input Data Requirements for Impact Assessment
              Model  ......................   2-20
  2-6         Default Trip Generation Rates  ..........   2-27
  2-7         Parti cul ate and Sulfur Oxide Emission Factors  .   .   2-29
  2-8         Single Family Detached or Attached Land Use
              Based Emission Factors ..............   2-31
  2-9         Multiple Family Land Use Based Emission Factors   .   2-32
  2-10        Commercial and Wholesale Land Use Based
              Emission Factors .................   2-33
  2-11        Office-Professional Land Use Based Emission
              Factors  .....................   2-34
  2-12        Education Land Use Based Emission Factors  ....   2-35
  2-13        Other Land Use Based Emission Factors  ......   2-36
  2-14        Estimated National  Industrial Land Use Based
              Emission Factors by Two Digit 1967 Standard
              Industrial Classification Code ..........   2-37
  2-15        Typical Uncontrolled Emission Factors for
              Electric Utilities   ...............   2-42
  C-l         Definition of Endogenous Model Variables .....   C-3
  C-2         Definition of Exogenous Model Variables  .....   C-5
  C-3         Metric Units of Variables Used In the Predictive
              Equations and Worksheets .............   C-16

-------
Figure Number
                       LIST OF FIGURES AND WORKSHEETS
Title
Page
     1-1      Technical Approach to the Development of a
              Statistical Model for Predicting the Growth
              Effects of Major Wastewater Projects 	   1-4
     2-1      Overview of the Land Use and Air Quality
              Impact Assessment Procedure  	   2-18
     2-2      Normal Seasonal Heating Degree Days  	   2-38
     2-3      Normal Seasonal Cooling Degree Days  	   2-39
     2-4      Annual Air Conditioner Compressor Hours  	   2-40

Worksheet Number                  Title                           Page

     1/1M*    Input Data Record  	  2-43/2-60
     2/2M     Land Use Projections	2-46/2-63
     3/3M     Final Land Use Projections  	  2-47/2-64
     4/4M     Default  Disaggregation Equations 	  2-48/2-65
     5/5M     Confidence Intervals 	  2-49/2-66
     6/6M     Motor Vehicle Trips  	  2-51/2-68
        7      Vehicle  Miles Traveled (VMT) 	  2-52
        7M     Vehicle  Kilometers Traveled (VKT)   	  2-69
        8      Categorized VMT	2-53
        8M     Categorized VKT	2-70
     9/9M     Motor Vehicle  Emission Factors  	  2-54/2-71
     10/10M    Composite  Motor Vehicle Emission Factors 	  2-55/2-72
     11/11M    Total Motor Vehicle  Emissions   	  2-56/2-73
     12/12M    Stationary Source  Emissions  	  2-57/2-74
     13/13M    Emissions  Summary  	  2-58/2-75
*The first set of  numbers  refers  to worksheets  in  English units,
 while the second  set refers  to metric  units.
                                           VI

-------
 I.  INTRODUCTION

     A.   STUDY OVERVIEW

         Pursuant to 40 CFR 51.12(d)-(h),  State  Implementation  Plans  must
contain provisions to prevent any national  ambient  air  quality  standards
from being exceeded.  These provisions  are  called Air Quality Maintenance
Area (AQMA) plans, and estimating the  air  quality impact  of major  land
use and urban development projects is  a necessary part  of AQMA  planning
[1-6].   In addition, the National  Environmental  Policy  Act [7]  and the
Council  on Environmental Quality (CEQ)  "Guidelines  on the Preparation of
Environmental Impact Statements" [8]  require  the consideration  of  secondary
impacts from major projects.   CEQ states that:

         "Many federal  actions,  in particular those that  involve the
         construction or licensing of  infrastructure investments (e.g.,
         highways, airports,  sewer systems, water resource projects,
         etc.), stimulate or induce secondary effects in  the form  of
         associated investments  and changed patterns of social  and
         economic activities.  Such secondary effects through their
         impacts on existing community  facilities and activities,  or
         through changes in natural conditions,  may often be more
         substantial than the primary effects of the original action
         itself [8]."

This has been a particular concern for  wastewater systems, since their
primary impacts tend to be small,  i.e., sewers and  treatment plants generally
improve water quality, but they  may lead to significant negative secondary
impacts.  The probable large  indirect  impacts (redirecting growth  and in-
ducing development) of a new or  expanded regional sewage  treatment facility
on ambient air quality, and the  need  for some procedure to ascertain  what
its impact will be, is recognized in  the AQMA Guideline series([1 ]:A-7ff,
[4]:21ff).  To date, EPA has  not developed a  model  to estimate  what the
ambient impacts will be for use  in AQMA planning.  Thus,  it was the purpose
of this study to develop such a  model.

         This study is entitled  the Growth Effects  of Major Land Use  Projects
(GEMLUP-II), and it addresses the induced  growth effects  of wastewater major
                                    1-1

-------
projects.  Similar research on two other major project types, large residential
developments and large industrial/office parks, has been performed previously
and is reported on in the three volume set of GEMLUP-I final  reports [9-11].
The secondary air quality impacts of wastewater projects are determined by
the land use growth induced by such projects.  That is, the air impacts are
emissions from (1) residential complexes that appear at the end of and along
new sewer lines, (2) other service-oriented land uses (commercial, industrial,
office, government) that relate to residential development, and (3) motor
vehicles used as transportation between development areas.  Thus, the key to
understanding secondary air quality impacts is to first understand the growth
effects of a wastewater facility on land use in a region.  The objectives of
the study effort were:  (1) to develop and test a path-analytic causal
model  that represents the induced land use ten years after construction of a
wastewater major project, (2) to develop and validate a simplified predictive
model  of induced development, (3) to test and correct the GEMLUP-I VMT model
[11],  and (4) to develop worksheets  that can be used to predict induced land
use and associated air pollution emissions.  For these purposes, data were
collected from forty  (40) case study wastewater projects nationwide.

         The study project was divided into four major phases:

         I.  Definition of basic concepts and initial model specification,
         II. Data collection,
         III.Causal analysis of the land use model using path analysis,
         IV. Development of predictive equations for the land use model
             and worksheet procedures.

Two  separate technical reports are to be prepared.  This is the second volume
report and covers Phase IV of the study.  The remainder of this chapter gives
an outline of the general study approach, and a summary of results and con-
clusions for the entire study.  Chapters II and Appendix A summarize the
technical performance  on Phase IV, including the development of the GEMLUP-II
model  worksheets.   The technical performance on Phases I-III is summarized
in the Volume I final report  [12].  Appendix B gives a.glossary of terms used
in this report, and Appendix  C summarizes definitions for the variables used
in the GEMLUP-II model.
                                     1-2

-------
     B.   GENERAL APPROACH

         The principal  objectives  of  this  study were  to  formulate  a  statis-
tical methodology to predict air pollutant emissions  from:

         •   Induced land development  associated with  the construction  and
            operation of wastewater facilities  in  a community.*
         •   Motor vehicular traffic associated with the  induced land
            development.

The ability to accurately predict the secondary development induced  by major
wastewater  collection and treatment projects  is dependent on understanding
the complex interrelationships inherent in such a  model.  Thus, an important
objective was to formulate and test a causal  theory of induced development
using path  analysis.  Additional objectives were  to  test and correct the
motor vehicle traffic model developed as part of  the  previous GEMLUP-I
study [11]  for use in the air pollutant emissions  projection procedure, and
to integrate the predictive land use  model, the traffic model, and land use
based emission factors in a set of easy-to-use  worksheets.

         The approach to fulfilling these objectives  involved the  sequential
execution of four separate project phases shown schematically in Figure 1-1
and  summarized below.

         In Phase I, our primary interest was to  define the basic  concepts
on which this study  is  based.  The first step in  this process was  to
determine the modeling approach.  Next, the infrastructure causal  relation-
ships of wastewater  facilities in communities were studied and the knowledge
used to define "induced development"  in the model.  The concepts of  a
"wastewater major project" and the "area of analysis" for induced development
were also studied and defined.  Knowing the important causal relationships
enabled us to develop a list of model variables representative of the relevant
factors involved  (e.g., the major project, land use,  regional growth).
*
 The causal model for induced development produced by this study does not
 take  into account the effects of mitigating measures.  Variables measuring
 restrictions on on-site disposal and hookups to existing interceptor
 lines were included in the development of the model, however, see Volume I,
 Section  II [12]. '

                                      1-3

-------
         PHASE  I
INITIAL MODEL  SPECIFICATION
   PHASE II
DATA COLLECTION
                                                         PHASE III
                                                      CAUSAL ANALYSIS
                                                                                                     PHASE IV
                                                                                        DEVELOPMENT OF PREDICTIVE  EQUATIONS
                      DEFINE
                     MODELING
                     APPROACH
                     IDENTIFY
                  INFRASTRUCTURE
                  RELATIONSHIPS
                  _1
                   DEFINE MAJOR
                   PROJECT ADD
                    AREA OF
                    ANALYSIS
                  INITIAL CHOICE
                  OF VARIABLES
                                  COEFFICIENT
                                   STABILITY
                                    ANALYSIS
                                                                   DETERMINE
                                                                   .NET CAUSAL
                                                                    EFFECTS
PRELIMINARY
CASE  STUDY
  SURVEY
  LIST OF
 POTENTIAL
CASE STUDIES
                                                                                          MODEL
                                                                                        VALIDATION
                                                                                         PREDICTIVE
                                                                                       LAND USE MODEL
                                                                                         EQUATIONS
                                            FINAL  PATH
                                          ANALYTIC LAND
                                            USE MODEL
                                              CAUSAL
                                         ANALYSIS REPORT
                                            (VOLUME I)
  LIST OF
    DATA
REQUIREMENTS
                                                                                                     SL'I-"!ARY  RE:
                                                                                                           E  :::
CASE STUDY
SELECTION
                                                                                                        DEVELC?
                                                                                                       E'-'ISSIO'-S
                                                                                                       PROJECTION
                                                                                                       VIORKSHEETS
   SPECIFY
 PRELIMINARY
    MODEL
                                                                                        STEPWISE
                                                                                        REGRESSION
                                                                                        ANALYSIS
   DATA
 COLLECTION
                                              THEORY
                                             TRIMMING
                                                         VALIDATE
                                                          GEMLUP
                                                         TRAFFIC
                                                          MODEL
 INITIAL
ANALYTIC
    E MODEL
                                                                COMPUTER
                                                                DATA  FILE
                                                                     FIGURE  1-1
                                   TECHNICAL  APPROACH  TO THE  DEVELOPMENT OF A  STATISTICAL MODEL
                                  FOR  PREDICTING THE  GROWTH EFFECTS OF MAJOR WASTEWATER  PROJECTS

-------
Finally,  specific causal  relationships between the  variables  were  hypothesized
and formalized in an initial  causal  model.

         The principal  objective of Phase II  was to collect a sufficiently
large, diverse, and thereby representative,  cross-sectional  data base on which
to develop the model.   Due to the critical  importance of the quality of this
data base and the large manpower effort required to collect such data, the
first task required a  careful selection of 40 case  study wastewater major
projects  distributed on a nationwide basis,  which had the potential, upon
construction, to induce a significant quantity of land development in their
communities and for which all the requisite data were available.  To this end,
a case study mail survey based upon the screening of over 15,000 federally
funded wastewater collection and treatment projects nationwide was performed.
Final selection of the 40 case study projects was made, a data collection
training  course was conducted for field personnel,  followed by site visits
to the case study areas to collect the required data.

         The objectives of Phase III were to develop the final causal model
and validate the GEMLUP-I VMT model [11].  The initial causal model was
tested using the set of case study data and the statistical techniques of
path analysis.  This approach verified which of the hypothesized causal
relationships were significant, trimmed those that were not, and determined
model parameters for the final causal model.  Tasks were also performed to
trace the direct and indirect effects the model variables have on one another
and to validate and correct the GEMLUP-I VMT model.

         In Phase IV,  predictive equations of induced land use associated
with wastewater major projects were developed and validated.  These predic-
tive equations, the GEMLUP traffic model, and GEMLUP land use based emission
factors  [10] were then used as the basis for an emissions projection procedure,
The procedure was developed in worksheet form to serve as a generalized
analytical tool for use by planners and environmental engineers in predicting
the induced land use and air pollutant impacts associated with major waste-
water collection and treatment systems.
                                       1-5

-------
     C.  SUMMARY OF STUDY RESULTS AND CONCLUSIONS

         1.  Volume I - Model Specification and Causal Analysis [12]

             A theoretical model of the Growth Effects of Major Land Use
Projects has been developed.  This model represents the total land use in
the drainage basin of a wastewater collection and treatment system (the
major project), ten years after its construction.  The model represents the
process of induced land use growth in the following 9 land use categories:

             Residential              Education
             Commercial               Recreation
             Office-Professional      Wholesale/Warehouse
             Manufacturing            Other
             Highways  (Non-expressway)

The assumption of a single  basic causal structure for induced development,
and the use of cross-sectional  data from 40 diverse case study major projects
throughout the United  States, allowed a static approach to the testing of
the theoretical model,  using path analysis.

             Path analysis  is a set of  statistical techniques useful in test-
ing theories and studying the logical consequences of various hypotheses
involving  causal relations.  It is not  capable of deducing or generating
causal  relations, only testing  them.  The causal analysis of induced land use
development  in  the  current  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 consis-
tent  estimates  of the  path  coefficients 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 9 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 in an itera-
tive  process to trim unneeded and undesirable  paths from the models.  The
                                        1-6

-------
trimming process eliminated almost half of the  paths  in  the  models  as
originally specified.   The  statistical  problems of multicollinearity,
suppressor variables and identification were eliminated  through the approach
used to trim the initial model.

             The final  models of land use development show that strong statis-
tical  relationships exist between the variables representing the 9  categories
of total land use and the other model variables representing induced and
non-induced land use growth processes.   The results indicate that the final
                                                                           2
model  explains more than half of the variance in the case study data with R
values ranging from 0.27 to 0.82 and averaging  0.54.   The residuals of the
final  regressions do not exhibit any trends or  patterns, indicating the
                                   2
remaining unexplained variance (1-R ) is not due to poor specification of
the model, but rather due to wide variance in the case study data (i.e., the
problem of trying to develop one generalized model for a broad range of
situations).  An analysis of the stability of the model  coefficients deter-
mined that, in general, the coefficient values  have low variance (± 15%)
and exhibit no extreme instabilities.  An analysis of the net causal effects
in the land use model  indicates that reserve collection  system capacity of a
wastewater major project is a significant causal factor  for induced land use
growth, principally in the residential, manufacturing, education and highways
categories.  By contrast, treatment plant capacity was not found to be an
important causal factor.

             The GEMLUP-I VMT model [11] was validated using transportation
data from 11 of the 40 case study major projects.  Based on the validation
results, it was concluded that revisions to both the default predictive
equations for trip length and the default values for trip generation rates,
were necessary.  The revised VMT model  was validated and found to have an
average error (imprecision) of 23%, with no statistically significant bias.

         2.  Volume II - Predictive Equations and Worksheets

             The development of predictive equations for land use development,
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

                                     1-7

-------
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.

             Predictive equations were developed for the 9 land use cate-
gories used in the causal analysis.  In addition, predictive equations were
developed for disaggregating residential and commercial land use totals into
3 subcategories each.  Logit analysis [14] was used with ordinary least squares
                                                          2
regression to develop the disaggregation equations.  The R  statistics
averaged 0.67 and 0.76 for the land use and disaggregation equations, re-
spectively.  Thus, the predictive equations explained the majority of the
variance in the dependent variables.  Overall F statistics indicate all
equations are significant at or below the 1 percent level.  The average error
associated with use of the predictive land  use equations is ±63 percent.
These errors are reasonable for a generalized tool, applicable nationwide,
and are  less than those  introduced subsequently in the impact assessment
procedure [16].

             A model validation using the techniques of cross-validation and
weight validity index [15] concluded that 8 of the 9 land use predictive
equations are generalized enough to produce good predictions on an independent
sample.   Conflicting results were obtained  for the OTHER equation, which may
not be well generalized.

             A coefficient stability analysis testing for worst case  in-
stabilities concluded that no  significant instabilities occur in 7 of the 9
land  use predictive equations.  The exceptions, COMM and OTHER, contain
instabilities which indicate that the true  functional form of these equations
changes  with the magnitude of  the data.

             The  predictive land use equations were included in an impact
assessment  procedure that estimates the total air  pollutant emissions
associated  with the  induced development from a wastewater major project.
This  impact assessment procedure was formalized in a set of easy-to-use
worksheets, presented in Section II.D.  These worksheets serve as an
operational tool for environmental engineers or planners to assess the
secondary air quality impacts  associated with new or expanded wastewater
facilities.                              _
                                       I -o

-------
II.  PHASE IV - DEVELOPMENT OF PREDICTIVE  MODEL

     The objective of the fourth and final  phase  of this  study  was  to
create and validate a predictive model  to  estimate  the  amount of  induced
land use (by category type) associated  with major wastewater  projects.
These predictive equations were then incorporated along with  the  revised
VMT model  [12] and GEMLUP land use based emission factors [10]  into an  emis-
sions projection procedure in worksheet form.

     A.  LAND USE PREDICTIVE EQUATIONS

         1.   Approach

             The development of predictive equations for  land use de-
velopment, separate from the model equations obtained in  the  causal analysis
(see Volume I [12]), was necessitated by the simultaneity of  the  causal
relationships, i.e., causal equations include independent variables whose
values will  not be known in the future. Therefore* it was necessary to de-
velop 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 are  biased.   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  is 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
coefficients 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.
 See Appendix B for a definition of terms.   See Appendix C for a definition
 of model variables.
                                    2-1

-------
             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 represented
the 9 categories of total land use analyzed in the causal analysis and 6
percentage disaggregation categories for residential and commercial  land use

                 • RES   = Residential
                 • COMM  = Commercial
                 • OFFICE= Office-professional
                 • MANF  = Manufacturing
                 • HIWAYS= Non-expressway highways
                 • EDUC  = Education
                 • REC   = Recreation
                 • WHOLE = Wholesale/warehousing
                 • OTHER = Hotel/Motel, culture, religion

                 Residential

                 • SFDET = Single family detached
                 • SFATT = Single family attached
                 • MF    = Multiple  family

                 Commercial
                  • PC.OMM1= Buildings with <50,000 ft  GLA*
                  • PCOMM2= Buildings with 50-100,000 ft2 GLA
                  • PCOMM3= Buildings with >100,000 ft2 GLA

Although it would have been possible to develop predictive equations directly
for  the land use  subcategories, it is advantageous to predict total, aggre-
gate  residential  and commercial land use and then disaggregate the totals.
There are  two reasons for such an approach.  First, results of the GEMLUP-I
study indicate  that average errors associated with aggregated predictive
equations  tend  to be smaller.  Second, by not directly incorporating the
disaggregation  of residential and commercial land uses into density classi-
fications  in the  predictive equations, we allow the user of the predictive
equations  to enter in a  land  use density distribution applicable to some
future  time period of interest, different from that of 1960-1970.  This
 *
  Gross  Leasable  Area
                                    2-2

-------
second concern is motivated by the  shift in density of  new housing  construction
in the past few years away from single family detached  developments,  which
predominated during the 1960s.  Thus,  we believe that this approach ensuired  a
more generalized, predictive model.

             The next step was to decide which independent variables to
include in the predictive equations.   A maximum of 6 independent variables were
included in each equation to keep the  predictive model  simple and easy-to-use,
to avoid possible degrees of freedom problems, and to keep confidence intervals
small.  The set of potential independent variables included all  exogenous  and
instrumental variables*from the initial causal model.  The choice of independent
variables was dictated by the stepwise regression analysis, i.e., at each  step
the analysis chose from the set of all independent variables the one which
explained the most additional variance in the dependent variable.  Since the
analysis can produce several different predictive equation forms depending  upon
the independent variables included in  the analysis, and their priority, a  set  of
statistical criteria were developed for application to  each variable before  its
final acceptance in the model equation.  These criteria were the following:

                 • A test to ensure an independent variable was  significant,
                   e.g., a minimum F statistic value of 1.7, corresponding  to
                   the 10% significance level.**
                 • A test to ensure the choice of the model form which explains
                   the greatest amount of variance in the dependent variable,
                   viz. the requirement that the chosen final form have the
                   highest adjusted R2 value.
                 • A test to ensure the absence of multi col linearity, viz.
                   3 < 1.0.

              The stepwise  regression  analysis  was  applied,  as described
 above,  using data  from all  40 case studies  and programs  from the SPSS  soft-
 ware  package [13].   It was  decided that it  was preferable  to use the  full
 set of 40 samples  in developing  the equations since  this  would  produce  the
 *  See Appendix B for a definition of terms.   See Appendix C for a
    definition of model  variables,  in both English and metric units.
 **  That is,  there  was at most  a  10 percent  chance  of  accepting  a  variable
    as  significant  in  the regression when  it actually  was  not.

                                    2-3

-------
most valid, generalized model.  All variables were used in unstandardized
form.  The results were predictive, linear equations in both English and
metric units, discussed in the next section, with full statistical output
summarized in Appendix A.

         2.  Discussion of Results

             a.  Land Use Equations

                 The land use predictive equations obtained by applying the
previously discussed objective criteria to the stepwise regression analyses
are  summarized below.  Summary statistics for these equations are shown in
              2
Table 2-la.  R  values indicate the predictive equations are explaining
the  majority of variance in the dependent variables,  i.e., the mean value
for  this statistic was 0.67.  The overall F  statistics indicate all of the
predictive equations are significant at or below the  one percent  level.  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.  The
mean of the coefficients of variation for the land use equations  is 0.63,
indicating that the average error encountered in the  use of these predictive
equations will be ±63 percent.  These errors are reasonable for a generalized
tool, applicable nationwide, and are less than those  introduced subsequently
in the impact assessment procedure [16].

                 The predictive equations shown below constitute a set of
equations applicable to an area of analysis where a wastewater major project
of a certain size range will be or already has been built.  They do not
constitute a general land use predictive model.

                   A major project is defined as the construction or extension
                   of interceptor or collector sewer lines in a community in
                   the United States that affected an increase in absolute
                   system collection capacity of 1.0 million gallons per day
                   (MGD) or more.  Based on  the distribution of data used in
                   developing the model, the increase in collection capacity
                   should not exceed 100 MGD.
                   The area of analysis is defined as the legal service area
                   of the major project in the base year, and it must contain
                   a significant amount of vacant, developable land.  Based on
                   the distribution of data  used in developing, the model,
                   the size of the area of analysis should be 5,000 to 75,000
                   acres.

                                 2-4

-------
                       TABLE 2-1 a
SUMMARY STATISTICS OF THE LAND USE PREDICTIVE EQUATIONS
Dependent
Variable
RES
COMM
OFFICE
MANF
WHOLE
HI WAYS
EDUC
REC
OTHER


Dependent
Variable
RP1
RP2
CP1
CP2
Number of
Predictors
6
6
6
6
6
5
6
6
6
TABLE
SUMMARY STATISTICS OF THE
Number of
Predictors
6
5
5
5
R2
0.74
0.57
0.70
0.63
0.76
0.52
0.75
0.65
0.67
2-1 b
2 Coefficient
R a of Variation
0.69
0.50
0.65
0.56
0.72
0.45
0.71
0.58
0.61

0.38
0.81
0.59
0.89
0.67
0.53
0.36
0.59
0.88

DISAGGREGATION EQUATIONS
R2
0.82
0.84
0-.73
0.65
R2
R a
0.76
0.80
0.69
0.60





                           2-5

-------
The use of these equations should be limited to situations in which the major

project and areas of analysis are in these size ranges.  Also, any application

of the predictive equations should be qualified by the error range indicated

by the coefficients of variation shown in Table 2^1.


                 (1) English Units.


RES    = - 8,692 VACANT + 5,347 LAND + 16.42 RECAP1 +  11.24 MANJOB - 18,804 STAY
         + 487.2 DRIVE + 13,466

COMM   = 4,302 SERVED - 4,515 VACANT - 28.84 KIDS + 6,249 ACCESS - 17.04 JOBCHG
         - 2.626 RECAP! + 5,376

OFFICE = 719.5 RRMILE + 33,478 OZONED + 36.70 PEAK + 0.3334 DUACRE - 1,791
         IZONED + 572.0 POPDIF - 459.9

MANF   = 2,591 RRMILE - 3,890 VACANT + 2.978 MANJOB +  85,923 OZONED + 234.7
         VACOFF +22.7 AIRPRT + 611.2

WHOLE  = 166.5 DRIVE + 86.28 VACOFF + 3.340 OFFJOB + 6,043 EMPOP + 6,949 UNEMP
         - 0.1065 PERCAP -1,911

HIWAYS = 35.22 RRMILE + 22.98 LAND - 0.1167 GOVT + 0.5379 RECAP + 25.94 CROSS
         + 11.35

EDUC   = -3,186 POOR + 723.0 SERVED + 437.5 LAND + 95.17 LIMITS - 405.2
         PERCHG - 1.415 COST + 428.9

REC    = 270.4 RRMILE + 2.631 CAPAC2 + 322.1 PERCHG +  0.9742 OFFJOB   3,226
         VACHSE - 557.8 IZONED + 59.52

OTHER  = 1,134 RRMILE - 0.3020 PERCAP - 12,175 VACHSE  + 649.0 PERCHG -
         295.4 INTDEN + 403.8 RZONED + 3,087

                 (2) Metric Units.   The predictive model equations in this

section are reported in metric units.   The conversion of units involved only

the dimension of distance.  The following conversion factors were used:


                     mile2 -»• km2,    multiply by 2.589

                     1,000 ft2 -> m2, multiply by 92.90
                              2
                     acres -> m ,     multiply by 4,047

                     miles -* km,     divide by 1.609
                         -1    -?
                     acre   + m  ,    divide by 4,047

                     mile"  ->• km  ,  divide by 1.609
                                  2-6

-------
RES    = -2 148 VACANT +  1.321  LAND  +  0.0041  RECAP1  +  0.0072  MANJOB  -  4.646
         STAY + 0.3117 DRIVE  +  3.327

COMM   = 987.5 SERVED   1,036 VACANT - 6.621  KIDS  +  3,714  ACCESS  -  10.13
         JOBCHG -  0.6029  RECAP1  +  1,234

OFFICE = 102.7 RRMILE + 7,685 OZONED + 8.424  PEAK  +  0.1981  DUACRE -  411.0
         IZONED +  131.3 POPDIF - 105.6

MANF   = 369.6 RRMILE - 893.0 VACANT + 1.770  MANJOB  +  19,724  OZONED  +  53.87
         VACOFF +  3.239 AIRPRT + 140.3

WHOLE  = 98.23 DRIVE + 19.80 VACOFF + 1.985 OFFJOB + 1,387 EMPOP  + 1,595
         UNEMP - 0.0244 PERCAP - 438.7

HIWAYS = 0.0087 RRMILE + 0.0091 LAND - 0.00005 GOVT  +  0.0002  RECAP + 25.94
         CROSS + 0.0045

EDUC   = -731.3 POOR + 170.0 SERVED + 100.4 LAND + 21.85 LIMITS - 93.02
         PERCHG - 0.3248 COST + 98.45

REC    = 168.1 RRMILE + 2.631 CAPAC2 + 322.1  PERCHG +  2.522 OFFJOB - 3,226
         VACHSE - 557.8 IZONED + 122.9

OTHER - 161.8 RRMILE - 0.0693 PERCAP - 2,795 VACHSE + 150.0 PERCHG - 67.81
         INTDEN + 92.70 RZONED + 708.5


              b.  Disaggregation Equations

                 Predictive equations for disaggregating residential and
commercial land uses  into subcategories were developed using logit analysis
[14].  This approach  fits the percentage data (P)  to a logistic S curve.
The result is a non-linear structural equation that relates the range 0 to 1
with  a set of independent variables (Xn.Xp,	X  ),  viz:


                              (b1x1+b9x9+	+b x  )  ,
                P   =   (1 + me   ] ]  2 2       n n T1             (1)
P  is a general term for the previously defined percentage variables SFDET,
SFATT, MF, PCOMM1, PCOMM2, and PCOMM3.  This approach avoids two problems
which would arise if linear, stepwise regression analysis was used, namely
(1) the value for P would not be constrained between 0 and 1 (i.e., 0% and
100%), and (2) the sum of the percentages predicted for all subcategories
of a given land use type would not sum to 1 (i.e., 100%).
                                   2-7

-------
                 To translate equation (1) to a form that can be solved
with linear, stepwise regression, the following transformation was made:

                 In (P/l-P) = ln(m) -Vrb2x2~' ' • ''Vn       (2)

This is equivalent to:

                 Y = a0 + alXl + a2x2+....+anxn                (3)


             where: m  =  e
                     bn  '  -an

                  To  constrain  the  total  of  predicted  subcategory percentages
 to  1.0,  it was  necessary  to apply  logit  analysis  in an  iterative manner.
 First,  the percent of single family  detached  and  large  commercial develop-
 ments were predicted.   Then the  residual  percentages  were  predicted.  Thus,

      Residential
      %  Single  family detached  =  RP1* =  SFDET
      %  Multiple family  =  RP2  (1-RP1), where RP2 = MF/(SFATT + MF)
      %  Single  family attached  =  1-RP2 -RP1

      Commercial
      	   r)    •fc^f
      % >100,000 f1T GLA   = CP1  = COMM3
      % 50-100,000 ft2 GLA = CP2  (1-CP1),  where CP2 = PCOMM2/(PCOMM1+PCOMM2)
      % <50,000 ft2 GLA = 1-CP2 - CP1

                       p
                  The R  statistics for the logit analysis,  shown in Table  4-lb,

 indicate excellent predictive ability for the disaggregation equations,  with

 value ranging from 0.65 to 0.84, and  averaging 0.76.  The complete predictive

 equations for RP1, RP2, CP1, and CP2  are  summarized below.
 *
   The statistical  output listed in Appendix A refers to RY1  which is simply
   the transform of RP1 used to solve for the coefficient values,  i.e.,  RY1  =
   (RP1/1-RP1).  This applies to RP2, CP1 and CP2 in a similar manner.
 **
   Gross Leasable Area.
                                     2-8

-------
                 (1)  English  Units.
                 RP1  = (1  + 92.8EXP?- 0.0137  KIDS + 0.000918 CAPCHG -  0.784  SERVED
                       - 22.7 OZONED + 2.98 STAY + 0.181  HOSCHG)H
                 RP2  = (1  + 1.24  EXP(-  4.38"HSECHG + 5.48 POOR  +  34.1  OZONED
                       - 0.732 RRMILE - 0.0365  PEAK))-!
                 CP1  = (1  + 0.104 EXP(0.215 VACOFF - 0.507 LIMITS +  0.0232
                       CAPAC2 - 0.0159  KIDS + 0.0603 TLIMIT))-'
                 CP2  = (1  + 0.585 EXP(1.54  PHASE + 0.00459 TRANS  + 0.0178
                       DISCED + 0.00417 GOVT  -  1.48 VACANT)H
                 (2)  Metric Units.
                     The disaggregation equations in metric  units are  identical
to those for English  units with two  exceptions:   (1) the  coefficient for  RRMILE
in RP2 changes to -0.454,  and (2) the coefficient for DISCED in CP2  changes  to 0.0111.
     B.   MODEL VALIDATION
         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  a  fairly small
data sample  (40) was available for developing  predictive land use  equations,
the question of validity was  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.  In addition, a  more precise weight validity index was
also computed.
         1.   Cross-Validation
             Cross-validation provides  a measure of the overall validity  of
the predictive equations.   The procedure involved splitting the case study
data sample  of 40 into two random samples of  20 each. The first  sample of
20 consisted of case  numbers* 1,2,3,5,6,7,10,12,13,14,18,21,24,26,27,29,30,
32,38, and 40; the second sample  of  20 consisted of the  remaining case num-
bers.  The coefficient values in  the predictive equations were  recomputed
using the first data  sample of 20.  These  coefficient values, specifying
*
  See Table 3-3 in Volume I [12] for a definition of case study numbers.
**                                                      Y
  EXP is the exponential function (e).  Thus, EXP(x) = e .
                                     2-9

-------
 9  land  use  predictive equations,  were  used  in  conjunction with data from the
 second  sample of 20 to predict the  values of the  dependent  variables  in the
 second  sample.   Statistical  comparisons  were then made  between actual  and
 predicted values for the  dependent  variables in  the  second  sample.

              The correlation coefficients  (R)  between actual  and  predicted
 values  are shown in Table 2-2, and  graphs of the  actual  versus fitted
 (predicted) values are summarized in Appendix  D.   In this application, the  R
 statistic represents a coefficient  of  validity of the predictive  ability of
 each equation.   Values in Table 2-2 range  from -0.04 to 0.68'.  All  R  values
 but one are statistically significant  at the 2%  level or better.   The ex-
 ception is the OTHER equation, where R is  not  significantly different from
 zero.  These results can  be interpreted  that 8 of the 9 equations are generalized
 enough  to produce good predictions  using an independent sample.   Only the OTHER
 equation  may not be well generalized.  The poor performance of  this  equation
 could,  however, be attributed to non-homogeneties in the data sample.s caused
 by the  small sample size.

          2.  Weight Validity  Index

             The weight validity index is a new statistical  technique  [15]
which essentially performs a validation without the need for a separate
independent sample or having to split the original sample, as in  cross-
validation.  Since the entire  data sample of 40 is used  in the test, it gives
more precise estimates of validity for the equations being tested.

             The procedure involved first obtaining the  R statistic for an
equation estimated on the full sample of 40 (see  Table 2-la).  Next, the
quantity p is computed as:
                    f                           I  1/2
             P =     1 -  (N-4)(l-R2)("l  + 20-R2)^
                            N-n-1   \     N-n-1 /J

                   where:   N = sample size = 40
                           n = number of independent variables (see Table  2-la)
                               /x
Then the weight validity index p  is estimated  by:
                  /^      /\
                  Pc = 2 p- R
                                       2-10

-------
                      TABLE  2-2
     COEFFICIENTS OF VALIDITY  BETWEEN ACTUAL  AND
PREDICTED LAND USE FROM THE  CROSS-VALIDATION  ANALYSIS
Dependent
Variable
RES
COMM
OFFICE
MANF
HI WAYS
EDUC
REC
WHOLE
OTHER
Coefficient of
Validity (R)
0.64
0.56
0.46
0.39
0.42
0.68
0.68
0.63
-0.04
Statistical
Significance Level
1%
1%
1%
2%
1%
1%
1%
1%
— ^









                           2-n

-------
The values for pc are summarized in Table 2-3.  To interpret the results the
values of R and pc are compared for a given equation.  If there is a signifi-
cant drop from R to pr, then the equations are not well generalized.  In
                     U   /v
the current application, p  is only 3-9% less than R.  This small drop in
value indicates that aV\_ the predictive equations can be used confidently
on a new sample and that they are all well generalized for predictive appli-
cations.

     C.  COEFFICIENT STABILITY ANALYSIS

         To test the stability of the coefficient values in the land use
predictive equations, a technique was devised which is very sensitive to
instabilities.  The analysis simulated a "worst case" in terms of instabili-
ties, and thus is more stringent than the jackknifing approach used in the
causal analysis  [12].  The procedure involved reestimating the model coef-
ficients when excluding a cluster of 2 or 3 samples that had the greatest
potential for instabilities.  For each equation, the independent variables
with the two largest 3 weights were identified and, together with the de-
pendent variable, examined for skewness in their distributions.  Where extreme
values occurred, the case study number was noted.  All such numbers were
combined to see  if 2 or 3 case studies dominated.  These were then  the data
samples excluded from  the stability analysis  regressions (see Table  2-4).
The percentage change  in coefficient values and the standard error  of the
          *
regression  were then  computed.  These values are listed in Table 2-4 as well.
Note that the excluded values may or may not  be associated with large residuals,
In fact, many times a  small residual on an extreme value indicates  the ex-
treme value has  a significant impact on the regression line and thus more
potential for instabilities.  Because of this, residuals were ignored in
choosing the case studies to exclude.

         The results in Table 2-4 reveal significant changes in the model
coefficient values.  The range in percent change is 1% to 111%, with an
average of 38% overall.  However, since there is a mixture of both  + and -
signs in the percentages, there may be compensating effects, i.e.,  the final
 *
 On  a  common  degrees  of  freedom  basis.
                                      2-12

-------
                                 TABLE 2-3

                 WEIGHT VALIDITY INDICES FOR THE LAND USE
                             PREDICTIVE EQUATIONS
Dependent
Variable
RES
COMM
OFFICE
MANF
WHOLE
HIWAYS
EDUC
REC
OTHER
Weight Validity
Index
0.83
0.69
0.79
0.74
0.84
0.66
0.83
0.75
0.77
Regression
R Value*
0.86
0.75
0.84
0.79
0.87
0.72
0.87
0.81
0.82
Coefficient
of Validity**
0.64
0.56
0.46
0.39
0.42
0.68
0.68
0.63
-0.04
**
 From Table  2-1 a.

r
 From Table  2-2.
                                    2-13

-------
                               TABLE 2-4

              SUMMARY  STATISTICS OF THE PREDICTIVE EQUATIONS
                      COEFFICIENT STABILITY ANALYSIS
Dependent Case Studies
Variable Excluded
RES 7,11,13





COMM 18,35





OFFICE 3,18,35





MANF 18,35





Percent Change in Independent
Standard Error of Variables
Regression
- 9.9% VACANT
LAND
RE CAP 1
MANJOB
STAY
DRIVE
-36% SERVED
VACANT
KIDS
ACCESS
JOBCHG
RE C API
-15% RRMILE
OZONED
PEAK
DUACRE
IZONED
POPDIF
-20% RRMILE
VACANT
MANJOB
OZONED
VACOFF
AIRPRT
Percent Change in Model
Coefficient Values
+ 9
-11
-14
-47
+61
-73
-60
+59
+83
-86
*
+81
-54
-55
-34
-13
+72
-20
-64
+37
- 1
-46
-55
-81
*Variable not significant in regression excluding skewed  cases
                                      2-14

-------
            TABLE 2-4 (CONTINUED)

SUMMARY STATISTICS OF THE PREDICTIVE EQUATIONS
        COEFFICIENT STABILITY ANALYSIS
Dependent Case Studies Percent Change in
Variable Excluded Standard Error of
Regression
HIWAYS 6,13,18 - 2.9%




EDUC 3,13 - 3.3%





REC 11,18,27 - 5.6%





WHOLE 7,14 - 9.5%





Independent
Variables
RRMILE
LAND
GOVT
RECAP
CROSS
POOR
SERVED
LAND
LIMITS
PERCHG
COST
RRMILE
CAPAC2
PERCHG
OFFJOB
VACHSE
I ZONED
DRIVE
VACOFF
OFFJOB
EMPOP
UNEMP
RERCAP
Percent Change in Model
Coefficient Values
-81
-20
+28
-16
-49
+ 7
- 4
- 5
- 1
- 4
-Hll
-18
+15
-54
-12
+28
+ 35
-25
-13
-19
-31
-69
+25
                           2-15

-------
                          TABLE 2-4 (CONTINUED)
              SUMMARY STATISTICS OF THE PREDICTIVE EQUATIONS
                      COEFFICIENT STABILITY ANALYSIS
Dependent  Case Studies  Percent Change in  Independent  Percent Change in Model
Variable     Excluded    Standard Error of   Variables     Coefficient Values
                             Regression

OTHER        12,18            -44%            RRMILE              -73
                                              PERCAP              -68
                                              VACHSE              -81
                                              PERCHG              -75
                                              INTDEN              -79
                                              RZONED              +40
                                       2-16

-------
predicted values may or may not be  changing.   The  standard  error of  the  re-
gression is an indicator of the precision of  the predicted  values.   Thus,
changes in the standard error,  when excluding certain samples,  provide  infor-
mation about the stability of an equation.  The percent changes in  the  standard
error listed in Table 2-4 range from - 2.9% to -44%.   Note  that all  values  are
negative, indicating that, as would be expected, the  regression does-a  better
job on a narrower range of data (i.e., that which  excludes  hvghly skewed
samples.) For 7 of the 9 equations, the percent change under worst  case
conditions is -20%.   This indicates that no significant instabilities occur
in these equations.   The other two  equations, COMM and OTHER, have  percent
changes of 36% and 44%, respectively.   This indicates in both cases  that an
equation different from the true predictive equation  does significantly better
on a narrower range of data.  Thus, the excluded points are causing  instabili-
ties in the COMM and OTHER equations.   What these  worst case instabilities
represent are the fact that the functional form of the land use relationships
change with the magnitude of the data.  Thus, a single generalized  equation
(i.e., that used in the GEMLUP-II model) cannot be as precise as one developed
for only a portion of the range of data.

     D.  EMISSION PROJECTION WORKSHEETS

         The predictive land use equations discussed in the last section, serve
as the input to a two-part air quality assessment  procedure outlined in
Figure 2-1.  First, the predictions of future land use in the area  of analysis
of a major project are translated into stationary  source emissions  (for all
criteria pollutants) using a set of land use based emission factors [10].
Second, the revised VMT model  [11,12] estimates vehicle miles traveled, and
hence mobile source emissions, generated by transportation  actively servicing
the land use types.  These two emission components are then summed to give
the user the total air pollutant emissions associated with  the induced develop-
ment from a wastewater major project.

         This .impact assessment procedure has been formalized in a set of
worksheets, presented in this  section.  These worksheets serve as an opera-
tional tool that can be easily used by environmental  engineers or planners
to assess the secondary air quality impacts associated with new or expanded
wastewater facilities.
                                       2-17

-------
 MAJOR PROJECT
CHARACTERISTICS
BASE YEAR
CONDITION
   PROJECTED
REGIONAL GROWTH
                          PREDICTIVE
                           EQUATIONS
                           PROJECTED
                          FUTURE LAND
                              USE
LAND USE BASED
   EMISSION
    FACTORS
                         TRAFFIC
                          MODEL
  STATIONARY
    SOURCE
   EMISSIONS
                         MOBILE    \
                         SOURCE     ]
                        EMISSIONS   /
                           TOTAL AIR
                           POLLUTANT
                           EMISSIONS
                          FIGURE 2-1

           OVERVIEW OF THE LAND USE AND AIR QUALITY
                 IMPACT ASSESSMENT PROCEDURE
                             2-18

-------
         1.   Definitions

             The predictive equations  used  in  the  worksheets  constitute  a
set of equations applicable to  an area of analysis where  a  wastewater  major
project of  a certain size range will  be,  or already has been,  built.   They
do not constitute a general  land use  model.  They  produce projections  of
total  land  use in the area of analysis ten  years after the  initiation  of
the major project.   In this regard,  the following  definitions  apply:

             A major project is defined as  the construction or extension
             of interceptor or  collector sewer lines  in a community  in
             the United States  that  affected an increase  in absolute
             system collection  capacity of  between 1.0 and  100.0  million
             gallons per day.
             The area of analysis is  defined as the legal service area
             of the major project in  the base  year, and it  must contain
             a significant amount of  vacant, developable  land.  It should
             be between 5,000 and 75,000 areas in  area.
             The year t,  or year of  initiation, corresponds to the year
             the major project's new  or expanded collection system first
             carried wastewater flows.  If  completion of  the  project was
             (or is to be) phased, it should be completed before  the
             year t + 5.

The use of the worksheets should be  limited to situations in  which the major
project and areas of analysis are in  the size  and  time ranges  noted  above.

         2.   Input Data Requirements

             The data required  for application of  the worksheets  are  sum-
marized in Table 2-5.  Here the variable names used in the  worksheets  are
defined (in English units) and  data  sources are listed.   The  data required
are generally available from regional  planning agencies and the facility
plans, and the total data collection task will require approximately two
to three mandays of effort.  Appendix C contains corresponding metric  units
for each model variable,  as well as  a cross-index  relating  worksheet variable
names and predictive equation variable names.   Photocopies  of worksheet  1
(English units, starting on page 2-43) or 1M (metric units, starting on
page 2-60) should be used to record  actual  values  when collecting data.
These are contained in Sections 4 and 5, respectively, along  with all  other
worksheets.
                                        2-19

-------
                                  TABLE 2-5

             INPUT DATA REQUIREMENTS FOR IMPACT ASSESSMENT MODEL
Variable Name
Description (English Units1)
                                                             Data Source
County Area
Vacant Developable

Vacant Undevelopable
Zoned Residential
Zoned Office
Zoned Industrial
Onsite Restrictions
                                      County-City Data
                                         Book
                                       Planning Agency


                                       Planning Agency



                                       Planning Agency



                                       Planning Agency



                                       Planning Agency
                                       Planning Agency
                                       or Local
                                       Government
                       Area  of  county  in  square miles


                       Vacant developable acreage  expected
                       in  area  of analysis in year t2

                       Vacant undevelopable acreage ex-
                       pected in  area  of  analysis  in
                       year  t

                       Acres of land expected to be zoned
                       for residential  use in area of
                       analysis in year t
                       Acres of land expected to be zoned
                       for office use  in  area of analysis
                       in  year  t

                       Acres of land expected to be zoned
                       for industrial  use in area  of
                       analysis in year t

                       Categorical variable to  indicate  the
                       severity of existing or  expected
                       governmental restrictions on on-lot
                       sewage disposal  during the  years  t
                       to  t+10.  Coded3 as follows:
                        4 = on-lot disposal prohibited
                        3 = prohibited except  on  large
                             lots
                        2 = permitted but percolation test
                             required
                         1 = permitted but package plants
                             prohibited
                         0 = no restrictions

                       The number of years between year  t
                       and year t+10 that it is  expected
                       that  any on-site sewage  disposal
                       restrictions will  be in  effect
                       (i.e.,  a value  of  0 to 10)
                       Number  of limited  access  interchanges   Planning  Agency
                       expected in county in year  t+5

                       Number  of limited  access  interchanges   Planning  Agency
                       expected in the area of  analysis  in
                       year  t+5

I3ee Table C-3 in Appendix C for corresponding  metric  units.

2t is the year of major project initiation.

3When restrictions are a mix of the five  codes  given,  use the most restrictive
 condition (i.e., the highest  value).
Restriction Years
                                       Planning Agency
                                       or Local
                                       Government
County  Interchanges
Limited Access
                                   2-20

-------
                            TABLE 2-5 (CONTINUED)

             INPUT DATA REQUIREMENTS FOR IMPACT ASSESSMENT
Variable Name
Description (English Units)
Data Source
Transit Stops
County Growth
Future Population
Population Growth


Office Vacancy



Future Houses


Median Price
Future Income


Future Employment



Future Medicals


CBD Distance




Airport Distance



Track
Number of transit stops (bus and
commuter rail) expected in area of
analysis in year t

Percent1 change in county population
projected for the years t to t+10

Projected SMSA2 population in year
t+10


Percent1 change in SMSA2 population
projected between years t and t+10

Percent3 of office buildings in area
of analysis in year t that are ex-
pected to be vacant

Projected housing units in SMSA2 in
year t+10

Projected median price of 1 acre of
vacant residential land ($) in area
of analysis in year t
Projected median family income in
SMSA2 in year t+10

Projected SMSA2 employment in year
t+10


Projected hospital employment in
SMSA2 in year t+10

Distance in miles'* from centroid of
area of analysis to centroid of
nearest central business district
in year t

Miles from centroid of area of
analysis to centroid of nearest
commercial airport in year t

Miles of railroad track in area of
analysis in year t
Planning Agency
Planning Agency


Planning Agency
or BEA Projec-
tions [28]

Planning Agency
or OBERS

Planning Agency
or Realtor
Planning Agency
or OBERS

Planning Agency
or Realtor
Planning Agency
or OBERS
Planning Agency
or OBERS Pro-
jections [18]

Planning Agency


Highway Road Map
Highway Road Map
USGS Topo-
graphical Map
:A value of 10% is coded as 0.1.

2If the area of analysis is not in an SMSA,  use county data.
 instances two different variables may end up with the same
3A value of 10% is coded as 10.

4Air miles, not road miles.  Central  business district used is that in the
                                       In these
                                     value.
 nearest city with a population of 100,000 or more.
                                     2-21

-------
                            TABLE 2-5 (CONTINUED)

             INPUT DATA REQUIREMENTS FOR IMPACT ASSESSMENT MODEL
Variable Name
Description (English Units)
Data Source
Area of Analysis
Sewered Land
Interceptors
Collection Capacity
Peak Flow
Treatment Capacity
Population  Served
 Project  Cost
 Federal  Funds
 Phasing
 SMSA Area
Area of analysis in acres
Acres of land within 5,000 ft of
the major project interceptor sewer
in the area of analysis in year t
Running length of interceptor sewer
lines in miles going through rela-
tively undeveloped land1 in area of
analysis in year t
Total hydraulic design capacity of
wastewater major project collection
system in million gallons per day
(mgd) in year t2
Anticipated peak flow in the waste-
water major project collection
system in mgd in year t
Total hydraulic design capacity of
the major project wastewater treat-
ment plant in mgd in year t
Population served by the major
project facility in year t
Total major project construction
cost in thousands of $
Federally funded share of major
project cost  in thousands of $
Categorical variable to indicate
whether the completion of the col-
lection network will be phased over
several years
   1 = phasing will occur
   0 = no phasing
Area of SMSA3 in square miles
Facility Plan or
Water Resources
Plan
Facility Plan


Facility Plan



Facility Plan



Facility Plan


Facility Plan


Facility Plan

Facility Plan

Facility Plan

Facility Plan
County-City Data Bk.
  See Appendix E for directions  on how to  gather  these  data.

 ^ess than  one dwelling unit per acre.

 2Year t or  up to 5  years after  year t if  it  is a phased  project.   Note  this
  model  should not be applied to major projects which are  not  completed by year
  t+5 (see Definitons on page 2-19).

 3If the area of analysis is not in an SMSA,  use  county data.   In  these in-
  stances, two different variables may end up with  the  same  value.
                                    2-22

-------
                            TABLE 2-5 (CONTINUED)
             INPUT DATA REQUIREMENTS FOR IMPACT ASSESSMENT MODEL
Variable Name
Description (English Units)
 Data Source
Tract Area
Dwelling Units

Current Houses

School Kids

Vacant Houses

Nonmobility

Median Income

Current Income

Poverty

Index One

Index Two

Government

Current Employment
Unemployment
Area of census tracts1 in sq.  miles
Census tract housing units in 100s
in year t (round off to nearest
100 units)
Total housing units in SMSA2 in
year t
Population 0-14 years of age in
census tracts in year t
Percent3 vacant available dwelling
units in
Percent3 of families in year t who
were in the same house in year (t-5)
Median income of familes ($) in
county** in year t
Median family income ($) in SMSA2 in
year t
Percent3 of total families with
income below the poverty level in

Consumer Price Index5 for the
year t
Consumer Price Index5 for the year
of federal funding
Total county expenditures in
millions of $ in year t
Total SMSA2 employment in year t
Percent3 unemployment in
 Census  Tracts  Map
 U.S.  Census

 U.S.  Census

 U.S.  Census

 U.S.  Census

 U.S.  Census

 County-City Data
 Book
 U.S.  Census or
 OBERS
 U.S.  Census

 (See footnote 5)

 (See footnote 5)

County Government

U.S. Census or OBERS
U.S. Census
 See Appendix E for directions on how to gather these data.
Census tracts which most closely approximate the area of analysis.   If area
 is untracted, use data for the municipality.  This applies  to all  tracted
 census variables.
2If the area of analysis is not in an SMSA,  use county data.   In these in-
 stances, two different variables may end up with the same value.
3A value of 10% is-coded as 0.1.
^County containing most of the area of analysis.
5Use U.S. Department of Labor statistics for the nearest major city, based
 on 1947-49 average prices being =100.0.  If t or the year  of federal
 funding are not the current year, use consumer price index  data for the
 current year, adjusted by the expected annual inflation rate.
                                   2-23

-------
                            TABLE 2-5 (CONTINUED)

             INPUT DATA REQUIREMENTS FOR IMPACT ASSESSMENT MODEL*
Variable Name
Description (English Units)
Data Source
Office Workers
Manufacturing
Workers

Current Medicals
Drivers
Office employment in census tracts
in year t

Manufacturing employment in census
tracts in year t

Hospital Employment in SMSA1 in
year t
100s of workers who drive to work
in year t in the county (round off
to nearest 100 workers)
U'.S. Census


U.S. Census


U.S. Census or
OBERS

U.S. Census
 See Appendix E for directions on how to gather these data.

 If the area of analysis is not in an SMSA, use county data.  In these in-
 stances, two different variables may end up with the same value.
                                     2-24

-------
         3.  Instructions

             First, decide whether you will  be using English or metric units, and
then photocopy a set of the appropriate set  of worksheets.   Next, collect all the
required data (as described in the previous  section), enter the values on Worksheet
1*, perform the indicated operations, and record the answers in the boxes labeled
"Inputs to Worksheets."  Finally, fill in the photocopies of Worksheets 2 through
13 using the instructions below.

             a.  Projection of Future Land Use

                 Worksheet 2 computes the values of the 9 land use variables.  Fill
in all the blanks on the right-hand side of  the equations with the appropriate data
from Worksheet 1.  Perform the arithmetic operations and record the results in the
blanks on the left-hand side of the equations.

                 Worksheet 3 computes total  land use in the area of analysis.  First,
copy the land use projection data from Worksheet 2 boxes into column (1).  Next,
enter the projected percentages for disaggregation of residential and commercial
land use in column (2).  These values must be between 0.0 and 1.0, and the three
percentages for residential and commercial uses must each sum to 1.0.  The per-
centages should take account of local factors which will affect the density of
development in the area of analysis.  In the absence of local data, default values
can be estimated using the equations in Worksheet 4 (the input data for Worksheet
4 all come from boxes in Worksheet 1).  Next, divide the Area of Analysis (first
box on Worksheet 1) by 10,000 and enter the  value in column (3).  Final land use
projections (4) are now obtained  by taking the product of columns (1), (2) and (3).

                 Worksheet 5 is another optional worksheet which can be used to
estimate 90% confidence intervals for any one of the land use projections in Work-
sheet 3.  First, enter the name of the land  use category at the top of the worksheet.
Next, under Predictor Variable Name, list the name of the 5 or 6 predictors from
the appropriate equation in Worksheet 2.  Next, enter the values for these predictors
from Worksheet 2 in the spaces to the right  of the names.  If only 5 predictors are
used, enter 0.0 on line 6.  The co-variance  data required is summarized in the sta-
tistical output in Appendix A, and Page A-l  gives you an index to where to look for
these data.  In order to read the statistical output, it is first necessary to
translate predictor names into the acronyms  used in the model, and Table C-3 pro-
vides a cross-index to do this.  Record the  covariance data on Worksheet 5 and per-
form the indicated arithmetic operations. An example of this procedure is shown on
page 2-83 and discussed below.  In the example on Page 2-83, a confidence interval
is being estimated for the "Single Family Detached" land use category so this is
recorded at the top of the Worksheet.  Since this category evolved from the "Resi-
dential" equation on Worksheet 2  (Page 2-80), we list the names and values of the
6 predictors from the "Residential" equation in the spaces at the top of Worksheet
5.  Then we go to Table C-3 (pp.  C-16 thru C-18), find the acronyms for the pre-
dictor names, and record these at the top of Worksheet 5 as well.  Next, we go to
Page A-l and in the index find that page  A-2 contains the covariance data we
need.  At the bottom of Page A-2  is a matrix from which we extract values corres-
ponding to a pair of variables.  For example, the first covariance value needed in
Worksheet 5 is for ."(D  x CD", that is the  covariance of variable (T) (which is
VACANT) with itself.  From the matrix we find "0.716E+07" at the intersection of
VACANT and VACANT which is then copied onto  Worksheet 5, and so forth.
*The discussion in this section will  use worksheets for English units.  The in-
 structions for metric worksheets are identical, except each worksheet number
 has a suffix M, i.e., 1  becomes 1M,  etc.

                                         2-25

-------
             b.  Calculation of Motor Vehicle Traffic

                 The worksheets presented in this and following sections are based
on the revised GEMLUP-I traffic model [11,12].  Due to its simplistic approach,
the appropriateness of transportation parameters used in the model is extremely
important.  Thus, although default values are provided for most parameters, it is
considered desirable to use local sources of transportation data whenever possible.

                 Worksheet 6 is used to compute total vehicle trips in the area of
analysis for two trip purposes - Work and Other.  First, compute a value for the
Effective Radius by completing the calculation shown at the top of Worksheet 6.
This definition uses a value based on a circle with an area equivalent to that
of the actual area of analysis.  That is,

                 Effective Radius =  ((Area of Analysis/640)/Tt)1/2

Next, enter the values for Projected Land Use in column (1).  Enter values for Work
and Other trip generation rates in columns (2) and (4), respectively.  If local
data are not available, default values are given in Table 2-6.  Compute Work trips
by multiplying columns (1) and (2);  compute Other trips by multiplying columns
(1) and (4).  Compute Residential Work Trips and Residential Other Trips by sum-
ming the first three rows of columns (3) and (5), respectively.  Compute total
Work Trips and Total Other Trips by  summing all items in columns  (3) and (5).

                 Worksheet 7 is used to compute vehicle miles traveled (VMT) in
the Area of Analysis and in the "Impact Area".  VMT in the Impact Area are defined
to include both the VMT within the Area of Analysis, as well as VMT outside the
Area of Analysis but occurring because of the presence of the major project and
its induced land uses.  First, enter values for the Work and Other trip lengths.
If local data are not available, use the following default equations:

Work  Trip  Length   =   0.00447*  p°'22  S^'49                    (in miles*)

                                                          fl
                                }0.18 s  1.40 +  p0.26  s  1.251    ^-n m11esj

                                                          J
          where:   p   = Future  Population

                  S-,  = Average  network  vehicle  speed  for work
                       trips  in miles per hour

                  Sp  = Average  network  vehicle  speed  for other
                       trips  in miles per hour

Next,  determine  a  value for  the  proportion  of  Work and Other Trip Lengths
that  are  less  than  the Effective Radius  (default = 0.40).   Enter  data  on  the
next  four  lines  from Worksheets  6  and  7  and compute  values  for  the VMT
variables.   Next,  enter the proportion of VMT  occurring in  the  peak  hour  of
the day  (default  = 0.10).  The proportion of  off-peak  hour  VMT  is 1  minus  the
peak  hour  proportion (default  =  0.90).   Next,  enter  the proportions  of VMT
traveled  on  three  facility types  (exoressways.  arterials,  local streets)  for
*To convert  to  kilometers  for Worksheet 7M, multiply by 1,609,
                                    2-26

-------
                                TABLE  2-6
                      DEFAULT TRIP GENERATION RATES*
Land Use Type
Single Family Detached
Single Family Attached
Multiple Family
Large Commercial
Medium Commercial
Small Commercial
Office - Professional
Manufacturing
Wholesale
Education
Other
Recreation
Trips Per Day
Per Measure
Dwelling Units
Dwelling Units
Dwelling Units
1 ,000 Square Feet
(Square Meters)
1 ,000 Square Feet
(Square Meters)
1 ,000 Square Feet
(Square Meters)
1 ,000 Square Feet
(Square Meters)
1,000 Square Feet
(Square Meters)
1,000 Square Feet
(Square Meters)
1,000 Square Feet
(Square Meters)
1,000 Square Feet
(Square Meters)
Acres
(Square Meters)
Work Trip
Rates
1.8
1.5
1.0
0
.0
0
16
(0.17)
5.0
(0.05)
4.0
(0.04)
0
0.13
(0.0013)
0
Other Trip
Rates
9.0
7.0
5.0
40
(0.43)
64
(0.69)
67
(7.2)
0
0
0
4.0
(0.04)
5.0
(0.05)
42
(169,970)
*Values in parentheses are in metric units.
                                          2-27

-------
Work and Other type trips.  The six proportions must sum to 1.0,  If local trans-
portation data are unavailable, estimates can be made using the procedure described
in Volume III of the GEMLUP-I final report [11], or by using methods presented in
other standard publications [20,21],  Next, enter average speeds for peak and off-
peak hours, and the three facility types.  If local data are not available, de-
fault values are provided.  Next, copy the indicated trip data from Worksheet 6
and compute Sum A and Sum B.  These are used to calculate the Heavy Duty Correc-
tion factor, and finally, the proportion of vehicles in various classes1.

                 Worksheet 8 is used to compute VMT by various categories of area,
time of day and facility type.  Enter VMT data in the columns indicated with arrows,
enter proportion data on the appropriate lines from boxes on Worksheet 7, and per-
form the indicated arithmetic operations.

             c.  Calculation of Mobile Source Emissions

                 The methodology for computing vehicular emissions is derived from
the EPA publication Mobile Source Emission Factors [19].  Since the worksheets
in this section rely heavily upon the data in that publication, it is essential
that a copy be available before beginning the calculations.  A computer program
is available that calculates composite emission factors using data from reference
19, and so could be substituted for Worksheets 9 and 10.  A copy of this program
can be obtained from the Office of Transportation and Land Use Policy, U.S.
Environmental Protection Agency, AW-445, 401 M Street S.W., Washington, D.C.,
20460 (202-755-0603).

                 Worksheet 9 is used to compute an average emission factor (in
grams per mile, or grams per kilometer) for a given set conditions involving ve-
hicle type, speed and pollutant.  Since there are 42 vehicle types, 6 vehicle
speeds (reflecting peak or off peak hours, in conjunction with 3 facility types),
and 3 pollutants3,  it is necessary to fill out 72 separate copies of Worksheet 9
for a given area of analysis.  If one representative speed for the entire area is
used, the total number will be reduced to only 12, but the results will necessarily
be less accurate.  The first step in completing Worksheet 9 is to record carefully
all of the descriptive data requested at the top of the worksheet.  Next, fill in
the model years in column (2), starting with t + 10 and decreasing down to t - 2.
Values for columns (3),  (4),  (6) and (7)1* should be transcribed from the EPA
publication [19]5.  Next, compute the Total Emission Rate in column (5) by sum-
ming columns (3) and (4).  The Model Year Total Emissions in column (8) are then
obtained by taking the product of columns (5), (6), and (7).  The average emis-
sion factor is the sum of column (8).

i Note a proportion value for Light Duty Truck/Gas is not calculated since a fixed
  value of 0.118 is used.

2 Emission factors for motorcycles are not included in the worksheets since these
  are usually negligible in an area total.  If a particular case study contains
  relatively large amounts of motocycle traffic, emissions can be calculated
  using the procedures in reference [19].

3 There are really five criteria pollutants which motor vehicles emit.  However,
  average emission factors for particulates and sulfur oxides are relatively
  invariable.  Appropriate values are listed in Table 2-7.

it There are other correction factors for such things as air conditioning (A. ),
  vehicle load (L ), trailer towing (U..-DW), and humidity (H-D).  These correction
  factors make little difference in composite emissions, thus we have not included
  them in the worksheets.  If deemed significant, the emissions can be corrected
  using the procedures in reference [19].
5 When using Worksheet 9M, use the metric value for Ci  , in grams per kilometer.
                                         2-28

-------
                                 TABLE 2-7

               PARTICULATE AND SULFUR OXIDE EMISSION FACTORS
                                Particulates               Sulfur Oxides
    Vehicle Type               gr/mi    gr/km             gr/mi      gr/km
Automobile/Gas                  0.54     0.33              0.13      0.08
                               (0.25)*  (0.15)*

Light Duty Truck/Gas            0.54     0.33              0.18      0.11
                               (0.25)*  (0.15)*

Heavy Duty/Gas                  1.31+    0.81+             0.36      0.22

Heavy Duty/Diesel               1.31     0.81               2.80      1.70
*Unleaded gasoline
+Assumes an average of 8 tires per vehicle

Source:  [22].


                 Worksheet 10 is used to calculate a composite emission factor
for a given combination of vehicle speed and pollutant.  Since there are 6
vehicle speeds and 5 pollutants, it is necessary to fill out 30 separate
copies of Worksheet 10.  The first step in completing Worksheet 10 is to
again carefully record the descriptive data requested at the top.  Next,
enter the appropriate Average Emission Factors from Worksheets 9 and the
vehicle class proportions from the bottom of Worksheet 7.  Emission factors
for sulfur oxides and particulates are listed in Table 2-7.  Column (3) is
the product of columns (1) and  (2).  Compute the Composite Emission Factor
by summing column (3).

                 Worksheet 11  is used to calculate Total  Motor Vehicle Emis-
sions in both the Area of Analysis and the Impact Area for a given pollutant.
Since there are 5 pollutants,  it is necessary to fill  out 5 separate copies of
Worksheet 11.   The first step  is to record the pollutant at the top of the
worksheet.  Next, Average Route Speeds from Worksheet 7, corresponding to  the
various conditions,  should be  listed in column (1) and VMT data from Worksheet
8 should be listed in column (2).   Next, record in column C3) the values for
Composite Emission Factors from Worksheets 10.   Total  Emissions in column  (4)
are computed by taking the product of columns (2) and (3).  Area of Analysis
and Impact Area totals are then obtained by summing column C4) in two parts
and multiplying by the units correction factor of 0,805.


                                    2-29

-------
               d.  Emission Summary

                   The emission estimates provided by the worksheets in this
section are totals for an entire year (in pounds or kilograms of pollutant)
in both the Area of Analysis and the Impact Area.  These emission projections
correspond to conditions in year t + 10.

                   Worksheet 12 is used to calculate Stationary Source Emissions
for a given combination of fuel type (oil, gas, or electricity) and pollutant.
Since the first two fuel types each generate 5 types of pollutants, 10 combina-
tions result.  When electricity is the fuel, the only "pollutant" is kilowatt-
hours (which are later converted to air pollutant emissions at an off-site
generating plant).  Thus, it is necessary to fill out 11 separate copies of
Worksheet 12.  The first step in completing Worksheet 12 is to record the fuel
type and pollutant at the top of the worksheet.  Next, enter the amounts of
total land use in column (1) from Worksheet 6 (note, Total Manufacturing Land
Use is recorded at the 'bottom).  Next, determine for each land use type, the
proportion of the total that will be using the fuel which is recorded at the
top of the worksheet  (i.e., this should be a value between 0.0 and 1.0) and
enter these values in column (2).  Note that for a given land use type, the
proportion values used in the 11 copies of this worksheet must sum to 1.0.
If local data are not available, national statistics from the U.S. Census
[23,24] may be used instead.  Next, record the process, space heating, and
space cooling emission factors in columns (3), (5), and (7), respectively.
These values can be obtained from Tables 2-8 through 2-13*.  The space heating
and space cooling emission factors must first be multiplied by the appropriate
degree-day or operating-hour statistics, displayed in Figures 2-2 through 2-4.
Next, a composite Industrial Emission Factor, based on the expected mix of SIC
codes in manufacturing development, should be computed from Table 2-14 and re-
corded at the bottom  of the worksheet.  Process Emissions, in column (4), are
calculated by taking  the product of columns (1), (2), and (3).  Space Heating
Emissions, in column  (6), are the product of columns (1), (2) and (5), while
Space Cooling Emissions in column (8) are obtained by multiplying columns (1),
(2), and (7).  Columns  (4), (6), and (8) are then separately summed to obtain
Total Process, Space  Heating and Space Cooling Emissions, respectively.  Total
Industrial Emissions  are obtained by performing the indicated multiplication.

                   Worksheet 13 is used to calculate total emissions (in
pounds/year or kilograms/year) for each of the 5 criteria pollutants in both
the Area of Analysis  and the Impact Area.  First, summarize the 11 copies of
Worksheet 12 in  rows  (1) through (4) and sum to obtain Total Stationary Source
Emissions in row (5).  Next, record the values of Total Motor Vehicle Emissions
from Worksheet 11 in  rows (6) and (11).  Compute the Total Emissions, Area of
Analysis, in row (7), as the sum of rows (5) and (6).  Next, enter values
for the Electric Utility Emission Factors in row (8), from either local data
or Table 2-15 (assuming either coal, oil, or gas powered generators).  Then,
bring the value  for Total Kilowatt-Hours down and record it in each blank in
row (9).  The Electric Utility Emissions, in row (10), is then just the product
of rows (8) and  (9).  Compute the Total Emissions, Impact Area, in row  (12),
as the sum of rows (5), (10), and (11).
*If Worksheet 12M is used, all of these emission factors must be converted
 from pounds to kilograms by multiplying by 0.454.


                                   2-30

-------
                                                        TABLE 2-8
                           SINGLE FAMILY DETACHED OR ATTACHED LAND USE BASE EMISSION FACTORS
                            pound of pollutant (or kilowatt-hours) per measure
                             PM
                                   SO.
                CO
   HC
NO.
kWh
Measure
ro
i
GO
Space Heating
      Electricity
      Gas
      Oil
Air Conditioning
    Central
      Electricity
      Gas
    Room
      Electricity
Process
    Hot Water
      Electricity
      Gas
      Oil
    Cooking
      Electricity
      Gas
    Miscellaneous
                          2.6 x  10"4    1.5  x 10~5   5.1 x 10"4
                          2.2 x  10"3    3.2  x 10~2S  1.1 x 10"3
                           1.8  x  10~4    1.1  x  10~5   3.5 x  10"4
                           3.0  x  10"1    1.8  x  10"2   6.0  x 10"1
                           2.5
3.7 x 10"]S  1.2
                           1.1  x  10"1    6.6  x  10"3    2.2  x  10"1
                          2.0 x  10"4   2.6  x  10"3
                          6.6 x  10"4   2.6  x  10"3
                          1.4 x 10"4  1.8 x 10"3
2.4 x 10"1   3.0
7.5 x 10"1   3.0
                          8.8 x 10"2  1.1
                        3.8       dwelling unit-ht.d.d.
                                  dwelling unit-ht.d.d.
                                  dwelling unit-ht.d.d.
                        4.7       dwelling unit-op.hr.
                                 dwelling unit-op.hr.
                                                           a.c. unit-operating hour
         1.4x10    dwelling  unit-year
                   dwell ing  unit. year
                   dwelling  unit -year

         3.5x10    dwelling  unit-year
                   dwel ling  unit-year
         7.9x10    dwelling  unit-year
        Note:  A 1600 square foot dwelling unit is assumed.
               'S1  represents the percent  sulfur  in oil, by weight.

-------
                                                        TABLE  2-9
                                     MULTIPLE FAMILY  LAND  USE  BASED  EMISSION  FACTORS
Activity
PM
pound
S0x
of
pol
lutant
CO
(kilowatt-hours)
HC
per
measure
NOX
kWh
Measure
ro
co •.
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
     1.7 x 10"2S
             2.4 x 10
             5.7 x 10
             -4
             -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.
dwell ing unit- ht.d.d.
dwelling unit-ht.d.d.


dwell ing unit-op.hr.
dwell ing unit«op.hr.
dwell ing unit-op.hr.

 a.c. unit-op.hr.
                                                                                       1.1 x 10     dwelling unit*year
Gas
Oil
Cooking & Dryer
Electricity
Gas
Mi seel laneous
2
2

_
1
-
.4 x 10'1
.0


.2 x 10"1

1
2

_
7
-
.4 x ID''
.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
2

_
1
-
.4
.4


.2

dwelling un
dwelling un

+3
3.8 x 10 dwelling un
dwelling un
+3
4.4 x 10 dwell ing un
     Note:    A 900 square foot dwelling unit is assumed.
             'S' represents the percent sulfur in oil, by weight.

-------
                                                   TABLE 2-10
                           COMMERCIAL AND WHOLESALE LAND USE BASED EMISSION FACTORS
pound of pollutant (or kilowatt-hours) per measure
Activity PM SOV CO HC N0¥ kWh
X X
Space Heating
Electricity - - - - - 1.3
Gas 9.8 x 10"5 5.9 x 10"6 2.0 x 10~4 7.8 x 10"5 9.8 x 10"4 -
Oil 1.7 x 10"3 1.2xlO~2S 2.9 x 10"4 3.3 x 10"2 4.4 x 10"3 -
Air Conditioning
Electricity - - - - - 5.2
Process
Hot Water
Electricity - - - - - 5.0xl02
Gas 2.4 x 10"2 1.4xlO"3 4.8 x 10"2 1.9xlO~2 2.4 x 10"1
Oil 5.2 x 10"1 3.65 9.1 x 10"2 l.OxlO*1 1.4
Lighting - - - - - 8.0 x TO3
Auxiliary - - - - - 3.6 x 103
Equipment
Appliances - - - - - 2.0 x 10
Refrigeration - - - - - 8.9 x 10
Measure

3 2
NT ft
103ft2
103ft2

3 2
itrfr


103ft2
3 2
3 2
i(rfr
3 ?
lo^fr
103ft2

103ft2
3 2
l(Tft

•ht.d.d.
•ht.d.d.
•ht.d.d.

•cl.d.d.


•year
•year
•year
•year
•year

•year
•year
Note:  'S' represents the  percent sulfur in oil, by weight.

-------
                                                        TABLE  2-11
                                   OFFICE-PROFESSIONAL  LAND USE  BASED EMISSION FACTORS
ro
u>
Activity
Space Heating
Electricity
Gas
Oil
Air Conditioning
Electricity
Gas
Oil
Process
PM

_
9.4 x 10"5
1.7 x 10"3

_
7.4 x 10"5
1.3 x 10"3
-
pound of pollutant (or kilowatt-hours) per measure
SO CO HC NO kWh
X A

_
5.6 x 10"6
1.2 x 10"2S

_
4.4 x 10"6
9.1 x 10"3S
-

_
1.9 x 10"4
2.9 x 10"4

_
1.5 x 10"4
2.3 x 10"4
-

_
7.5 x 10"5
3.3 x 10"2

..
5.9 x 10"5
2.6 x 10"2
-

1.9
9.4 x 10"4 -
4.4 x 10"3 -

1.5
7.4 x 10"4 -
3.4 x 10"3 -
2.8 x 10+4
Measure

103ft2-ht.d.d.
103ft2-ht.d.d.
103ft2-ht.d.d.

103ft2«cl.d.d.
103ft2-cl.d.d.
103ft2* cl .d.d.
103ft2-year
  Note:   'S1 represents the  percent sulfur in oil,  by weight.

-------
ro
i
CO
en
                                                       TABLE 2-12


                                       EDUCATION LAND USE BASED EMISSION FACTORS
Activity
Space Heating
Electricity
Gas
Oil
Air Conditioning
Electricity
Gas
Oil
Process


_
8.
1.

_
2.
4.
-
PM


0 x IO"5
2 x IO"3


3 x 10"5
1 x 10

pound of pollutant (or kilowatt-hours) per measure
SO CO HC NO kWh Measure
X X

_
4.8
8.5

_
1.4
2.8
-


x 10"6
x IO"3


x 10"6
x IO"3


_
1.6 x
2.1 x

_
4.6 x
7.1 x
-


ID'4
io-4


io-5
ID'5


_
6.4
2.4

_
1.8
8.0
-


x IO"5
x IO"2


x IO"5
x IO"3


-
8.
3.

-
2.
1.
-


0x10
2 x IO"3


3 x IO"4
1 x IO"3


1.7 103ft2 • ht.d.d.
103ft2 • ht.d.d.
103ft2 • ht.d.d.

4.7 x IO"1 103ft2 ' cl.d.d.
103ft2 * cl.d.d.
103ft2 • cl.d.d.
7.1 x 103 103ft2 'year
    Note:   'S'  represents the percent  sulfur  in  oil,  by weight,

-------
ro

u>
01
                                                        TABLE 2-13


                                          OTHER LAND USE BASED EMISSION FACTORS

Activity
PM
pound of pollutant (or kilowatt-hours) per measure
SO CO HC NO kWh
X *»
Measure
Space Heating



Air




Electricity
Gas
Oil
Conditioning

Electricity
Gas
Oil
_
9.4 x 10"5
1.4 x 10"3


_
2.3 x 10"5
4.1 x 10"4
mm
5.6 x 10"6
9.9 x 10"3S


-
1.4 x 10"6
2.8 x 10"3S
_
1.9 x 10'4
2.5 x 10"4


-
4.6 x 10"5
7.1 x 10"5
_
7.5 x 10"5
2.8 x 10"2


-
1.8 x 10"5
8.0 x 10"3
_
9.4 x 10"4
2.8 x 10"2


-
2.3 x 10"4
1.1 x 10"3
Process _____
1.7 1
1
1

_1
4.7 x 10 ' 1
1
1
1.2 x 10+4 1
3 2
0 ft
03ft2
03ft2

•3 0
vn
n3 2
03ft2
3 2
•ht
•ht
•ht


•cl
•cl

-------
                                    TABLE  2-14

             ESTIMATED  NATIONAL  INDUSTRIAL  LAND USE  BASED  EMISSION
       FACTORS  BY  TWO DIGIT  1967  STANDARD  INDUSTRIAL CLASSIFICATION CODE
Pounds of pollutant (or
SIC Partic-
Code ulates SO
J\
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
10 &
640
1220
580
60
60
110
3120
10
100
1060
510
170
4030
3060
140
220
220
680
950
39 80
500
1020
540
40
70
80
3090
20
460
2780
380
170
2670
2380
120
180
200
480
700
130
3 2
KWH of electricity) per 10 ft floor area per year
CO HC NO KwH
X
13
25
14
1.4
3.4
2.2
69
0.68
11
55
10
4.7
720
61
3.5
4.7
5.3
13
18
3.5
3.3
14
8.1
0.84
2.3
1.2
40
0.48
8.1
38
5.8
2.9
38
34
2.1
2.7
3.2
6.8
9.5
2.4
130
230
140
15
45
21
690
9.5
160
730
97
52
610
570
36
46
56
110
150
44
38,000
48,000
68,000
16,000
22,000
14,000
85,000
25,000
181,000
426,000
50,000
18,000
78,000
297,000
33,000
31,000
56,000
54,000
38,000
31 ,000
Note:   The following is  assumed:   2%  sulfur  in  coal
                                 10%  ash  in  coal

                                0.2%  sulfur  in  distillate  oil
                                1.75% sulfur in residual oil

       1967 SIC codes are used  because of data  availability.  The
       1972 SIC code manual  provides  conversions  between 1967 and
       1972 codes  [25].
                                       2-37

-------
                            Figure 2-2:  NORMAL SEASONAL HEATING DEGREE DAYS ( BASE 65°F ) 1941-1970
ro
i
CO
co

-------
ro
GO
               Figure 2-3:  NORMAL SEASONAL COOLING DEGREE DAYS ( BASE 65°F )  1941-1970

-------
ro

-fa.
o
       FIGURE
2-4: ANNUAL AIR CONDITIONER COMPRESSOR-OPERATING  HOURS  FOR HOMES  THAT ARE NOT
     NATURALLY VENTILATED.

-------
                                                     TABLE 2-15

                           TYPICAL UNCONTROLLED EMISSION FACTORS FOR ELECTRIC UTILITIES


coal
oil
gas
PM

5.23
6.34
1.19
Ibs. emissions per kWh sold to customer
SO.. CO HC
•3
x 10' JA
x 10"4
x 10"4
x ,,
1.53 x
1.26 x
7.13 x
10 "S
10"2S
io-6
4.03 x
2.38 x
2.02 x
-4
10 '
io-4
io-4

1
1
1

.21 x
.58 x
.19 x
_4.
10 4
ID'4
io-5
NO

x 9
2.21 x
8.32 x
8.32 x
10 u
io-3
io-3
ro
        Note:   A 33.3% overall plant efficiency is assumed for coal-fired plants [26].
               A 31.6% overall plant efficiency is assumed for oil- and gas-fired plants [26]
               A ]Q% transmission loss 1s assumed [27].
               'S'  and 'A1 represent, respectively, the percent sulfur and ash in the fuel
               by weight.

-------
         4.  English Unit Worksheets

             Summarized in this section are all English unit worksheets,
Metric unit worksheets begin on Page 2-60.
                                           2-42

-------
                                                                     WORKSHEET 1

                                                         •INPUT DATA RECORD IN ENGLISH UNITS
Variable Name
                                Value/Computation
Inputs To Worksheets   Units
    Area of Analysis            	                                       = |_
         Vacant Developable     	
         Vacant Undevelopable   	
    Vacant Land               = Vacant Developable  T  (Area of Analysis - Vacant Undevelopable)  = |_
         Median Price           	
         Median Income          	
    Land Cost                 = Median Price * Median  Income                                   = |_
         Collection Capacity    	
    Peak Flow                   	                                       = |_
    % Collection Reserve      = 100*  (Collection Capacity - Peak Flow)  v  Peak Flow             = f
         Manufacturing Workers  	
         Tract Area           = 	
    Manufacturing Density     = Manufacturing Workers  * Tract Area                             = [_
    Normobility                 	                                       = |~
         Drivers                	
T3        County Area            	
00   Driver Density
         Sewered Land
    Sewer Service
         School Kids
         Dwelling Units
    Kid Density
         Limited Access
    Interchanges
         Current Employment
         Future Employment
         SMSA Area
    Employment Growth
         Track
    Railroads
                          = Drivers -f County Area
                          = Sewered Land - Area of Analysis
                            Schools Kids T Dwelling Units
                          = 640* Limited Access * Area of Analysis
                          = (Future Employment - Current Employment) - SMSA Area
                                                                                                                   Acres
                                                                                                                   Acres
                                                                                                                   Acres
                          = 640* Track •? Area of Analysis
     Zoned Office
                       Million Gallons Per Day
                       Million Gallons Per Day
                       *
                       Employees
                       Square Miles
                       Employees Per Square Mile
                       *
                       100s of Drivers
                       Square Miles
                       100s of Drivers Per Square Mile
                       Acres
                       *
                       Children
                       100s of Dwell ing Units
                       Children Per 100 Dwelling Units
                       Interchanges
                       Interchanges Per Square Mile
                       Employees
                       Employees
                       Square Miles
                       Employees Per Square Mile
                       Miles
                       Railroad Miles Per Square Mile Land
                       Area
                       Acres
 Unitless
                                                                                                                                                  ENGLISH

-------
                                                                WORKSHEET 1 (CONTINUED)
                                                          INPUT DATA RECORD IN ENGLISH UNITS
Variable Name
  Value/Computation
Inputs  to Worksheets   Units
Office Zoning
House Density
     Zoned Industrial
Industrial Zoning
Population Growth
Office Vacancy
Airport Distance
     Office Workers
Office Employees
     Future Population
Employee Ratio
Unemployment
Future Income
Government
Collection Reserve
     Interceptors
Interceptor Density

Poverty
Onsite Restrictions
County Growth
     Project Cost
     Federal Funds
     Index One
     Index Two
     Population Served
= Zoned Office r Area of Analysis
= 100* Dwelling Units T Tract Area

= Zoned Industrial * Area of Analysis
  Office Workers •=• Tract Area
= Future Employment * Future Population
= Collection Capacity - Peak Flow
  640* Interceptors T Area of Analysis
                       Dwelling Units Per Square Mile
                       Acres
                       Miles
                       Employees
                       Employees  Per  Square  Mile
                       People
                                                                                     J   Millions  of $
                       Million  Gallons  Per  Day
                       Miles
                       Miles  of Interceptor Pipe Per Square
                       Mile Land Area
                                                                                         Thousand
                                                                                         Thousand
                                                                                         People
 Unitless
                                                                                                                                                 ENGLISH

-------
                                                               WORKSHEET 1  (CONTINUED)




                                                         INPUT DATA RECORD IN ENGLISH UNITS
Variable Name Value/Computation

S ewe i" Costs = 1,000* ((Project Cost T Index One) - (Federal Funds - Index Two)) T Population Served =

Treatment Capacity

Vacant Houses
County Interchanges
Interchange Density = Interchanges r (640* County Interchanges -f County Area) =
Zoned Residential


Current Income
Income Growth = Future Income - Current Income =
Current Medicals
Future Medicals


Current Houses
Future Houses
Housing Growth = (Future Houses - Current Houses) T Current Houses =

Restriction Years

Phasing =

Transit Stops =

CBD Distance

Inputs to Worksheets





























Units

$ Per Person

Million Gallons
Per Day
*
Interchanges
*
Acres
*

$
$
Employees
Employees
*

Dwelling Units
Dwelling Units
*

Years

it

it

Miles

Unitless
                                                                                                                                                ENGLISH

-------
                                                                         WORKSHEET 2


                                                                     LAND USE PROJECTIONS
ro
i
Residential
Commercial
Office-
Professional
Manufacturing
Wholesale
Highways
Education
Recreation
Other
= (-8692 *
- (18804 *
= (4302 *
- (17.04 *
= (719.5 *
- (1791 *
= (2591 *
+ (234.7 *
= (166.5 *
+ (6949 *
= (35.22 *
+ (25.94 *
= (-3.186 *
- 405.2 *
= (270.4 *
- (3226 *
= (1134 *
- (245.4 *

Vacant
Land )
Nonmobility)
Sewer
Service)
Employment
Growth )
Railroads)
Industrial
Zoning )
Railroads)
Office
Vacancy)
Driver
Density)
Unemployment)
Railroads)
Interceptor
Density )
Poverty)
County
Growth)
Railroads)
Vacant
Houses)
Railroads)
Interchange
Density )

+ (5347 *
+ (487.2 *
- (4515 *
- (2.626 *
+ (33478 *
+ (572.0 *
- (3890 *
+ (22.7 *
+ (86.28 *
- (0.1065
+ (22.98 *
+ 11.35 =
+ (723.0 *
- (1.415 *
+ (2.631 *
- (557.8 *
- (0.3020
+ (403.8 *
Land
Cost)
Driver
Density)
Vacant
Land )
%Col lection
Reserve
Office
Zoning)
Population
Growth )
Vacant
Land )
Airport
Distance)
Office
Vacancy)
Future
* Income)
Land
Cost)
Mil f
10,000
Sewer
Service)
Sewer
Costs)
Treatment
Capacity )
Industrial
Zoning )
Future
* Income)
Residential
Zoning )

+ (16.42 *
+ 134P6 =
- (28.84 *
+ 5376 =
+ (36.70 *
- 459.9 =
+ (2.978 *
+ 611.2 =
(
+ (3.340 *
- 1911 =
- (0.1167
Per
Acres
+ (437.5 *
+ 428.9 =
+ (322.1 *
+ 59.52 =
- (12175 *
+ 3087 =
% Collection
Reserve )+ (11.24 *
- • Dwnl linn Unite Por
10,000 Acres*
Kid
Density) + (6249 *

Per 10,000 Acres
Peak Flow) + (0.3324 *

Per 10,000 Acres
Manufacturing
Density ) + (85923 *

Per 10,000 Acres
Office
Employees) + (6043 *

Per 10,000 Acres
Government) + (0.5379 *
Land
Cost) + (95.17 *

Per 10,000 Acres
County
Growth) + (0.9742 *

10,000 Acres
Vacant
Houses) + (649.0 *

Per 10,000 Acres

Manufacturing
Density
Interchanges)
House
Density)
Office
Zoning)
Employee
Ratio )
Collection
Reserve )
On-Site
Restrictions)
Office
Employees)
County
Growth)

         10,000  Acres  refers  to  Area  of Analysis
                                                                                                                                                 ENGLISH

-------
                                                                       WORKSHEET 3

                                                                FINAL  LAND USE PROJECTIONS
                (1)
       Land Use Projections
         From Worksheet  2
              (2)
        Disaggregation
         Percentages
      (3)
Area of Analysis
   T By 10,000
           (4)
Final  Land Use Projection
     (1)  * (2) * (3)
                                                                                                                                 (Per Area of Analysis^
 Residential
% Single Family
                  Total Single Family Detached =
                             Dwelling Units
 Residential
% Two Family
                  Total Single Family Attached =
                             Dwelling Units
 •Residential
% Multi Family
                  Total Multiple Family =
                             Dwelling Units
 Commercial
% Large Commercial  '
                  Total Large Commercial =
                             1,000 Square Feet
 Commercial
% Medium Commercial 2
                  Total Medium Commercial =
                             1,000 Square Feet
 Commercial
% Small  Commercial   3
                  Total Small Commercial =
                             1,000 Square Feet
 Office-Professional
 Manufacturing
 Wholesale
 Education
 Other
 Recreation
 Highways
I
                                                  Total Office-Professional =
                                                  Total Manufacturing =
                                                  Total Wholesale =
                                                  Total Education =
                                                  Total Other =
                                                  Total Recreation =
                                                  Total Highways =
                                                                   1,000 Square Feet
                                                                   1,000 Square Feet
                                                                   1,000 Square Feet
                                                                   1,000 Square Feet
                                                                   1,000 Square Feet
                                                                   Acres
                                                                   Lane Miles
          i
            Large commercial = commercial development with floor area > 100,000 ft  .
          7                                                                                        2
           Medium commercial = commercial development with floor area between 50,000 and 100,000 ft
          o                                                                      2
            Small commercial = commercial development with floor area < 50,000 ft  .
                                                                                                                                                ENGLISH

-------
                                          WORKSHEET 4 (Optional)

                                     DEFAULT DISAGGREGATION EQUATIONS
% Single                                    Kid
Family     =!-(!+ 92.8 * EXP (-0.0137 * Density
                                                                           Income
                                                              - 0.000918 * Growth
                                                                    Sewer
                                                          - 0.784 * Service
                      Office
             - 22.7 * Zoning
                                         2.98 * Nonmobility
                                              Hospital
                                      0.181  * Growth
SMulti
Family
                  Single
           = (1-% Family
                      Office
             + 34.1 * Zoning
                                Housing
)  * (1  4 (1  + 1.24* EXP (-438 * Growth
+ 5.48 * Poverty
                                       - 0.732 * Railroads
                                   - 0.0365 * Peak Flow
% Two
Family     = (1-X Single Family
                                          - % Multiple Family
% Large                                    Office
Commercial =1*0+ 0.104 * EXP (0.215 * Vacancy
                                                                       Onsite
                                                             - 0.507 * Restrictions
                        Kid
             - 0.0159 * Density
                                                     Restriction
                                            0.0603 * Years      	))
                                                                       Treatment
                                                            + 0.0232 * Capacity
% Medium          Large
Commercial = (1-% Commercial
                        CBD
             + 0.0178 * Distance
                                     _)*(!*(!+ 0.585 * EXP (1.54 * Phasing 	


                                           + 0.00417 * Government           - 1.48 * Vacant Land
                                                                    Transit
                                                        + 0.00459 * Stops
% Small
Commircial  = (1-3J Large Commercial
                                             - % Medium Commercial
                                          )   =
Note:  EXP is the exponential function.  Thus EXP(X) = e .
                                                                                                                           ENGLISH

-------
                             WORKSHEET 5 (OPTIONAL)



                    CONFIDENCE  INTERVALS FOR PREDICTED LAND USE
Final  Projected  Land  Use  Category
PREDICTOR VARIABLE
Name
1.

2.
3.
4.
5.
6.
Constant Term
COMPUTING VARIANCE OF DEPENDENT
Covariance
Predictor Term from
Variables Appendix A
0 x 
-------
Covariance
Predictor Term from
Variables Appendix A
© x ® x
(D x (D x
® x © x
(D x (5) x
(D x © x

-------
                                   WORKSHEET 6

                               MOTOR VEHICLE TRIPS
           Effective Radius  (in miles) = 0.0223 *-^(Area of Analysis, in acres)

                                                  Miles
(1)
Total Land Use
Single Family
Detached
Single Family
Attached
Multiple
Family
(2)
Work Trip
Rates



(3)
Work Trips
(1)*(2)



(4)
Other Trip
Rates



(5)
Other Trips
(1)*(4)



                        Residential
                        Work Trips
Residential
Other Trips
Large
Commercial
Medium
Commercial
Small
'Commercial
Office-
Professional
Manufacturing
Wholesale
Education
Other
Recreation




























i







                        Total  Work
                        Trips
Total Other
Trips
*See Page 2-26 for a definition of Effective Radius.
                                      2-51
                                                                            ENGLISH

-------
                                                              WORKSHEET 7



                                                     VEHICLE MILES TRAVELED (VMT)
Work Trip
Length Miles

Other Trip Work (Default Other
Length Miles Proportion = 0.40) Proportion

VMT?,, [= (Total Work * Work Trip ) - (Residential
IU Trips
Length Work Trips

VMTflu | = (Total Work ^ Effective ) - (Residential
AW Trips
Radius Work Trips

VMTT(V |= (Total Other * Other Trip ) - (Residential
10 Trips
Length Other Trips

VMTnn |= (Total Other * Effective ) - (Residential
AU Trips
Peak Hour Proportion
Facility
Work
Expressways
Arterial s
Local Streets

Trip Sum A I = Manul

Radius Other Trips

(Default = 0.10) Off Peak Hour Pro
* Work
Proportion
* Work
Proportion
* Other
Proportion
* Other
Proportion

portion (
(Default
= 0.40)
* Effective )
Radius
* Effective )
Radius
* Effective )
Radius
* Effective )
Radius
3efault = 0.90)
Proportions Average Route Speeds (miles per hour)
Other Peak
(
(
(
"acturing Work + Manufacturing Other

Off Peak
Default = 37)
Default = 20)
Default = 15)
+ Wholesale Work

(Default = 45)
(Default = 28)
(Default = 18)
+ Wholesale Other

Trip Sum B [ |= Total Work Trips + Total Other Trips

HD Correction | = (!

Automobile/Gas Proportion 1
Heavy Duty/Gas Proportion |
Heavy Duty/ Diesel Proportion
>um A - Sum B) - 0.05, if less than 0, set equal to 0.
1 = 0.804 - HD Correction
1 = 0.346 + 0.3 * KD Correction

[ |= 0.032 + n.2 * HD Correction



*Subscripts for VMT variables are defined  as  follows:   I  =  impact area,  A  =  area of analysis, W = work trips, 0 = other trips.
                                                                                                                                    ENGLISH

-------
AREA OF ANALYSIS

VMTAp, | 1 = Peak

VMTAp/,.| |= Peak

VMTAp. | |= Peak

VMTAdEl |= Off Peak

VMTAtfifl 1= Off Peak

VMTA(f| ] I = Off Peak
IMPACT AREA

VMTIprl |= Peak

VMT.pA.| |= Peak

VMT.p. | |= Peak

VHTTny | |= Off Peak

VMTI0v| |= Off Peak

VMT7n-, | |= Off Peak

ilOKK-'lE'/T J
CAiilCORI~ r
U"irA,,i
* ( * Expressways
Work
* ( * Arterials
Work
* ( * Local
Work
* ( * Expressways
Work
* ( * Arterials
Work
* ( * Local
Work
* ( * Expressways
Work
* ( * Arterials
Work
* ( * Local
Work
* ( * Expressways
Work
* ( * Arterials
Work
* ( * Local
Work
4 v . 
-------
         Vehicle Type =
         Pollutant = 	
         Speed = 	
                           WORKSHEET 9
                 MOTOR VEHICLE EMISSION FACTORS

(Automobile Gas,  Light Duty Truck Gas, Heavy Duty Gas,  or Heavy Duty Diesel)
                                       Region = 	 (Low alt.,  High alt.,  or Calif.)
                                                                            °F
      (CO, NO. or HC)
             A
miles per hour
         Year of Impact Assessment (t+10) =
Ambient Temperature =
Cold Starts =
                                                                          Hot Starts =
(1)
Vehicle
Age
(Years)
1
2
3
4
5
6
7
8
9
10
11
12
>13
(2)
Model
Year













(3)
Base Emission
Rate
«W













(4)*
Hydrocarbon Crankcase/
Evaporate Emission Rate
(H^













(5)
Total Emission
Rate
(3)+(4)













(6)
Fraction of
Travel














(7)
Speed/Temp. /Cold-Hot
Starts Correction Factor
'ripstwx'













(8)
Model Year
Total Emissions
(5)*(6)*(7)













IS}
en
         Note Hi = 0 for CO and NOX; for these pollutants,
         enter the values for C-   directly in column (5).
                                                              Average Emission Factor

-------
                                WORKSHEET 10
                  COMPOSITE MOTOR VEHICLE EMISSION FACTORS
                   Speed
               (miles per hour)
        Pollutant
        (CO, NOV, HC, SOV or Particulates)
               X        X
Vehicle Class
      (1)
Average Emission
     Factor
     (2)
Vehicle Class
  Proportion
                                                                     (3)
                                                                   Product
Automobile/Gas
Light Duty Truck/Gas
Heavy Duty/Gas
Heavy Duty/Diesel
                          0.118
                                         Composite Emission Factor
                                   2-55
                                               ENGLISH

-------
         WORKSHEET 11



 TOTAL MOTOR VEHICLE EMISSIONS



Pollutant

Condition
Peak, Expressways
Peak, Arterial s
Peak, Local Streets
Off Peak, Expressways
Off Peak, Arterial s
Off Peak, Local Streets
(1)
Average
Speed
i






(2)
VMT
Data
VMTAp£
VMTAPA-
VMTAPL
VMTA(JE
VMTAO*
VMTA(JL
Area of Analysis, Total Motor Vehicle Emissions
Peak, Expressways
Peak, Arterial s
Peak, Local Streets
Off Peak, Expressways
Off Peak, Arterial s
Off Peak, Local Streets






VWIPE
VHTIpA.
VHTIpL
VMTIO'E
VMTIOA-
VHTIOL
(3)
Emission
Factors






Sum
(4)
Total Emissions
(2)*(3)




!


* 0.805
=

(in pounds/year)






Sum
*
Impact Area, Total Motor Vehicle Emissions -
C1n p<

|





0.805
•
)unds/year)
               2-56
                                           ENGLISH

-------
                                                                      WORKSHEET 12



                                                               STATIONARY SOURCE EMISSIONS
r-o
i
en
            Fuel Type
(Gas,  Oil,  or Electricity)
  Pollutant
(CO,  HC,  NO.  SOV,  Particulates,  or Kwh)
           X     X
           TOTAL EMISSIONS
(1)
Total Land Use
SF Detached
SF Attached
Mult. Family
Large Commercial
Med. Commercial
Small Commercial
Office-Professional
Wholesale
Education
Other
(2)
Fuel
Proportion










(3)
Process
Emission
Factor










(4)
Process
Emissions
(D*(2)*(3)










(5)
Space Heating
Emission
Factor










(6)
Space Heating
Emissions
(D*(2)*(5)










(7)
Space Cooling
Emission
Factor










(8)
Space Cooling
Emissions
(D*(2)*(7)










         Process
Space Heating
       Space Cooling
                                 Industrial
                   =  Total  Manufacturing Land Use



                                *  Fuel  Proportion
                             * Industrial Emission Factor
                                                                                                                                                   ENGLISH

-------
5.   Metric Unit Worksheets
    Summarized in this section are all  metric unit worksheets.
                           2-59

-------
                                                                            WORKSHEET 1M
                                                                  INPUT DATA RECORD IN METRIC UNITS
      Variable Name
                            Value/Computation
                                                                   Inputs to Worksheets   Units
CTl
O
Area of Analysis
     Vacant Developable
     Vacant Undevelopable
Vacant Land
     Median Price
     Median Income
Land Cost
     Collection Capacity
Peak Flow
% Collection Reserve
     Manufacturing Workers
     Tract Area
Manufacturing Density
Nonmobility
     Drivers
     County Area
Driver Density
     Sewered Land
Sewer Service
     School Kids
     Dwelling Units
Kid Density
     Limited Access
Interchanges
     Current Employment
     Future Employment
     SMSA Area
Employment Growth
     Track
Railroads

     Zoned Office
                                = Vacant Developable * (Area of Analysis - Vacant Undevelopable)
                                = Median Price T Median Income
                                = 100* (Collection Capacity - Peak Flow) -.  Peak Flow
                                  Manufacturing Workers * Tract Area
= Drivers •=• County Area
                                = Sewered Land T Area of Analysis
                                = School  Kids v Dwelling Units
                                = 1,000,000* Limited Access T Area of Analysis
                                = (Future Employment - Current Employment)  * SMSA Area
                                = 1,000,000* Track r Area of Analysis
                                                                                                                          Square Meters
                                                                                                                          Square Meters
                                                                                                                          Square Meters
Million Gallons Per Day
Million Gallons Per Day
*
Employees
Square Kilometers
Employees Per Square Kilometers
*
100s of Drivers
Square Kilometers
100s of Drivers Per Square Kilometer
Square Meters
*
Children
100s of Dwelling Units
Children Per 100 Dwelling Units
Interchanges
Interchanges Per Square Kilometer
Employees
Employees
Square Kilometers
Employees Per Square Kilometer
Kilometers
Railroad Kilometers Per Square
Kilometer Land Area
Square Meters
       Unitless
                                                                                                                                                      METRIC

-------
                                                               WORKSHEET 1M (CONTINUED)
                                                           INPUT DATA RECORD IN METRIC UNITS
Variable Name
  Value/Computation
Inputs to Worksheets    Units
Office Zoning
House Density
     Zoned Industrial
Industrial Zoning
Population Growth
Office Vacancy
Airport Distance
     Office Workers
Office Employees
     Future Population
Employee Ratio
Unemployment
Future Income
Government
Collection Reserve
     Interceptors
Interceptor Density

Poverty
Onsite Restrictions
County Growth
     Project Cost
     Federal  Funds
     Index One
     Index Two
     Population Served
= Zoned Office * Area of Analysis
= 100* Dwelling Units v Tract Area

= Zoned Industrial T Area of Analysis
  Office Workers -f Tract Area
= Future Employment * Future Population
= Collection Capacity - Peak Flow
  1,000,000* Interceptors - Area of Analysis
                       Dwelling Units  Per Square Kilometer
                       Square Meters
                       Kilometers
                       Employees
                       Employees  Per Square Kilometer
                       People
                       Mill ions  of $
                       Mi 11 on Gallons  Per Day
                       Kilometers
                       Kilometers  of  Interceptor Pipe Per
                       Square Kilometer Land Area
                                                                                          Thousand $
                                                                                          Thousand $
                                                                                          People
 Unitless
                                                                                                                                                   METRIC

-------
                                                              WORKSHEET 1M  (CONTINUED)
                                                         INPUT DATA RECORD  IN ENGLISH UNITS
Variable Name
  Value/Computation
Inputs to Worksheets   Units
Sewer Costs
Treatment Capacity
= 1,000* {(Project Cost * Index One) - (Federal Funds * Index Two)) * Population Served =
Vacant Houses              	
     County Interchanges   	
Interchange Density      = Interchanges -f (640* County Interchanges * County Area)
     Zoned Residential     	
Residential Zoning       = Zoned Residential * Area of Analysis
     Current Income        	
Income Growth
     Current Medicals
     Future Medicals
Hospital Growth
     Current Houses
     Future Houses
Housing Growth
Restriction Years
Phasing
Transit Stops
CBD Distance
  Future Income-Current Income
= (Future Medicals - Current Medicals) * Current Medicals
= (Future Houses - Current Houses) * Current Houses
                       $ Per Person
                       Million Gallons
                       Per Day
                                                                                                                 Interchanges
                                                                                                                 *
                                                                                                                 Square Meters
                       Employees
                       Employees
                       *
                       Dwell ing Units
                       Dwell ing Units

                       Years
                       *
                                                                                                                 Kilometers
 Unitless
                                                                                                                                                METRIC

-------
                                                                         WORKSHEET  2M

                                                                     LAND  USE  PROJECTIONS
CTl
CO
Residential
Commercial
Office-
Professional
Manufacturing
Wholesale
Highways
Education
Recreation
Other
= (-2.148 *
- (4.646 *
= (987.5 *
- (10.13 *
= (102.7 *
- (411.0 *
= (369.6 *
+ (53.87 *
= (98.23 *
+ (1595 *
= (0.0087 *
+ (25.94 *
= (-731.3 *
- (93.02 *
= (168.1 *
- (3226 *
= (161.8 *
- (67.81 *

Vacant
Land )
Nonmobil ity)
Sewer
Service)
Employment
Growth )
Railroads)
Industrial
Zoning )
Railroads)
Office
Vacancy)
Driver
Density)
Unemployment)
Railroads)
Interceptor
Density )
Poverty)
County
Growth)
Railroads)
Vacant
Houses)
Railroads)
Interchange
Density )

+ (1.321 *
Land
Cost)
Driver
+ (0.3117 * Density)
- (1036 *
Vacant
Land )
%Col lection
- (0.6029 * Reserve )
+ (7685 *
+ (131.3 *
- (893.0 *
+ (3.239 *
+ (19.80 *
- (0.0244 *
+ (0.0091 *
+ 0.0045 =
+ (170.0 *
- (0.3248 *
+ (2.631 *
- (557.8 *
- (0.0693 *
+ (92.70 *
Office
Zoning)
Population
Growth )
Vacant
Land )
Airport
Distance)
Office
Vacancy)
Future
Income)
Land
Cost)

10,000
Sewer
Service)
Sewer
Costs)
Treatment
Capacity )
Industrial
Zoning )
Future
Income)
Residential
Zoning )

+ (0.0041 *
+ 3.327 =
- (6.621 *
+ 1234 =
+ (8.424 *
- 105.6 =
+ (1.770 *
+ 140.3 =
+ (1.985 *
- 438.7 =
- (0.00005
lometers Per
Square Meter
*• (100.4 *
+ 98.45 =
+ (322.1 *
+ 122.1 =
- (2795 *
+ 708. r> =
%Collection Manufacturin
Reserve ) + (0.0072 * Density

10,000 Acres*
Kid
Density) + (3714 * Interchanges)

Per 10,000 Square Meters
House
Peak Flow) + (0.1981 * Density)

Square Meters Moor Area
Manufacturing Office
Density ) + (19724 * Zoning)

Per 10,000 Square Meters
Office Employee
Employees) + (1387 * Ratio )

Per 10,000 Square Meters
Collection
* Government) + (0.0002 * Reserve )
5
Land Ons ite
Cost) + (21.85* Restrictions)
«•
Per 10,000 Square Meters
County Office
Growth) + (2.522 * Employees)

Per 10,000 Square Meters
Vacant County
Houses) + (150.0 * Growth)
c>> ,., 	 Uni. 	 rl 	 „ 	 ,
Per 10,000 Square Meters

        10,000  Square  Meters  refers  to Area  of Analysis
METRIC

-------
                                                                           WORKSHEET 3M

                                                                    FINAL  LAND USE PROJECTIONS
                    (1)
           Land Use Projections
            From Worksheet 2M
              (2)
        Dlsaggregation
         Percentages
      (3)
Area of Analysis
   4 By 10,000
           (4)
Final  Land Use Projection
     (1) * (2) * (3)
                                                                                                                                     (Per Area of Analysis
     Residential
% Single Family
                  Total Single Family Detached =
                             Dwelling Units
     Residential
% Two Family
                  Total Single Family Attached =
                             Dwelling Units
     Residential
% Multi Family
                  Total Multiple Family =
                             Dwelling Units
     Commercial
% Large Commercial
                                                       1
                  Total Large Commercial
                             Square Meters
     Commercial
% Medium Commercial
                  Total Medium Commercial =
                             Square Meters
     Commercial
% Small Commercial
                  Total Small Commercial =
                             Square Meters
IS)
I
en
     Office-Professional
     Manufacturing
     Wholesale
     Education
     Other
     Recreation
     Highways
                                                  Total Office-Professional =
                                                  Total Manufacturing
                                                  Total Wholesale =
                                                  Total Education =
                                                  Total Other =
                                                  Total Recreation
                                                  Total Highways =
                                                                   Square Meters
                                                                   Square Meters
                                                                   Square Meters
                                                                   Square Meters
                                                                   Square Meters
                                                                   Square Meters
                                                                   Lane Kilometers
                  1                                                                    2
                    Large commercial  = commercial  development with floor area> 9,290 m .
                  2                                                                                      2
                    Medium commercial  = commercial  development with floor area  between 4,645 and 9,290 m .
                  3                                                                   2
                  Small  commercial  = commercial  development with floor area< 4,645 m .
                                                                                                               METRIC

-------
                                                                   WORKSHEET 4M (Optional )

                                                              DEFAULT DISAGGREGATION  EQUATIONS
          % Single
          Family
                                 Kid
= 1  T (1 + 92.8 * EXP (-0.0137 * Density
                                          Income
                             -  0.000918 *  Growth
                                               Office
                                      - 22.7 * Zoning
                            + 2.98 * Nonmobility
                                              Hospital
                                     +  0.181 * Growth
                  Sewer
        - 0.784 * Service
            J)
          % Multi
          Family
       Single
  (1-% Family
                                               Office
                                      + 34.1  * Zoning
                                 Housing
J  *  (1  T  (1  +  1.24*  EXP  (-438  *  Growth
+ 5.48 * Poverty
                            - 0.454 * Railroads
                                    - 0.0365 *  Peak  Flow
          % Two
          Family
= (1-% Single Family
         -  % Multiple  Family
CTi
          % Large
          Commercial
                                Office
= 1 T (1 + 0.104 * EXP (0.215 * Vacancy
                                      Onsite
                            -  0.507  *  Restrictions
                                                 Kid
                                      - 0.0159 * Density
                                          Restriction
                               + 0.0603 * Years
                                       J)
                     Treatment
          + 0.0232 * Capacity
          % Medium
          Commercial
       Large
  (]-% Commercial
                                                 CBD
                                      + 0.0111  * Distance
    J  * (1  T  (1  + 0.585  * EXP  (1.54  *  Phasing 	


          +  0.00417 * Government      ,     -  1.48  *  Vacant  Land
                  Transit
      + 0.00459 * Stops
                                                                                              J))
          % Small
          Commercial
  (1-% Large Commercial
            -  % Medium Commercial
          Note:   EXP 1s the exponential  function.  Thus EXP(X) = e .
                                                                                                                                                     METRIC

-------
                           WORKSHEET 5M,(OPTIONAL)
                    CONFIDENCE INTERVALS FOR PREDICTED LAND USE
Final Projected Land Use Category
PREDICTOR VARIABLE
     Name
                                         2.
                                         3.
                                         4.
                                         5.
                                         6.
     Constant Term
                                         7.   1.0
COMPUTING VARIANCE OF DEPENDENT VARIABLE
(T) x  (f) x  Covariance	
(T) x  (?) x  Covariance 	
(l) x  (§) x  Covariance 	
(I) x  (4) x  Covariance
(T) x (5) x  Covariance
(T) x (e) x  Covariance
(V) x ^7/ x  Covariance
(?) x (T) x  Covariance
(?) x (2) x  Covariance
(?) x (3) x  Covariance
(2) .x (T) x  Covariance
(?) x (£) x  Covariance
(?) x (6) x  Covariance
(?) x (T) x  Covariance
(T) x CD x  Covariance
(§) x (?) x  Covariance
(D x (§) x  Covariance
(5) x (£) x  Covariance
(3) x (§) x  Covariance
(5) x (6) x  Covariance
(f) x (?) x  Covariance
(4) x (T) x  Covariance
(4) x (2) x  Covariance
(4) x (§) x  Covariance
(4) x (J) x  Covariance
(4) x © x  Covariance
(7) x © x  Covariance
(4^ x (z) x  Covariance
                                          a.
                                        = b.
                                        = c.
                                        = d.
                                        = e.
                                        = f.
                                        = £:
                                          h.
                                        = i.
                                        = k.
                                        = 1.
                                        = m.
                                        - n.
                                        = o.
                                        = £1.
                                        - r.
                                        - s.
                                        - t.
                                        •= u.
                                        = V.
                                          w.
                                          X.
                                        - z.
                                          aa.
                                        = bb.
                                   2-66
                                                                          METRIC

-------
                          WORKSHEET  5M (OPTIONAL)
                    CONFIDENCE  INTERVALS FOR PREDICTED LAND USE
£) x (2) x  Covariance
Ji) x CD x  Covariance
&) x (T) x  Covariance
[6) x (§) x  Covariance
*&) x (§) x  Covariance
[D x CD x  Covariance
5) x (T) x  Covariance
5) x (D x  Covariance
5) x © x  Covariance
5) x (4) x  Covariance
2) x (?) x  Covariance
5) x © x  Covariance
f?) x (7) x  Covariance
Sum (a) through @)  =
Standard Deviation of Dependent
    Variable ="/  @
t - statistic of predictive equation

If Final Projected Land Use is Resi-
    dential or Commercial, set equal
    to the disaggregation percentage
± Confidence Interval
    -  $6) x  $4)  x  ©  T 100
                                         ee.
COMPUTING VARIANCE OF DEPENDENT  VARIABLE
(?) x CD x  Covariance 	=  cc.
(?) x (D x  Covariance 	-  dd.
(?) x (§) x  Covariance 	
(5) x (4) x  Covariance 	
(?) x (?) x  Covariance 	
© x (6) x  Covariance 	
© x (?) x  Covariance 	
® x CD x  Covariance
                                          ff.
                                         gg.
                                       = hh.
                                       = 11.
                                       - JJ.
                                       = kk.
                                       =  11.
                                         mm.
                                       = nn.
                                       = oo.
                                         PP.
                                         go.
                                       - rr.
                                       = ss.
                                       = tt.
                                       = uu.
                                          vv.
                                       - ww.
                                          43.
                                          44.
                                          46.
                                          48.
                                               1.69
                                          47.   100
                                  2-67
                                                                         METRIC

-------
                                  WORKSHEET 6M

                              MOTOR VEHICLE TRIPS
     Effective Radius (in km) = 0.000564 * yj (Area of Analysis,  in  meters2)

                                                Kilometers
(1)
Total Land Use
Single Family
Detached
Single Family
Attached
Multiple
Family
(2)
Work Trip
Rates



(3)
Work Trips
0)*(2)



(4)
Other Trip
Rates



'(5)
Other Trips
(D*(4)



                        Residential
                        Work Trips
Residential
Other Trips
Large
Commercial
Medium
Commercial
Small
Commercial
Off ice -
Professional
Manufacturing
Wholesale
Education
Other
Recreation




























•







                        Total  Work
                        Trips
Total Other
Trips
*See Page 2-26 for a definition of Effective Radius.
                                        2-68
                                                                            METRIC

-------
                                                               WORKSHEET  7M
                                                     VEHICLE  KILOMETERS TRAVELED (VKT)
Work Trip
Length

VCT!H

VKTflM

VKTIO ;

VKT.Q
t>o Peak Hour Proporf
i
CTl

-------
                                                       WORKSHEET 8M

                                                      CATEGORIZED VKT
AREA OF ANALYSIS

VKTAPF | 1 = Peak

VKTAPA'l 1= Peak


1= Peak

VKTAOF 1 1= Off Peak

VKW 1 1 = of f Peak

VKTA1t. | 1= Off Peak
ro
^ IMPACT AREA
VKTTPF | |= Peak

VKTIPA'I l= Peak

VKTTPI | |= Peak
It'L ' 	 • •
VKT-mV | | = Off Peak

VKTirfA* 1 1 = Off Peak
m
70
0 VKTT(J, 1 1 = Off Peak
1UL ' 	 ' ' 	
* (
* (
* (
* (
* (
* (
* (
* (
* (
* (
* (
* (

* Expressways
Work
* Arterials
Work
* Local
Work
* Expressways
Work
* Arterials
Work
* Local
Work
* Expressways
Work
* Arterials
Work
* Local
Work
* Expressways
Work
* Arterials
Work
* Local
Work
+ * Expressways ]
Other
+ * Arterials !
Other
+ * Local 1
Other
+ * Expressways !
Other
+ * Arterials \
Other
+ * Local ]
Other
+ * Expressways )
Other
+ * Arterials )
Other
+ * Local )
Other
+ * Expressways }
Other
+ * Arterials )
Other
+ * Local )
Work
 *Subscripts  for VMT  variables are defined as follows:  I = impact area, A = area  of analysis, F = peak hour
  0 = off-peak hour,  E =  expressways, A = arterials, L = local streets.

-------
                                                                        WORKSHEET 9M
                                                             ' MOTOR  VEHICLE  EMISSION  FACTORS
          Vehicle Type
          Pollutant = _
          Speed = 	
(Automobile  Gas,  Light Duty Truck  Gas,  Heavy Duty Gas,  or Heavy  Duty  Diesel)
                                       Region =  	 (Low  alt.,  High  alt.,  or Calif.)
                                                                            °F
      (CO, N0x, or HC)
miles per hour
         Year of  Impact Assessment  (t+10) =
Ambient Temperature =
Cold Starts =
                                                                          Hot Starts =
(1) .
Vehicle
Age
(Years)
1
2

3

4

5
6
7
8
9


10
n
12
£13
(2)
Model
Year



















(3)
Base Emission
Rate



















(4)*
Hydrocarbon Crankcase/
Evaporate Emission Rate
(H^


(5)
Total Emission
Rate
(3)+(4>


|




1








i
T






•

(6)
Fraction of
Travel








(7)
Speed/Temp. /Cold-Hot
Starts Correction Factor
^ripstwx'




























(8)
Model Year
Total Emissions
(5)*(6)*(7)


















  ro
•yo
i—i
o
         Note H-J = 0 for CO and NOX; for these pollutants,
         enter the values for C.   directly in column (5).
                                                              Average Emission Factor

-------
                                WORKSHEET 10M


                  COMPOSITE MOTOR VEHICLE EMISSION FACTORS
                   Speed
               (miles per hour)
        Pollutant
        (CO, NO. HC, SOV or Particulates)
               A        A
Vehicle Class
      (1)
Average Emission
     Factor
     (2)
Vehicle Class
  Proportion
  (3)
Product

0)*(2)
Automobile/Gas


Light Duty Truck/Gas


Heavy Duty/Gas


Heavy Duty/Diesel
                          0.118
                                         Composite Emission Factor
                                       2-72
                                                                         METRIC

-------
 TOTAL MOTOR VEHICLE EMISSIONS



Pollutant

Condition
Peak, Expressways
Peak, Arterial s
Peak, Local Streets
Off Peak, Expressways
Off Peak, Arterial s
Off Peak, Local Streets
(1)
Average
Speed






(2)
VKT
Data
VKTApE
VKTApft.
VKTApL
VKTAO'E
VKTArfA'
VKTAOl
Area of Analysis, Total Motor Vehicle Emissions
Peak, Expressways
Peak, Arterials
Peak, Local Streets
Off Peak, Expressways
Off Peak, Arterials
Off Peak, Local Streets
-





VKT,pE
VKT,pA.
VKT,pL
VKTIO'E
VKTIOV
VKTIOL
(3)
Emission
Factors






Sum
(4)
Total Emissions
(2)*(3)


|



.
* . 0.365
(in ki






Sum
*
Impact Area, Total Motor Vehicle Emissions
(in k-

lo grams /year)






.
0.365

ilograms/year)
               2--73
METRIC

-------
                                                                  WORKSHEET 12M



                                                           STATIONARY SOURCE EMISSIONS
        Fuel Type
(Gas,  011,  or Electricity)
  Pollutant
(CO,  HC,  NOX,  SOX,  Partlculates,  or Kwh)
(1)
Total Land Use
SF Detached
SF Attached
Mult. Family
Large Commercial
Med. Commercial
Small Commercial
Office-Professional
Wholesale
Education
Other
(2)
Fuel
Proportion










(3)
Process
Emission
Factor










(4)
Process
Emissions
0)*(2)*(3)










(5)
Space Heating
Emission
Factor










(6)
Space Heating
Emissions
d)*(2)*(5)










(7)
Space Cooling
Emission
Factor










(8)
Space Cooling
Emissions
0)*(2)*(7)










       TOTAL EMISSIONS
         Process
Space 'Heating
       Space  Cooling
                             Industrial
                   =  Total Manufacturing Land Use



                               * Fuel Proportion
                            *  Industrial Emission Factor
O

-------
                                                        WORKSHEET 1 3M




                                                     EMISSIONS SUMMARY
Total Emissions
(1) Process
(2) Space Heating
(3) Space Cooling
(4) Industrial
(5) Total Stationary
Source Emissions
(1H2H3H4)
(6) Area of Analysis,
Total Motor
Vehicle Emissions
(7) Total Emissions
Area of Analysis
(5)+(6)
(8) Electric Utility
Emission Factors
(9) Total Kilowatt-
Hours
(10) Electric Utility
Emissions
(8)*(9)
(11) Impact Area,
Total Motor
Vehicle Emissions
[12)Total Emissions,
Impact Area
(5H10H11)
CO












HC












NOX












S°x












Particulates












Kilowatt-Hours





V
-*".

ro



en

-------
         6.  Example on Using Worksheets

             This section presents an example of the use of the computation
worksheets, in English units.  One example of each worksheet is filled out,
even though in an actual application, several copies of some of the work-
sheets would be required.  The data used in this example were collected
from case study #1, Willimantic, CT.
                                    2-76

-------
ro
                                                                         WORKSHEET 1
                                                              INPUT  DATA RECORD IN ENGLISH  UNITS
       Variable Name
                           Value/Computation
                                                                Inputs  To '..'o"''sheets   Units
Area of Analysis
     Vacant Developable
     Vacant Undevelopable
Vacant La^:'
     Median Price
     Median Income
Land Cost
     Collection Capacity
Peak Flow
% Collection Reserve
     Manufacturing  Workers
     Tract Area
Manufacturing Density
No mobility
     Drivers
     County Area
Criver Density
     Sewered Land
Sewer Service
     School Kids
     Dwelling Units
Kid Density
     Limited Access
Interchanges
     Current Eroployrest
     Future Employment
     SM3A Area
Employment Growth
     Track
Railroads

     Zoned Office
                                      6456
                                                                                      Acres
                                                                                      Acres
                                                                                      /V.rcs
      'ji D
        3
                                         Developable.^ (Area  cf Analysis - Vacant Undevelopable)
                                          /rOQ
                                =  Median Price v Median Income
                                           9.0
                                =  100*  (Collection Capacity  -  Peak Flow) * Peak Flow
                                                                        9.0
                                                                        76
                                           4.5*
                                  Manufacturing Workers •=• Tract Area
                                         O.S7Z
                                  Drivers * County Area
                                         3 ce^
                                  Sewered Land T Area of Analysis
                                           37	
                               =  Schools Kids 4 Dwelling  Units
                                  	o
                               =  640* Limited-Access 4 Area of Analysis
  	5/4	
- (Future F-.-loyiv-rt - Current Employment)  r SMSA Area

= 640* Track  T Area of Analysis
                                                                                                      0.248
Million C.i "lens  Per  Cay
Million Gallons  Per  Oey

Employees
Square Miles
Employees  Per  Square "ile
*
100s of Drivers
Square Miles
100s of Drivers  Per  Square Mile
Acres
*
Children
100s of Dwelling  Units
Children Per ICO  :v:cll ir; Jrits
Interchanges
Interchanges Per  Square '^le
Employees
Employees
Square Miles
Employees  Der  Square Mile
"'les
Railroad '-'iles Per Sq-jare "ile Land
Area

-------
ro
^J
co
                                                                   WORKSHEET 1 (CONTINUED)
                                                              INPUT DATA RECORD IN ENGLISH UNITS
      Variable Name
                           Value/Computation
Office Zoning
House Density
     Zoned  Industrial
Industrial  Zorirj
Population  Growth
Office Vacancy
Airport Distance
     Office Workers
Office Employees
     Future Population
Employee Ratio
Unemployment
Future Income
Government
Collection  Reserve
     Interceptors
Interceptor Density

Poverty
Onsite Restrictions
County Growth
     Project Cost
     Federal Funds
     Index  One
     Index  Two
     Population Served
                               = Zoned Office  v Area of Analysi*
                               = 100* Dwelling Units 4- Tract Area
                                         770
                               = Zoned Industrial T Area of Analysis
                               = Office Workers v Tract Area
Future  Employment 4- Future Population
       0.075-
       10067
                                          10.9
                               = Collection  Capacity - Peak Flow
                                          1.2.
                               = 640* Interceptors v Area of Analysis
                                          690
                                                             Inputs to  Worksheets   Units
                                                                                   Dwelling  Units Per Square f-'i's
                                                                                   Acres
                                                                                   Miles
                                                                                   Employees
                                                                                   Employees Per Square '-'He
                                                                                   People
                                                                                   ."•ill ions of 3
                                                                                   Million Gallons Per Day
                                                                                   Miles
                                                                                   Miles of Interceptor Pipe Per Square
                                                                                   Kile Land Area
                                       Ife  973
                                                                                   Thousand S
                                                                                   Thousand $
                                                                                   People
       Unit!ess

-------
                                                                     WORKSHEET 1  (CONTIiiUED)

                                                               INPUT DATA RECORD  IN ENGLISH UNITS
ro
 i
       Variable Name
                           Value/Computation
                                                                                       Inputs to  Worksheets   Units
                               = 1,000* ((Project Cost •=•  Index One) - (Federal  Funds  -•-  Index Two)) * Population  Served =   f Stf • 374-  j  $ per Person
                                       o.ozs
                               = Zoned Residential  - Area of Analysis
Sewer Costs
Treatment Capacity

Vacant Houses
     County Interchanges   	^» f	
Interchange Density     = Interchanges T (640* County  Interchanges v County Area)
     Zoned Residential
Residential Zoning
     Current Income
Income Growth
     Current Medicals
     Future Medicals
Hospital  Growth
     Current Houses
     Future Houses
Housing Growth
Restriction Years
Phasing
Transit Stops
CBD Distance
= Future Income  -  Current Income
            fe
            10
= (Future Medicals - Current Medicals)  r Current Medicals
           72
          a-76
= (Future Houses - Current Houses) T Current  Houses
                                                                                                                                            nil! ion Gal Ions
                                                                                                                            o.tt?
Interchanges
*
Acres

S
$
Employees
Employees

Dwelling Units
Dwell incj Units
*

Years
                                                                                                                                            Miles
        Unit! ess

-------
                                                                            WORKSHEET  2

                                                                        LAND USE  PROJECTIONS
           Residential
                (-8692 * Q*7S>L


               •  (18804 * C
                                 acant
                                Land  )
Commercial     = (4302 *
                               Sewer
                               Service)
                                              + (5347  *


                                              + (487.2 *
                                                                          Driver
                                                                          Density)
+ (16.42 *


+ 13466  =
  % Collection
 _Reserve     )  + (11.24  *

         Dwelling Units Per
         10,000  Acres*
                             Bx,/i Employment
                            rjQ Growth    )
                                                                     --.^Vacant
                                                         -  (4515 * 0* 7/5 Land  )

                                                                    n     %Collection
                                                         -  (2.626 *   /O   reserve    )   + 5376
                                                                                                                                            '••nu'Ecturing
                                                                              -  (28.84 *   C^7  Density)      + (6249 *  D    Interchanges)
                                                                                            4  * ^tf^M    jw« ~^*.«
                                                                                            / T/13   Per 10,000
           Office-
           Professional
              = (719.5 * 0*245Rail roads)     +  (33478 -

                             .,   Industrial
               - (1791  * OJIT3 Zoning     )    +  (572.0 *
                                                               Office
                                                               _Zoning)

                                                               Population
                                                                                                                                 House
                                                                                           x%                                    nouse
                                                                               +  (36.70 *   7*0 Peak Flow)    + (0.3324 * IQZi- Density)
                                                                          Growth    )    -459.9  =   _/_ O
                                                                                          —/—-^-^-  1,000 Square Feet
                                                                                               ~      Per 10,000 Acres
oo
o
Manufacturing = (2591
           Wholesale
               + (234.7 *
                               Railroads)

                                 Office
                                 Vacancy)
                                              - (3890 *


                                              + (22.7 *
                                                                  Qf?7j Land  )
                                                                       .  Airport
                                                                      /  Distance)
                        ._  5/55Driver
                (166.5 * W»Jl3Density)
                                                        n  i  Office
                                             +  (86.28 * ^» *-  Vacancy)
               + (5949
                       * d)<.Q7v5
                                          Unemp1oynient) _ (0.1065 *
                                                                Future
                                                                Income)
           /• oo o Manufacturing                ^,^ Office
+ (2.978 * be*?lgt Density     )+  (85923  * CM?^Z Zoning)
+ 611.2  =
_^^» ^ -   I ) WVJ W ^>^U'_lt IUC
93/   Per 10,000 Acres
                                                                                                Employees)    + (6043 *
                                                                                                       Per 10,000 Acres
                                                Employee
                                                Ratio   )
Highways      = (35.22 * 0*24% Railroads)
                                                        A crtf-jLand
                                             +  (22.98 * OOTlCost)
                                                                              - (0.1167
               +  (25.94 *
                                 Interceptor
                                 Density     )
                                                        + 11.35 =
                                                                     Miles Per
                                                                     10,000 Acres
                                                                                                                                     f  <*  Collection
                                                                                                         Government)    + (0.5379 *  fe>*o  Reserve    )
                            01 in
                           *l£p Poverty)

                        n  01 2 C°Unty
               - 405.2  * U» MT> Growth )
                                                         A £T7/^ewer
                                                (723.0 *  ^^ // Service)
                                             - (1.415 *
                                                               Sewer
                                                               Costs)
                                                                                         +  (437.5  *
                                                                                         +  428.9  =
                               + (95.17 *

                        1,000 Square Feet
                        Per 10,000 Acres
                                                                                                                               On-S-'te
                                                                                                                               Restrictions)
Recreation    = (270.4
                                          Railroads)
                         f\ A?iTVacan'l:
                 (3226 * 0*065 Houses)
                                                         <2  *>  Treatment                .. -7C. County
                                             +  (2.631 *  *'** Capacity )     +(322.1 * Ut^J)Growth)

                                                               [Industrial
                                                                                                                           _    Office
                                                                                                             +  (0.9742 *  O'Z~ Employees)
                                                                   t\ ;;/y-2lndustrial                 .	—,    ^   Acres Per
                                                        - (557.8 * ^"^^Zoning    )     + 59.52  =  I   O7»JLI 10,000 Acres
           Other
              = (1134 * 0>Z4& Rail roads)      -  (0.3020  *10C?67 Incoml)       - (12175
0                                 Interchange
                 v ._-,„.-,   _ Density    )
                                                                   _  ^..i Residential               i
                                                        + (403.3 * >?lf?rTZcrin9     )    +  3087   = L
                                                                                                             +  (649.0 *

                                                                                                      1 ,000 Square Feet
                                                                                                      Per 10,000 Acres
                                                                                                                                          County

-------
 I
00
                                                                                '.•.'O'K'-.KECT  3

                                                                         FINAL LAND I'SF  PROJEC
i                n:'
I       Lani  Use  Projections    i
|          From  V.'orksheet  2     i
!                               I
                                                        (2)
                                                  Disaggregation
                                                   Percintanes
                                                                      of  Analysis :
                                                                      3y  10.000
Final  Land Use Projection
     0) * (2) * (3)
         ! Residential
                                % Single Family    ,39$        «  6^5*6  j  Total  SinSle Far":i^ C-t-:chc'J =
                                                                                            Tf?1
resicer-f'al
Co-rercial
Commercial
Conercial
74*4
1473
14-73
( 473
% ;';.;! ti Family
% Large Commercial
',' Kedium Commercial
% Snail Medium
j
602 i -6454 Totel -Uipic Family = ^907 C -.-I-.; J«;ts '
! i
O&7 ' ^4^^ ! "r°':a^ Large Co~i~:3rcia'i = 9^9 1 -CGO Squ:*''- Fset '
All j £4££ ( Total r'edium Commercial = /A j^ 1 .COO Sc^are Feet '
• V 1 1 1 ** -f ^"W «^
1 1
rtA^ ' fiA^tm. ' Total Snail Co^aercial = / „ O^ 1 ,C".~ Squc^e Feet 1
• W«B : • ^™^^P . • • TV '
Office-Professional |Q QO

                  n
                                                                                             iota" Of"ice-Professi~r.a

-------
                                                                      '/.'OP.KS-iEET 4  (Optional)

                                                                 DEFAULT  C!SAf?TREGATIO\  CQUATIONS
             :i Single
             Family
                                                                        Y   + 0.00417 * Government   '
                                                                                                                     Transit     -.^
                                                                                                         + 0.00459  *  Stops       O
                                                                                       /   -  1.48 *  Vacant Land   ^'   __'•))
Commercial   j  Q.Q03»  = (1-- Large Commercial   Ot°vl - « ;-edium
                                                                                   Cornnercial
             Note:   EXD  is  the  exponential function.  Thus  EXP(X) = e  .

-------
                             WORKSHEET 5 (OPTIONAL)



                    CONFIDENCE  INTERVALS  FOR  PREDICTED  SAND  USE
Final  Projected Land Use Category
PREDICTOR VARIABLE
Name
VA£d/vd UjMrd
Ld/nuL tost
%£oMwM,ty.w
MdM- D&wtu
Nfliaw)bittU|
Drwt D^fy
Constant Term
= VACANT
= UND
PV^r RE^APl
«= MAJOJOb
^STAV
- bUWE

1.
2.
3.
4. j
5.
6.
7.
COMPUTING VARIANCE OF DEPENDENT VARIABLE
Covariance
Predictor Term from
Variables Appendix A Multiplier
0 x © x .Ilk
0 x © x -Ml
0 x © x .&VL
0 x © x .103
0 x © x --^
0 x © x -414
0 x © x -.?#£
(D x © x . \0<\
© x G) x -102.
© x © x -.46!
© x (5) x -. i&q
(2) x © x ~ • *3lr'
El x
*- ~*^ X
El x
E4 x
(•• / y
E^? x
> E! x
&~7 x
E4 x
E2 x
EG x
2. &5 x
1.0
2.0
2.0
2.0
2.0
2.0
2.0
1.0
2.0
2.0
2.0
2.0

0- fH4-
10.0
win
O.'jll
Q.^o&
1.0
Resultant
, a. 4,^30,415
— * Q ^^ T 1 i ^^ X^ ^
^J X ""^ ^3
^ U • / / ^"^ "^
= e. - 2,243,096
= fl 1*5^404
— y . / /
= h. 554X5^|
= i. 04,823
= j. - 34,5iq
= k. - 116,451
k, . a4 -311
                                2-83

-------
Covariance
Predictor Term from
Variables Appendix A Multiplier
\D x (Z) x "-423 £k x 2.0
© x © x . 141 £1 x i.o
® x © x -. U4 El x 2.0
© x © x -^46? ^Ar x 2.0
© x © x ""• 113 E3 x 2.0
© x © x — •5'2j3 E4- x 2.0
© x © x .U£> &1 x 1.0
© x © x -.13^ £^5 x 2.0
s~*k. ^*+. "^ K^ *^7 f™" /**)
© x © x •o2-/ t-Z- x 2.0
© x © x .2.60 E4 x 2.0
© x © x -4-31 E8 x i.o
s\ r\f} *-/*
© x © x -. "£ Eb x 2.0
© x © x ~ • \c^b E^ x 2.0
© x © x .4^4 £^ x i.o
© x © x -.1)3 E1^ x 2.0
© x © x - \O^> C-C3 x 1.0 =
Sum (a) through (a^ :
y
Standard Deviation of Dependent Variable =v(8) :
Disaggregation proportion:
(This is 1.0 for all land uses except
residential and commercial. The pro-
portion is the decimal equivalent of
% Single Family, etc. The numbers
come from Col. 2 of Worksheet 3.)
Confidence Interval ( + ) = (9) x (fcD x 1 .69 :
Resultant
1.- W*,5-24
m. 71,0^0
n. - 103, &5^
n 45£>. €>56
w • f
P • '
q.- "74^,6^0
r. 4(e>6&, J6&
s.-q,g)^^ifo
u • /
U • / *
v • •
w • •
x. -10,075,00?
y. 7, 166
56,041
d d • / /
1,C
-------
                        WORKSHEET 6

                    MOTOR VEHICLE TRIPS


Effective  Radius*= (0.0004973* Area of  Analysis

                = 	/*79       Miles
                                                                   1/2
(1)
Total Land Use
Single Family fl
Detached «"Zi
Single Family _
Attached °
Multiple 29 9
Family x.Twy
(2)
Work Trip
Rates
Ii8
l.sr
l.O
(3)
Work Trips
(1)*(2)
2461
0
2*09
(4)
Other Trip
Rates
9.0
7.0
5.0
(5)
Other Trips
(D*(4)
/7307
o
14 543*
                       Residential
                       Work Trips
                           5370
Residential
Other Trips
Large ^ -^
Commercial ' 3f
Medium .^ ^
Commercial '*'••&
Small . ^
Commercial f*/Q
Office- ^.y
Professional «*//
Manufacturing /£34-
Wholesale ^8f
Education 75"^
Other //^
Recreation 3£.?
0
O


-------
                                                               WORKSHEET  7



                                                      VEHICLE  MILES TRAVELED  (VMT)
Work Trip / ~
Length » • *

VMT?,., I70771&

VMT. 4/7ZO

VMTin 43/ 7?£

VMTnn 1 //£ f#/
Peak Hour Proport
/ Other Trip
* Miles Length
= (Total Work
Trips
= (Total Work
Trips

25455
= (Total Other
Trips 756* 1
= (Total Other
Trips ^ZSfc'T/

ion 0

«/0 (
Facility Proportions
ro
co Expressways
Arterial s
Local Streets

Trip Sum A 1 93 «

Work
O./tT
£.£/
O.i3

Other
0-10
O.ZI
O-tto

5-61 = Manufacturing Work

Trip Sum B \IOf QQ6\ = Total
HD Correction | O,

Work Trips 4

•O42.1 = (Sum A T Sum B)

Automobile/Gas Proportion 0»*


Heavy Duty/Gas Proportion 0.
Heavy Duty/Diesel

Proportion j


/, -». Work ^ ^^ (Default Other A ^^ (Default
V.Vf Miles Proportion CJ -*IU =0.40) Proportion t*«~V =0.40)
^ Work Trip x1 O/ ) - (Residential ^ --*- * Work ^ * * Effective . «^ )
Length fc»ob Work Trips 5 Of" Proportion O & Radius *''Y
^ Effective • ^ /7 ) - (Residential -.^ — _ * Work ^ * Effective * — ^ )
Radius /» / j Work Trips ^ O/v Proportion 0 »^r Radius '**i
+ Other Trip - ) - (Residential 7/0£« * Other ^ ^ Effective y — ^ )
Length £»O/ Other Trips -J't^i Proportion G *^T Radius 9 * rj
i, Effective . ^^ ) - (Residential  (Default = 15) / W (Default = 18)
ol 7 0 + Manufacturing Other O + Wholesale Work //S^ + Wholesale Other O
2S4S& + Total Other Trips 75"£4/
- 0.05, if less than 0, set equal to 0.
762* = 0.804 - HD Correction


DfcOl = 0.046 + 0.8 * HD Correction

o.O4d =

0.032 + 0.2 * HD Correction
''Subscripts  for  VMT  variables  are  defined  as  follows:   I =  impact area, A = area of analysis, W = work trips, 0 = other trips.

-------
                                                       WORKSHEET 8



                                                     CATEGORIZED VMT
AREA OF ANALYSIS                           +VrTTfl,/                                 iVMTAn4'
      	                                       HH                            _          r\-J
VXTW I75Z = Peak O.I * (4l fZO * Expressways C/.'»
APE Work
. * f». tf 1
VMTBnl, | 324 / |= Peak O. 1 * (4C 7/C * Arterials V- CrJ
APA Work
V>'7nn, | 1794-1= Peak CP.| * (4( 7/6 * Local U« 1,3
Ml L Work
VMT.Xr 1 757*51= Off Peak O .7 * ( 41 7ZO* Expressways V.f^
AUt Work
VMT,,,,-,, I 29 /$5[= Off Peak v»7 * (4*1 /IP * Arterials D,ZJ
AOA Work r
VMT,«', 123" /*A - Off Peak C/.T *(4l ItO * Local £/»Ld
""u V/ork
INS IMPACT AREA ^VMTj^
CO
VMTjpr 1 45 WO 1= Peak C»\ *( /TO 7/(i* Expressways Q.l^

...,,„,, .^^^-,, .v.-., •/ M/70 7/6* Arterials O-t-f
IPAL^ ' Work
Of 1*0**.  10 )
Other
+ f|Z59/* Arterials 0.2f )
Other
+ IIZS1I * Local 0,2O )
Other
+ I IC^V / * Expressways 0«IO )
+ I l£5"?f * Arterials 0 • 2| )
Other
+ WZ5"f|* Local 0*20 )
Other
+ nil rj» * Expressways 0«IO )
+ 4317% * Arterials 0 -2f )
Other
+ 431 71K * Local 0»ZO )
+ T*«7ffc * Expressways 0»|0)
Other
+ 431 7/6** Arterials 0 . t( )
Other
+ 43/7%.Loca1 0.10)
Work
 Subscripts  for  VMT  variables  are  defined  as  follows:   I  =  impact area,  A = area  of analysis   P = peak hour

  0  =  off-peak  hour,  E  -  expressways,  A  =  artcrials,  L  = local  streets.                       '               '

-------
Vehicle Type =
Pollutant =  	
Speed = 	
                                                                 WORKSHEET 9

                                                       MOTOR VEHICLE EMISSION FACTORS

                                       (Automobile Gas, Light Duty Truck Gas, Heavy  Duty Gas, or Heavy Duty Diesel)

                                       (CO, NO  , or HC)                       Region =   LOvu fl>U   (Low alt.,  High  alt., or Calif.)
                                             A                                                  ~~
                                 miles  per'hour                               Ambient Temoerature =      15      °F

    Year of Impact Assessment (t+10)  =     I9°O
CO
30
Ambient Temperature  =
Cold Starts  =
                                                                                                     Hot Starts =
(1)
Vehicle
Age
(Years)
1
2
3
4
5
6
7
8
9
10
11
12
>13
(2)
Model
Year
I
0.107
o.l(£
O.IOZ
0.09i
o, off
0.077
0.044
0.o4?
0.0-S3
0.0^3
D.cM-
(7)
Speed/Temp. /Cold-Hot
Starts Correction Factor
(ripstwx^
0*
0.63
0.6*3
^63
0.^3
o.GS
0.63
0.63
0.63
0.6?
0.63
0.63
0.63
(8)
Model Year
Total Emissions
(5)*(6)*(7)
0*143
o.z/s~
O.ZZ-?
0.2+0
0.25V
0.254-
0.2-f?
C.213
0.202
0.144
£.//£
^.or/
0. Z26
ro
co
CO
    Note Hi = 0 for CO and NOX;  for these  pollutants,
    enter the values for C,   directly  in  column  (5).
                                                                                         Average Emission Factor

-------
                        WORKSHEET  10



          COMPOSITE MOTOR  VEHICLE  EMISSION FACTORS
           Speed


Pollutant       CO
       (miles  per  hour)
(CO,  NO.  HC,  S0v  or  Particulates)
       X        X
Vehicle Class
Automobile/Gas
Light Duty Truck/Gas
Heavy Duty/Gas
Heavy Duty/ Diesel
(1)
Average Emission
Factor
Z.fco
9.*0
M7
28.7
(2}
Vehicle Class
Proportion
0. 762-
0.118
0.0*0
0.040
(3)
Product
|.
-------
 TOTAL MOTOR VEHICLE EMISSIONS



Pollutant      CO	
Condition
Peak, Expressways
Peak, Arterials
Peak, Local Streets
Off Peak, Expressways
Off Peak, Arterials
Off Peak, Local Streets
(1)
Average
Speed
37
2o
(6~
4f
30
)«
(2)
VHT
Data
VMTAPE
/ant
VMTAPA
32*1
VMTApL
H794
VMTAOE
/S7(&
VMTAO'A
tlltf
WTAdL
ZSI4&
Area of Analysis, Total Motor Vehicle Emissions
Peak, Expressways
Peak, Arterials
Peak, Local Streets
Off Peak, Expressways
Off Peak, Arterials
Off Peak, Local Streets
37
2o
/£-
45-
30
IB
VMT £
* ?ro
VMTJpA
/2<5*-
VHTIpu
IOVS6
mi6i
619/1
VMTIOA
111 886
VMTIO'L
f77 37/
0.805
869 Wi
)unds/year)
5^934-
2o4-f9S"
23o 147
4tl IZZ.
/ 5%0£S&
/74B 9oe
4 ml 338
0.805
^ 3ft> /a7
?unds/year)
                2-90
                                            ENGLISH

-------
                                                                WORKSHEET 12


                                                         STATIONARY  SOURCE EMISSIONS
       Fuel Type
Oil
      (Gas, Oil,  or Electricity)
Pollutant   CO
(CO,  HC,  NO. SO. Particulates,  or Kwh)
          X    A
(1)
Total Land Use
SF Detached |fl23
SF Attached Q
Mult. Family 2^0^
73?
Large Commercial
/0.5"
Med. Commercial
/ -90
Small Commercial
671
Office-Professional
Wholesale £»f
Education YOv
Other /76
(2)
Fuel
Proportion
0.6
0.6
0-6
0*4
0.6
€>-£»
0-t
0.6
0-6
O-&
(3)
Process
Emission
Factor
L£
1^
1.0
O.Of
o.o?
0.07
0
0.0^
0
0
(4)
Process
Emissions
(D*(2)*(3)
ise^
o
1745"
5^1
1
0.1
0
U
0
0
(5)
Space Heating
Emission
Factor
6.Z7
6.Z7
3-Z5"
/.6S"
/.65^
A 65"
/-65~
/.65"
i.LO
Af3
(6)
Space Heating
Emissions
(1)*(2)*(5)
7234-
C
5T673
^30
10
Z
664
£«
5Vo
151
(7)
Space Cooling
Emission
Factor
O
0
O.I/
o
o
o
O.IZ*
0
0.0*
0. c4
(8)
Space Cooling
Emissions
(D*(2)*(7)
O
o
/fa
o
0
o
4r
$
K
+
ro

10
      TOTAL  EMISSIONS
               Process    J I ~f 9    Space Heating  ^


Industrial   JT/OlO     Total  Manufacturing  Land  Use


                                     *  Fuel Proportion
                                                                            Space  Cooling
                                                                                            * Industrial Emission Factor

-------
                                                  WORKSHEET 13

                                                EMISSIONS SUMMARY
          Total Emissions
                                   CO
                HC
NO.
SO,
Participates
Kilowatt-Hours
       (1)  Process
                              3198
       (2) Space Heating
                             / SA10
       (3) Space Cooling
ro

ro
m
•2.
CD
[—
1—4
CO
       (4) Industrial
                                 rofe
       (5) Total Stationary
           Source Emissions
                              33
       (6) Area of Analysis,
           Total Motor
           Vehicle Emissions
                            86?  69?
       (7) Total Emissions
           Area of Analysis
                            903  3SS"
       (8) Electric Utility
           Emission Factors
                                     10
       (9) Total  Kilowatt-
           Hours
       (10)Electric Utility
           Emissions
           (8)*(9)
       (11)Impact Area,
           Total Motor
           Vehicle Emissions
      (12)Total Emissions,
          Impact Area
                           =F
4*70 707
                                                             	^	i	.	
                                                                                                 /v

-------
 III.  REFERENCES

 1.    Office  of Air  Quality  Planning  and  Standards,  Guidelines  for Air
      Quality Maintenance  Planning  and Analysis; Volume 2:  Plan  Prepa-
      ration, EPA  Publication  No.   EPA-450/4-74-Q027Research  Triangle
      Park, NC, 1974.

 2.    Office  of Air  Quality  Planning  and  Standards,  Guidelines  for Air
      Quality Maintenance  Planning  and Analysis; Volume 3:  ControT
      Strategies,  EPA  Publication No.  EPA-450/4-74-003,   Research
      Triangle Park, NC, 1974.

 3.    Office  of Air  Quality  Planning  and  Standards,  Guidelines  for Air
      Quality Maintenance  Planning  and Analysis; Volume 4:  Land  Use
      and  Transportation Considerations.  EPA  Publication No.  EPA-450/4-74-004,
      Research Triangle Park,  NC, 1974.

 4.    Office  of Air  Quality  Planning  and  Standards,  Guidelines  for Air
      Quality Maintenance  Planning  and Analysis; Volume 6:  Overview of
      Air  Quality  Maintenance  Area  Analysis,  EPA Publication  No.
       EPA-450/4-74-007,   Research  Triangle Park, NC,  1974.

 5.    Office  of Air  Quality  Planning  and  Standards,  Guidelines  for Air
      Quality Maintenance  Planning  and Analysis; Volume 9:  Evaluating
      Indirect Sources, EPA  Publication No.   EPA-450/4-75-001,  Research
      Triangle Park, NC, 1975.

 6.    Office  of Air  Quality  Planning  and  Standards,  Guidelines  for Air
      Quality Maintenance  Planning  and Analysis; Volume 12:   Applying
      Atmospheric  Simulation Models to Air Quality Maintenance  Areas,
      EPA  Publication  No.EPA-450/4-74-013,Research Triangle Park,
      NC,  1974.

 7.    National Environmental Policy Act of 1969, 42  U.S.C., Section 4321
      at seq.

 8.    Council on Environmental  Quality, "Guidelines  for Preparation of
      Environmental  Impact Statements," Federal Register,  38/147), Part  II,
      August  1,  1973.

 9.    Benesh, F.,  Guldberg,  P.,  and D'Agostino, R.,  Growth Effects of Major
      Land Use Projects:   Volume  I  -  Specification and Causal Analysis of
      Model.  EPA Publication No.  EPA-450/3-76-012a,  Research  Triangle
      Park, NC,  May  1976.

10.    Benesh, F.,  Growth Effects  of Major Land Use Projects:  Volume II  -
      Compilation  of Land  Use  Based Emission  Factors,  EPA  Publication No.
      EPA-450/3-76-0126, Research Triangle Park, NC, September  1976.
                                     3-1

-------
11.   Benesh, F.,  Guldberg, P.,  and D'Agostino,  R.,  Growth  Effects  of    ^
     Land Use Projects:   Volume III - Summary,  EPA  Publication  No.
     EPA-450/3-76-012C,  Research Triangle Park, NC, September 1976.

12.   Guldberg, P., D'Agostino,  R., and Cunningham,  R., Growth Effects  of
     Major Land Use Projects (Wastewater Facilities);  Volume i;  Model
     Specification and Causal Analysis, EPA Publication No.  EPA-450/3-78-Ol4a,
     Research Triangle Park, NC, March 1978.

13.   Nie, N.H., Hull, C.H., Jenkins, J.G., Steinbrenner, K., and Bent, D.H.,
     Statistical  Package for the Social Sciences (2nd Edition),  Mc(5raw-
     Hill, New York, NY, 1975.

14.   Finney, D.J., Statistical  Methods in Biological Assay (2nd Edition),
     Hafner Publishing Co., New York, NY, 1964.

15.   Personal communication, Dr. Edward E. Cureton, University of Tennessee,
     Knoxville, TN, 1977.  The substance of the weight validity index
     technique is contained in an article of Dr. Cureton's which has been
     accepted for publication in a professional journal, but has not yet
     been released.

16.   Guldberg, P., Benesh, F. and McCurdy, T., "Secondary Impacts  of
     Major Land Use Projects," Journal of the American Institute of Planners,
     43(3):  268-269, July 1977^

17.   Heise, D., Causal Analysis, John Wiley & Sons, New York, NY,  1976.

18.   U.S. Water Resources Council, 1972 Obers Projections, Volumes  I-VII,
     Washington,  DC, April 1974.

19.   Office of Transportation and Land Use Policy,  Mobile Source Emission
     Factors, Environmental Protection Agency,  Washington, DC,  January 1978.

20.   Cerighton, "Estimating Efficient Spacing for Arterials and Expressways",
     Highway Research Board Bulletin No. 253, Washington,  DC, 1960.

21.   National Research Council, Highway Capacity Manual, 1965,  Highway
     Research Board Special Report Number 87, Washington,  DC, 1965.

22.   Office of Air and Waste Management, Compilation of Air Pollutant
     Emission Factors, Third Edition, Environmental Protection Agency
     Publication No. AP-42, Research Triangle Park, NC, August 1977.

23.   Bureau of the Census, 1970 Census of Housing,  Washington,  DC.

24.   Bureau of the Census, 1972 Census of Manufacturers. Washington, DC.

25.   Office of Management and Budget, Standard Industrial  Classification
     Code Manual, Washington, DC, 1972.
                                      3-2

-------
26.   National  Coal  Association,  Steam Electric  Plant  Factors,  Washington,
     DC,  1973.

27.   Edison Electric Institute,  Statistical  Yearbook  of the  Electric
     Utility Industry,  New York,  NY,  1972.

28.   Bureau of Economic Analysis, Population, Personal  Income,  and  Earnings
     By State,  Projections to  2000, U.S.  Department of  Commerce,  Washington,
     DC,  October 1977.
                                       3-3

-------
                                APPENDIX A

                       COMPLETE STATISTICAL OUTPUT
                       OF THE PREDICTIVE EQUATIONS


     Covariance data for the following Final  Projected Land Use categories
can be found on the indicated page:

     Final Projected Land Use Category                      Page

     Single Family Detached                                 A-2
     Single Family Attached                                 A-2
     Multiple Family                                        A-2
     Large Commercial                                       A-4
     Medium Commercial                                      A-4
     Small Commercial                                       A-4
     Office-Professional                                    A-6
     Manufacturing                                          A-8
     Wholesale                                              A-10
     Education                                              A-14
     Other                                                  A-18
     Recreation                                             A-16
     Highways                                               A-12
                                     A-l

-------
           1.  RES Equation
                    EQUATION   1
                    ***********

                    SMPL  VECTOR
                        1   40

                    ORDINARY  LEAST SQUARES

                    VARIABLES...

                           RES
                           VACANT
                           LAND
                           RECAP1
                           MANJOB
                           STAY
                           DRIVE
                           C

                    MEAN  OF DEPENDENT VARIABLE  IS  8713.5469
£
INDEPENDENT
VARIABLE
VACANT
LAND
RECAPl
MANJOB
STAY
DRIVE
C
R-SQUARED =
ESTIMATED
COEFFICIENT
-8691.57422
5347.30078
16.4241486
11.2375050
-18803.8398
487.237305
13466.2617
0.7416
STANDARD
ERROR
2675.66455
1041.f2104
3.F32589U
3.43206120
6609.87891
215.450577
3235.0POP1

T-
STATISTIC
-3.24P37875
5.12264751
4. 28539181
3.27427197
-2.844807f 2
2.26148033
4.16257286

PEAN OF
VARIABLE
0.6f<694695
0.53672218
181.299986
504. 2f 3763
0.44994849
1.72454357
1.00000000

                     DURBIN-UATSON STATISTIC (ADJ. FOP   0 GAPS)  =   2.0923

                     NU^SER  OF OBSERVATIONS =   40

                     SUM  OF  SQUARED RESIDUALS =          .356145E+09

                     STANDARD ERROR OF THE REGRESSION =          3285.16

                     ESTIMATF OF VARIANCE-COVARI A\CE MATRIX  OF  FfTIPATFC  COEFFICIENTS

                1/ACAUT-0.7T6E + 07 -0.262E + 06  0.892E+02  0.203E + 04 -0.253E+07  0.219E + 06 -C.388E + C7
                 i_/»/wo-0.262E*06  0.109E + 07  0.102E + 04 -0.467E*02 -0.189E+06 -0.522E + 05 -0.423E + 06
                      «.«92E+02  0.102E+04  0.147E*02 -0.119E*01   0.b4PF+04 -0.113E+03 -0.529E+04
                             >0<» -0.*»67E + 02 -0.119E + 01  0.118E»02 -0.139r.»05  0-327F + 02  0
                             >-07 — O»189E*O6  O*b**6E*OA —O«13^E»O5   O»^»37E*Ofl -—0-222L*O& — 0 • :

-------
    LINF.
   PRINCETO\ UNIVERSITY
                                                           *    TSP
                                                                                        CF AlubST,
                                                                                                                    PAGE.
                 PLOT OF  ACTUALC)
                                                            FITTED<+)  VALUF.S
                                                                                                PLOT  OF  RESICUALS(O)
10
       ACTUAL
FITTED
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
?2
23
24
25
26
27
28
29
30
31
32
33
35
3ft
37
38
39
4C
8209.
2782.
8223.
5911.
6920.
6257.
.2891E+05
9706.
.1197E*05
5868.
.2357E*05
9987.
.1926E+05
7094.
6072.
3745.
9709.
.1517E*05
5100.
7540.
6771.
5017.
7372.
.1894E+05
5903.
582.0
6515.
4626.
6550.
.1498F. + 05
7772.
2112.
651f.
.1050T+05
369.0
.1322E+05
6313.
5375.
Tti39.
7484.
-12.56 +
.1046E+05
5511.
.1249E+05
7808.
.2351E+05
6300.
.1512E+05
. 1 155E+05
.1731E»05
9811.
.1978E*05
4616.
6941.
7860.
9520.
8073.
8030.
6669.
55*3.
2994.
4533.
.1769E+05
1451. *
14?7. *+
59CO.
5867.
8455.
. 1CC>6E + 05
463t.
?402.
6753.
.IS^SE+C^
37/0.
.1145E+05
7227.
6790.
bO 1 P.
                                                                                   RESIDUAL

                                                                                     7?5.
                                                                                     .279E+04
                                                                                    -.224E+04
                                                                                     400.
                                                                                    -.557E+04 0
                                                                                    -.155E*04
                                                                                     .540E+04
                                                                                     .341E+04
                                                                                    -.315E+04
                                                                                    -.568E»04 0
                                                                                     .625E*04
                                                                                     176.
                                                                                    -525.
                                                                                     .248E*04
                                                                                    -869.
                                                                                    -.412E+04
                                                                                     1B9.
                                                                                     .7C9E*04
                                                                                     871.
                                                                                     '.119E*0«
                                                                                     .202E*04
                                                                                    -675.
                                                                                     £15.
                                                                                    -.191E+04
                                                                                    -579.
                                                                                    -.329E»04
                                                                                    -237.
                                                                                    -.3tOE+04
                                                                                     .177E+0«
                                                                                    -.141E+01
                                                                                    0.0
                                                                                      .0
                                                                                          0.
                                                                             0 .
                                                                                                             0  .
                                                                                                               0
                                                                              c.
                                                                                                        0  .
                                                                                                             0  .

-------
2.  CQMM Equation

         ECUATION  ?
         SWPL VECTOR
            1  40

         ORDINARY LEAST  SQUARES

         VARIABLES...

               COMM
               SERVED
               VACANT
               KIDS
               ACCESS
               JOBCHG
               RECAP1
               C

         MEAN OF DEPENDENT  VARIABLE IS  1848.
INDEPENDENT
VARIABLE
SERVED
VACANT
KIDS
ACCESS
JOBCHG
RECAP1
C
ESTIMATED
COEFFICIENT
4301.81250
-4514.91016
-28.P41201B
6248.92187
-17.C374908
-2.62619495
5J75. 52344
STCNOARD
ERROR
1002.06226
1133.73901
11.1874228
2225.56396
7.56761165
1.75866601
1415.94165
T-
STATISTIC
4.29287338
-3.98231792
-2.57800198
2.80779171
-2.25136948
-1.49328709
3.79642963
KEAN OF
VARIABLE
0.60547256
0.60694695
99.0000000
0.09044975
36.6881866
181.299988
1.00000000
         R-SGUARED  =   0.5734

         DURBIN-UATSON STATISTIC (ADJ. FOR  C GAPS)  =   2.1626

         NUMBER  OF  OBSERVATIONS =   40

         SUP OF  SQUARED RESIDUALS =         .749141E+C8

         STANDARD ERROR OF  THE  REGRESSION =          1506.69

         ESTIMATE OF  VAPIANCE -COVAR IANCE MATRIX OF ESTIMATED  COEFFICIENTS
                                           -0.225E+06  -0.1P3F+04  -0.^68E+03 -0.207E+06
                      0.1P9E+P7  C.177r+04 -P.8f2E*0=   n.l2U + 0«   0.145E + 0? -C.102E + 07
      KiDS-0.211E*04   0.177E+04  0.i:5E+03 -0.731E+04  -0.332E+01   0.804E+00 -0.115E+05
     *<•<•' ,-0.225E*06  -C.862C + 05 -0.731F + P4  0.495E+JT7  -t .435 T + 04  ^Oj.326E+03  0.683E+06
     r?6>Ht-0.183E*04   C.121E+04 -0.3T?r + 01 -Q.435E + 04   0.573E + 02   C.177E + 01 -0.133E + 04
     RF.t-'i -0 ,56fiE+07>   0.145E + C^  O.^flE + OO -0.326E + 071   0.177E + 01   0.309L + 01 -0.420E + 03

-------
       LINE
PRINCETCf.  UNIVERSITY
                                                                    TSF
VERSlCf. OF  AUGUST,
                                          PLOT OF ACTUAL(*>  AfuC  FITTEO<+> VALUES
                                                                           PLOT  OF  RESIDUALS(O)
    in
           ACTUAL
                        FITTED
                                                              RESIDUAL
Ul
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
*4
35
36
37
3 ft
39
1419.
419.4
879.1
892.4
1474.
2778.
2959.
1794.
1206.
388.8
3862.
2470.
2971.
2154.
1118.
1104.
2027.
8732.
701.9
2776.
1462.
1897.
1640.
2771.
77".0
145.4
729. «
722.8
174.?
22't.t.
1197.
171.6
F07.6
1 n71 .
.11T7L+05
245.7
1510.
1875.
609. ?
1472.
-1478.
1162.
890.9
1901.
2632.
2488.
2340.
775.5
744.9
1936.
1693.
3948.
1559.
651.9
2505.
1035.
5986.
3724.
3446.
2373.
83.45
1464.
1812.
1888.
95.07
-823.1
1904.
754.0
3674.
C772.
-2C7.9
2012.
1971.
f- f C 7 .
-28f .2
2443.
3317.
^rc. 9
                         18*0.

-53.3
.190E+04
-283.
1.51
-428.
146.
471.
-546.
430.
-356.
.193E+04
777.
-978.
596.
466.
-.140E+04
992.
,275E*04
-.302E+04 0
-670.
-911.
.181E+04
176.
958.
-.111E+04
50.3
.155E+04
-.118E+04
-580.
-.144E+04
-.157E+04
380 .
-.120E+04
.466E+04
532.
-533.
-.144E+04
-192.
-.137F. + 04
0.0
0
.0
0
0
. 0.
.0
. . 0 .
. 0.
. 0 .
. 0.
.0
. 0 .
.0 .
. 0 .
. 0 .
0 . .
. 0.
0
• • .
. 0 .
. 0 .
.0
. .0 .
. 0 .
.0 . .
.0
0
.0 .
. 0.
0
0
. 0 .
.0 .
. . .
. 0 .
. c .
0
0
0

-------
3.  OFFICE Equation

        LGUATICN
         SMPL  VECTOR
            1   40

         ORC/INtRY  LEAST SQUARES

         VARIABLES...

               OFFICE
               RRHILF.
               OZONED
               PEAK
               DUACPE
               IZONED
               POPDIF
               C

         MEAN  OF DEPENDENT VARIABLE  IS    503.1760
INDEPENDENT
VARIABLE
RRHILE
OZPNED
PEAK
OUACRE
IZONED
POPDIF
C
ESTIMATED
COEFFICIENT
719.511719
33478.0352
36.6961060
0.33335096
-1790.61548
572.025391
-459.871338
STflNDARD
ERROR
137.541382
6491.65156
10.2877779
0.10663390
772.712891
271.222656
147.449326
T-
STATISTIC
5.23123741
5.15693092
3.56696129
3.12612534
-2.31730938
2.10906124
-3.11884308
MEAN OF
VARIABLE
0.46907347
0.00529748
5.88499546
692.524902
0.06514716
0.20634961
1.00000000
         R-SQUARED =  0.7017

         DUPBIN-UATSON STATISTIC  (ADJ. FOR   0  GAPS)  =  2.1360

         NUMBER OF OBSERVATIONS =    40

         SUM OF SOUAPtD RESIDUALS =          .293272E+07

         STANDARD ERROR OF THE REGRESSION  =          29fl.lll

         ESTIMATE OF VARIANCE-COVARIANCE MATRIX  OF ESTIKATEC COEFFICIENTS
     RRMTIE0.169E + 05  0.196E»06  0.148E + 03  -0.361E + 01 -0.40U + 05  0.182E + 04 -0.603E+0«
     -^v£,>0.19tE + 06  0.'»21E + Ofl -0.903E + 04  -0.31CE + 02 -C.184F + 07  0.267E + 06 -0.175E + 06
          Q.14PE+03 -O.S03E+04  0.106t*03   0.682E-01 -0.131E+04  0.170E+03 -0.fcl2E+03
                    -0.316E*02  0.f>82E-01   0.114E-P1  C.15CF+02  C.525E»01 -0.928E + 01
.^10 IE
                                O . 1 7 O f. » 0 3
                                                   Ci  r .597'. »0fc -0.1S7r»OS -n.
                                            O . T ^ E -r C 1 - O . 1 •=• 7 F » n e. ^O.7?eF.O5 -O.a3BE»Of>

-------
    LINE
                       PRINCFTOM UNIVERSITY
                                                               TSP
                                                                                        OF AUGUST, 15(9
                                                                                                                   PAGE
                                                                                                                           12
                                     PLOT  OF  ACTUAL<*>  AND FITTEDC+) VALUES
                                                                                                 PLOT  OF  RESIDUALS(O)
ID
       ACTUAL
                    FITTED
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
3f.
37
38
39
40
552.7
411.8
1687.
86.40
337.3
339.7
1169.
288.5
555.1
147.9
687.5
589.3
390.4
382.4
935.9
403.4
822.3
1999.
995.9
230.9
199.0
177.8
34.00
567.3
94.70
65.10
251.9
144.6
40.60
P.87.5
577.2
33.90
T85.5
?96.f,
2111.
94.60
"H9.7
380.1
98.60
185.2
1046.
-23.02
1474.
168.3
298.5
404.3
855.9
759.1
939.3
261.7
553.9
609. f
711.8
451.1
673.5
578.3
234.2
1626.
923.7
479.9
64.30
18.59
215.1
325.3
428.6
50.92
345.6
441.4
-208.6
896.8
b45.7
115.5
41^.8
431.2
1446.
13?. 6
581.4
-19.17
12?. 3
447.5
                                                   *    *
RESIDUAL

 -493.
  435.
  212.
 -ei.9
  38.fl
 -64.6
  713.
 -471.
 -384.
 -114.
  134.
 -20.3
 -321.
 -68.7
  262.
 -17b.
  588.
  373.
  72.2
 -249.
  135.
  159.
 -181.
  242.
 -334.
  14.2
 -53.7
 -297.
  249.
 -9.34
 -269.
 -81.6
 -34.3
 -135.
  666.
 -38.0
 -91.7
  399.
 -?3.7
 -262.
                                                                                                        0.0
                                                                                                        0.
                                                                                                         .0
                                                                                                        0.
                                                                                                               .0
                                                                                                   0.
                                                                                                         .0
                                                                                                   0.
                                                                                                              0.
                                                                                                              0.
                                                                                                        0.
                                                                                                        0.

-------
4.  MANF Equation

         EQUATION  4
         SKPL  VICTOR
            1   4C

         ORDINARY LEAST SQUARES

         VARIABLES...
               RRMILE
               VACAM
               P.ANJOB
               OZONED
               VACOFF
               AIRPPT
               C

         MEAN OF DEPENDENT VARIABLE  IS  1B05.8608
INDEPENDENT
VARIABLE
RRHILE
VACANT
MANJCB
OZONEP
VACOFF
AIPPRT
C
ESTIMATED
COEFFICIENT
2590.97559
-3890.12744
2.97811413
65922.3750
234.675339
22.7014465
611 .240967
STANDARD
ERROR
697.218994
1287. (5332
1.40823187
33632. S289
93.959^58
12.R664851
1068.74121
T-
STATISTIC
3.71615696
-3.02109814
2.11463928
2.55473232
2.49761772
1.76164722
O.E7192606
MEAN OF
VARIABLE
0.469D7347
0.60694695
204.263763
0.00529748
3.16749668
23.5024567
i.oooooooa
         R-SQUARED =  P.6301

         DURBIN-UATSON STATISTIC  (ADJ. FOR  0 GAPS)  =   2.4249

         NUMBER OF OBSERVATIONS =    4P

         SUM OF SQUAPEC RESIDUALS  =          .8e0977E»08

         STANDARD ERROR OF  THE  REGRESSION =          1605.Pt

         ESTIMATE OF VAPIANCE-CCVARIA\CE MATRIX OF  LSTIPATEC  COEFFICIENTS
           £p.HTLE    \Jf\(.f\KJT    AlAA/lOfi     0*<3A/£p    VAtafp     ftrkPRT
         r-0.486E*06  0.1f.9E + 06  -0.112E + 03  0.213E + 07 -0.938E + 04  -0.267E + 03 -0.283E»06
            It^r+Ob  0.16br+07  0.4?3£+03 -P.171E+07   0.660E+04  -0.'(75E*04 -0.107E+07
            112E + 03  0.«33E' + 03  0.19PE + 01 -O.llH + Of   0.1[59F + 02  -0.332E + 01 -0.529E + 03
          C.213E1+07 -G.171E+07  -0.111E+05  O.llJL+10   0.743E+06   0.498E+05 -0.721L+07
    V-V-f •- -0 .
                                                                   0.1C,BE»03  -D-

-------
    LINE
                                                                                         'F  AUGUST*
                                                                                                                           15
                                      PLOT  CF  4CTUAL(*>  AI-.C FJTTt;0(+) VALUES
                                                                                                    PLCT OF  f
ID
       ACTUAL
                    FITTED
to
1
2
7
1
5
6
7
8
9
10
11
12
13
11
15
16
n
18
19
20
21
22
23
21
25
26
?1
28
29
70
31
32
73
7<4
7.5
3^
77
Zf
39
10
2339.
191 .1
1209.
17«5.
1831.
20Q2.
6213.
1050.
511.2
267.7
736.8
672.7
1208.
715.^
1071.
3265.
553.1
.1011E+05
1126.
1859.
388.1
1271.
2365.
3106.
359.2
fO.fO
12».8
1011.
83.f.0
621. «•
1151.
121.8
1711.
517.7
.10f 6E+05
i7^.b
"512.
970.2
110.8
17.10
25M.
-1520.
1916.
1877.
1251.
2193.
5900.
2057.
1677.
976.5
525.2
82C.9
1392.
2 1 ? 7 .
96^.0
10f 2.
-326.2
671 1.
1 8 r> 0 .
2799.
531.1
&36.1
-28.29
2770.
871.7
-1271.
1126.
706.7
-53.61
?733.
330?.
-163.1
1791.
3219.
f 705.
-571.9
7f'4l.
1 1 f> 1 .
171.8
1 7 ?. » .
                                                                                            01
                                                                                       REJ1CUAL

                                                                                        -192.
                                                                                         .201ET
                                                                                        -707.
                                                                                        -P8.2
                                                                                         580.
                                                                                        -101.
                                                                                         313.
                                                                                        -.101E+01
                                                                                        -.116E+01
                                                                                        -709.
                                                                                     -118.
                                                                                     -183.
                                                                                     -.168E+01
                                                                                      106.
                                                                                     -S17.
                                                                                      £fc2.
                                                                                      .310E»01
                                                                                     -.312E+01
                                                                                      .206E+01
                                                                                     -116 .
                                                                                      125.
                                                                                      .239E+01
                                                                                      f.35.
                                                                                     -E15.
                                                                                      .173E+01
                                                                                     -S97.
                                                                                      338.
                                                                                      137.
                                                                                     -.211E+01
                                                                                     -.215E+01
                                                                                      288.
                                                                                     -180.
                                                                                     -.273E+01
                                                                                      .395E+01
                                                                                      907.
                                                                                      ??P.
                                                                                     -221.
                                                                                     -71.0
                                                                                     -.128E+01
                                                                                                     .0
                                                                                                         0.0
                                                                                                          0
                                                                                                          0
                                                                                                          .0
                                                                                                             0.
                                                                                                     .p

-------
5.  WHOLE Equation
         EQUATION  ?.
         ***********
              VECTOR
            1   40

         ORDINARY LEAST SQUARES

         VARIABLES...

               WHOLE
               OKI VE
               VACOFF
               CFFJOP
               EMPOP
               UNEMP
               PERCAP
               C

         MEAN  OF DEPENDENT VARIABLE  IS    477.0664


£




INDEPENDENT
VARIABLE
DRIVE
VACOFF
OFFJOB
EMPOP
ur:EMP
PERCAP
C
ESTIMATED
COEFFICIENT
166.454239
86.2756805
3.34034920
604?. 78906
6948.57422
-0.10651082
-1911.15161
STANDARD
ERROR
20.7087250
19.°433136
0.9781018?
1535.17920
2431.63574
0. "4552276
668. f 82324
T-
STATISTIC
8.03787899
4.32604504
3.41513348
3.93621063
2.85757160
-2.33972645
-2.85723114
MEAN OF
VARIABLE
1. 72454357
3.16749668
75.1841431
0.37607706
0.04694974
9726.22266
1.00000000
 R-SGUARED =   0.7597

 DURBIN-UATSON  STATISTIC (ADJ. FOR  0 GAPS)  =   2.2514

 NUMBER OF OPSERVATIONS  =   40

 SUM OF SQUARED  RESIDUALS =         .33509f,E+07

 STANDARD ERROR  OF  THE  REGRESSION =          Mfl.6£0

 ESTIMATE OF  VARIANCE-COVARIANCE MATRIX OF ESTIMATED  COEFFICIENTS
    Jflve     >jhCoFF      OFFjofi     2WPOP     v/ufwP
H 0.429E + C3   0.?,29E. + 02   0.6R5E + 00 -0.709E + 02   0.73SE + 04
 L C.685E + 00   0.4?4T*01   0.9b7E+00  0.22f>e+0.T -C.3?4f + 03
 '-0.709E+02  -0.414T+04   0.226F+03  0.?3fc,L+07  0.111L+07
                                                                 -0.209E + 00   C.709E + 03
                                                                  0.3S4E-01   0.434E + 03
                                                                  0.118E-01  -0.270E + 03
                                                                 -C.24CE+02  -0.711E+06
           » - i o ** r -*-o
                            .03  — o . zr
                                       - .:> ,1 -n_viiK»n»-
                                                                  0.1S2E-r2 -r.780E»06

-------
                       PPINCETO' UNIVERSITY
isr
VEPSKN CF AUGUST,
                                                                                                                    PfiGE
                                                                                                                           1R
                                      PLOT  OF  aCTUAL(')  AND FTTTEDC+) VALIES
                                  PLOT  OF  RESIDUALS(O)
ID
       ACTUAL
                    FITTED
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
??.
24
25
26
27
28
29
30
31
32
33
34
35
36
37
36
39
40
114.2
21.20
189.3
74.30
558.1
68.30
2535.
182.3
937.6
133.3
312.5
667.8
205.0
2364.
333.7
432.9
.0
551.5
425.0
1871.
346.8
1399.
300.6
760.6
646.0
14.50
218.3
361.4
20.50
228.4
570.4
27.40
671.0
1?6 .5
256.5
9.300
732.1
175.0
10.90
If 1.1
405.8
-215.0
233.9
36R.5
413.9
-109.7
2322.
28.37
1171.
545.6
-96.10
504.3
643.8
1993.
59.36
587.0
13P.5
797.1
669.4
1095.
402.6
77P.1
568.6
966.8
303.9
85.42
772.6
597.9
119.0
44^.5
14^.6
-5.30C
907.3
705.3
6 4 3 . fe
264.0
531.?
194.4
-340.2
237.5
                                                                                    RESIDUAL
                                                                                      236.
                                                                                     -44. 6
                                                                                     -294.
                                                                                      144.
                                                                                      198.
                                                                                      213.
                                                                                      154.
                                                                                     -233.
                                                                                     -412.
                                                                                      409.
                                                                                      164.
                                                                                     -439.
                                                                                      371 .
                                                                                      274.
                                                                                     -154.
                                                                                     -139.
                                                                                     -^46.
                                                                                     -244.
                                                                                      776.
                                                                                     -55. F.
                                                                                      620.
                                                                                     -26U.
                                                                                     -206.
                                                                                      342.
                                                                                     -70.9
                                                                                     -154.
                                                                                     -236.
                                                                                     -Sf.5
                                                                                     -221.
                                                                                      425.
                                                                                      33.3
                                                                                     -236.
                                                                                     -1*9.
                                                                                     -307.
                                                                                     -255.
                                                                                      201.
                                                                                     -19.4
                                                                                      351.
                                                                                     -56.4
                                                                                                       0.0
                                        C.
                                 0  .
                                   0
                                   .  0
                                        0.
                                       0 .
                                       0 .
                                   .0
                                 0  .
                                        0.
                                         •
                                        0.
                                            0 .
                                             0.
                                             0.
                                            0 .
                                             0.

-------
6.  HIWAYS Equation

        EQUATION
SMPL VECTOR
   1  40

ORDINARY LEAST SQUARES

VARIABLES...

      HlUAYS
      RRMILE
      LAND
      GOVT
      RECAP
      CROSS
      C

MEAN OF DEPENDENT VARIABLE  IS
                                          43. 3924
INDEPENDENT
VARIABLE
RRKILE
LAND
GOVT
RECAP
CROSS
C
ESTIMATED
COEFFICIENT
35.2186584
22.9822540
-0.11664993
0.53785098
25.9384308
11.3462658
STANDARD
ERROR
9.71726036
7.10037518
O.P4293521
0.23814219
16.6402893
8.29916477
T-
STATISTIC
3.6243*006
3.23676586
-2.71688175
2.25852776
1.55877209
1.36715698
MEAN OF
VARIABLE
0.469073*7
0.53672218
71.3199310
tO»2699714
0.23038459
1.00000000
         R-SQUARED =  0.5226

         DURBIN-UATSON STATISTIC (ADJ. FOR  0 GAPS) =  2.1316

         NUMBER  OF OBSERVATIONS =   40

         SUM  OF  SQUARED RESIDUALS =         18421.3

         STANDARD ERROR OF THE REGRESSION =         23.2767

         ESTIMATE OF VARIANCE-COVARIANCE MATRIX OF E«Tlr"ATED COEFFICIENTS
           \-f.HZLE.     L/WD       Govr       RtcfiP     CROSS      (ionsr*irr)
          0.944E*02 -0.163E+01 -0.966E-02 -0.552E-01 -0.205F*02 -0.374E*02
    .Lft'JO -0.163E+01  0.504E+0? -0.633E-01  0.34flE-01 -0.889E+01 -0.201E + 02
     iovr-0.966E-02 -0.633F-01  0.184E-02  0.1"1E-01  O.fc^r-02 -0.946f-01
    Ht^P -0.552E-01  0.348E-01  0.141E-04  0.567E-01  0.593E + 00 -0.714E + 00
     CFos$-0.2D?E*02 -0.869E+01  0.655E-02  0.593E»00  0.277E+03 -0.560E+02
                    -0.201E»02 -0.946E-01 -0.71tE«00 -0.560E + 02  0.689E»02

-------
    LINT
                       PRINCCTON UNIVERSITY
                                                               TbP
                                                                                VERSION  OF  AUGUST,  1969
                                                                                                                    >AGE   21
                                      PLOT  Of  ACTUAL(*>  AND
                                                                      VALUES
                                                                                                 PLCT  OF RESlLlALS(C)
ID
       ACTUAL
                    FITTED
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
lc
15
?0
21
22
23
24
25
?S
27
2P.
29
30
31
32
33
T A
-i c;
36
37
38
39
40
70.90
30.10
26.80
15.40
26.70
92.20
66.60
"7.00
84.00
55.00
57.50
f. 2 . 1 0
75.60
P3.50
C6.60
42.30
25.30
170.0
3.000
15.20
31.40
45. fO
13.40
80.20
30.30
4.1PC
4P.°0
30.80
32. TO
58.50
4A.30
5 . 3, 1 0
34.50
P * . B 0
48.20
14.00
36. 5C
27.70
1 5 . '. 0
P .4fO
38.94
25.83
55.48
-6.09?
33.33
62.04
68.37
41. P5
£2.55
36.26
32.46
61.43
hj.f'?.
49.46
53.88
54.73
37.48
111.9
42.40
46.81
53.30
2H.40
12.03
60.59
48.ft4
16.45
43.65
38.57
2.frM
43. P5
"3. 07
3?. 49
52.55
54.74
4P.59
2^ .59
37.54
42. P 4
-4.V24
37.61
                                                                                   RLSIOUAL

                                                                                      32.0
                                                                                      4.27
                                                                                     -28.7
                                                                                      21.5
                                                                                     -6.3T
                                                                                      30.2
                                                                                     -1.57
                                                                                      5.15
                                                                                      1 .01
                                                                                      18.7
                                                                                      25.0
                                                                                     -<>.33
                                                                                     -12.2
                                                                                      34.0
                                                                                      2.72
                                                                                     -12.4
                                                                                     -12.2
      O.C
0 .     •    •
            0
 .   0 .    .
            .0
      0.
 .     .0   .
       0
       .   0.
 .     .    .0
 .  0  .    .
 .  0  •    .
                                                                                               0
Z1.6
21.9
17.2
1.37
19. t
18.5
12.3
= .21
7.77
30.1
14.7
5.23
?7.2
•18. C
25.9
.395
11.6
1.04
•15.1
20.8
•25.2
0 .
0
•
•
•
.0
. 0
•
0
•
•
•
0.
.0
0 .
•
. C
•
. 0
•
0.
•
•
•
0
•
•
•
.0
•
•
•
.0
•
•
•
0
•
0.
•
•
•
•
•
0.
•
n
•
•
•
•
. 0
0 .
•
•
•
•
•
•
•
•
0
•

-------
7.  EDUC Equation

        EQUATION   7
         S"FL  VECTO"
            1   4P

         ORDINARY  LEAST  SQUARES

         VARIABLES...

               EOUC
               POOP
               SERVED
               LAND
               LIMITS
               PERCH6
               COST
               C

         MEAN  OF  DEPENDENT VARIABLE IS   662.7135
INDEPENDENT
VARIABLE
POOR
SERVED
LAND
LI-ITS
PE"CHG
COST
C
ESTIMATED
COEFFICIENT
-3185.58545
722.979492
437. 52£ 123
95.1653137
-405.2014U,
-1.41497135
428.885010
STANDARD
EFRCR
670.197266
144.C78339
83.0681152
30.2513428
180.390137
0.77642828
168.597427
T-
STATISTIC
-4.75320530
5.01796055
5.26707649
3.14562062
-2.24625015
-1.82241058
2.54384041
KEAr, OF
VARIABLE
0.15904957
0.60547256
0.53672218
2.47499943
0.25504965
45.4195862
1.00000000
         ^-SQUARED =  0.7545

         DURBIN-yATSON STATISTIC (ADJ. FOP  0 GAPS) -  2.0024

         NUMBER  OF OBSERVATIONS =   40

         SUM  OF  SQUARED RESIDUALS =         .192734E+07

         STAMDAPD  ERROR OF THE REGRESSION =         241.669

         ESTIMATE  OF V AP I Ah,CE-COVAR IANCE MATRIX OF ESTIMATED COEFFICIENTS
           POOfiv       SERVEQ     LflWO      LTMTTS    PERCHt      COST
          0.445E + C6  0.163L + 0?  Q.cr;56E + 04 -0.34°i: + 04  0.153E+Ob  -0.178E + 03  -C.709E + 05
    S£Rveo 0.153E + 05  0.208r + Cc; -C.933E*03 -C.108E + 03 -0.371E + 04  -0.717E + 01  -0.130E + 05
          0.55f-E + 04 -0.9?3F+Ci  0.69C.E+C4  0.645E + 03  0.1?ff + 04  -C.254E + 02  -C.482E + 04
                *94 -C.lCct+0:  O.S45E + 03  0.915r*0.' -0.f-18E + 03   C.227E + 01  -0.194E*04

-------
LIME
                   PRINCETON  UNIVERSITY
                                                            TSF
                                                            VERSICK  OF  AUGUST»  Iv/t9
                                                                                                PAGE
                                                                                                       24
                                  PLOT CF ACTUALi*)  AND  FITTEC(+>  VALUES
                                                                             PLOT CF RESICUALS(O)
   ACTUAL
FITTED
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
3C
31
32
33
34
3":
36
37
3P
39
4C
866.0
563.4
2135.
454.0
524.9
709.5
1554.
635.8
724.0
603.9
984.8
f67.fl
483.3
447.4
350.8
499.?
982.5
1517.
368.6
1279.
286.9
707.3
f55.0
1374.
359.2
69.70
346. C
715.5
147.4
11C6.
662.5
f: 1 . 1 0
£•43. 7
1037.
4 ri 7 . 6
95.50
615.5
607.9
11. 6P
160.1
751.0
443.0
1805.
529.0
816.1
965.0
1194.
787.6
725.6
425.9
485.5
760.6
f 15.8
214.9
576.0
505.3
llr.5.
1283.
424.3
1075.
4ie.5
411.5
407.2
1311.
580.5
-168.1
434.0
643.0
230.3
7BP.4
700.5
329.3
72?. 7
9f I .4
582.1
-50.90
7^4.7
11"6.
14C.1
3 c ' . 7
RESIDUAL

  115.
  120.
  229.
 -75.0
 -291 .
 -255.
  759.
 -152.
 -1.65
  178.
  499.
 -92.8
 -132.
  233.
 -2?5.
 -5.75
 -172.
  233.
 -55.7
  204.
 -129.
  29f. .
                                                                                  b3.0
                                                                                 -221.
                                                                                  238.
                                                                                 -f-b.O
                                                                                 -127.
                                                                                 -82.9
                                                                                  325.
                                                                                 -.'fl.O
                                                                                 -248.
                                                                                 -80.0
                                                                                  cO. 1
                                                                                 -125.
                                                                                  I4f. .
                                                                                 -149.
                                                                                 -C14
                                                                                                        0.0
                                                                                                         .  0   .
                                                                                                         .   0  .
                                                                                                       0  .
                                                                                                      0   .
                                                                                                        0.
                                                                                                         .0
                    .   0  .
                    .0   .
                    .   0  .
                    •     *
                        0.
                   ).
                    .   0  .
                    •     •
                    •  0   .
                    •     •
                    .0
                    •     •
                    .  0
                                                                                             .0
                                                                                           0  .

-------
8.  REC Equation

        EQUATION   H
        SMPL  VECTOR
            I   40

        ORDINARY  LEAST  SQUARES

        VARIABLES...

               REC
               RRWILE
               CAPAC2
               PERCHG
               OFFJOB
               VACHSE
               IZONED
               C

        MEAN  OF  DEPENDENT  VARIABLE IS   207.8000
INDEPENDENT
VARIABLE
RRWILE
CAPAC2
PERCH G
OFF JOB
VflCHSE
I ZONED
C
ESTIMATED
COEFFICIENT
270,401123
2.63108635
322.055176
0.97423780
-3225.57251
-557.780029
59.5172862
STANDARD
ERROR
56.?478333
1.06722641
91.3438721
0.35317051
1262.37156
294.C32471
71.8569489
T-
STATISTIC
4.8IJ7315B3
2.46534920
3,52574444
2.7SB54778
-2.55516815
-1.89700127
0.82B27461
I«E*N OF
VARIABLE
0,46907347
10.8374815
0*255114965
75.18*1431
0.03909981
4.06514716
1.00000000
         R-SQUARED  =   0.6471

         DURBIN-VATSON STATISTIC (ADO. FOR  0 GAPS) =  1.7751

         NUMBER  OF  OBSERVATIONS =   40

         SUM  OF  SQUARED "ESIDUALS =         498049.

         STANDARD ERROR OF THE REGRESSION =         122.851

         ESTIMATE OF  VARIAKCE-COVARIANCE MATRIX OF  ESTIMATED COEFFICIENTS
           RRNILE     cfwca.     PERCH &      OFFJOB    Vflcnse     rio/veo
        £ 0.316E+04  0.105E+01  0.628E+03 -0.475E+01  0.143E+05 -0.521E*04 -0.152E+04
          0.105E + 01  O.lltF + 01 -0.390E*01 -0.970L-02  0.128E»03 -0.266E + 02 -0.144E + 02
         i 0.628E + 03 -0.390E + 01  O.RT4E + 04  0.741E + 01 -O.^6?t + 04 -C.277E+04 -C.237E + 04
         -0 .475E + C1 -0.970E-02  0.741E + C1  C.12r.E + CO -C.£t?E + 02  0.229E + 02 -0.791E + 01

-------
    LUF
           10
                        PRI\CtTCN UMVFRSITY
                                                                 TSF-
                                                                                                      1'jfc1?
                                                                                                                       PAGE    27
                                      PLOT  CF  tCTUALC) £ND  FTTTEH
                                                                        VALUES
                                                                               PLOT  OF RESICUALS(O)
ID
       ACTUAL
FITTED
RESIDUAL
1
?
3

-------
              9.  OTHER Equation

                       ECUATIOM
                       SMPL VECTOR
                          1  40

                       ORDINARY LEAST SGUAKES

                       VARIABLES...

                             CTHEK
                             RRMILE
                             PFRCAP
                             VACHSE
                             PERCHG
                             INTDtfJ
                             R Z 0', f 0
                             C

                       MEAN OF DEPENDENT  VARIABLE. IS
                                151 .23136
3=.
 I
oo
                   INDEPENDENT
                     VARIABLE
                  ESTIMATED
                  COF.KFICIf NT
               1134.4^140

                 -0.30158252

             -12175.2109

                64A.Q79980

               -2^5.404785
                                               STANDARD
                                               EPF.CR

                                             180.363739
                                                                311 .26b602
 PERCAF

 VACHSE

 PERChG

 I \TOEf.'

 RZONED

 r.                  ?08f. 53931

    R-SQUARED =   0.6721

    DURBIi\-WATSON'  STATISTIC (ATJ. FOR   0  GAPS)  =  2.3041

    NUMBER OF OBSLKVATIONS =   40

    SUM OF SQUARE L  RESIDUALS =          .5244^1f.+07

    STANDARD ERROR  OF  THE REGRESSION =          39p.(66

    ESTIMATE OF  V AR I ANCE-CO VAR I ANCE '•'ATRIX  OF  E^TIMATtC COt
T-
STiTISTIC
6.26568620
-4.77312660
-2.51952839
1.50167999
-1.4P177052
1.46485615
4.45826244
MEAN OF
VARIABLE
0.46907347
9726.22266
0.03909981
0.25504965
0.18093479
0.33849955
1.00000000
^ri-E  0.325E
          -C.104E + 01   O.l^r'E + Ce  0.122r
           0.400'"-02   0.114F+n3 -f.
C.158E. + 06  0.114li + i;;   C.2itf. + 08 -C.
0.1??E*05 -C.lnlE*C2  -
                                                                      C.H3F + 01
                                                                      0.414t+ni
FFICIEf'TS
 R 7 OWED   f'ouim/"^
 0.122E+05  -0.200E+05
-0.368E+01  -0.398E+0?
 0.101L+06  -C.213E+07
 O.filJE+03  C.789E+05
                                                                                 0.101E.-»-CE -C.t, 0 *T- + Q =
                                                                                 C_/. Pt •C1" — C . !:•*•» • ft »

-------
       LIKE
              11
                          PRINCETON UNIVERSITY
                                                                  TSP
VERSION' OF AUGUST,
                                                                                                                       PAGE
                                                                                                                              30
                                        PLOT OF ACTUAL(') AND  FITTED(+)  VALUES
                                                                                                   PLOT CF RESIDUALS(O)
   ID
          ACTUAL
                       FITTED
to
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
85.70
50.00
353.2
486.1
83.70
548.6
578.2
308.8
603.7
403.4
76?. 6
1124.
304.4
477.8
866. 2
377.0
434.4
4044.
268.7
558.5
160.6
532.6
82.20
288.2
389.8
33.10
230.7
274.6
41.10
642.2
414.0
38.30
282.9
146.9
316.7
b8.30
1016.
309.1
45.40
146.3
279.8
-521.3
298.4
197.7
1.213
40.70
730.6
-26.88
811.8
869.2
64=. 2
853.9
453.3
1034.
627.2
720.1
625.0
2803.
761.4
889.6
370.6
257.4
-118.2
236.4
922.3
228.6
451.3
409.3
83.55
234.0
712.9
16.05
253.9
754.9
274.7
.8671
631.5
182.8
-321.9
474.^
                                            *    +
                                               »*
   RESIDUAL

    -154.
     571.
     5«.8
     288.
     62.5
     508.
    -152.
     336.
    -208.
    -486.
     118.
     270.
                                                                                                        0.0


                                                                                                         .0
                                                                                                         •
                                                                                                         .0
    -557.
     239.
    -3t3.
    -191.
     .124E+04
    -493.
    -331.
    -210.
     ?75.
     200.

    -532.
    -195.
    -221.
    -I?1).
    -42.4
     108.
    -299.
     22.2
     29.0
    -608.
     42.0
     r.7.4
     3H5.
     126.
     3(7.
    -?.2fi.
                                                                                                     .0
.  0
                                                                                                          .0
                                                                                                         C.
                                                                                                     .0

-------
  10.  RY1  Equation

  PREDICTIVE FQUATION REGRESSION

  FILE   PREDICT  (CREATIO^ DATE =  02/17/78)



  DEPENDENT VARIABLE..    PY1

  VARIAELE(S) ENTERED ON STEP Nl'itER   6..
                                                                 ("L-/17/7P
                             U L T  I
                       HOSCHG
  MULTIPLE R
  R SQUARE
  ADJUSTED P SQUARE
  STANDARD ERROR
0 .9052F.
0.81953
0.75581
0.28002
ANALYSIS
RCGRESSlO
RESIDUAL
VARIABLE

KIDS
CAPCHG
SERVED
OZONED
STAY
HOSCHG
(CONSTANT)
                    VARIABLES  IK  THE  EQUATION  	

                     B          BETA      STL IRKCR  b
 0.13729120-01
-0.91820930-03
 0 .7842406
  22.70408
 -2.980871
-0.1814667
  4.531282
                                O.f.2662
                               -0.71293
                                0.36504
                                0.37C74
                               -0.44621
                               -0.27103
                       0.00246
                       0.00017
                       0.24114
                       7.17613
                       0.91678
                       o.oai-u
        31.204
        21.16u
        10.57?
        10.010
        10.t72
         ".957
ro
o
l l '. l l r
b^ GL'<'
t .
17 .
VAKIABLt
'. U.'-C^E
V A C h S F
1NCOL
VACOFF
UM-'iF
OFFJOC
UUJOL
i1 A i1. o c b
oor s
i\ pf^nh
FOCR
VALUf
<*OC> £
UMV
KELJOB
RELUW
FELOFF
R E L C" A f-1
GCVT
LAf- ;•
POP JEN
DISC6D
R R M I L t"
Accrss
INTL.LN
TRANS
A I H F ' R T
VACANT
RZON'ED
CZONED
I70NFC
LIKITS
POLICY
TLI'-'IT
TPOLIC

CF r, TUAFtS
C . C 5 3 a 2
1 .33304
F.ETA I'
-P. 26276
-0.01227
-0.090?"
-0.11146
0.1167?
-0.0C793
0.0912P
0 .076?4
0 .15144
-0.02842
-0.03401
C . 0 i c (. Z
-P.0?777
0.11174
0.02776
C .1^166
0.13P"9
0 .1 J5 7>
r. . c r 5 ? P
0.06^93
0 .13291
-0.0234"
-0.05721
-0.2250-:
-n. 14391
0.12^7
0.0997"
0.00995
-O.OOOCb
-0.195^1
- 0 . C 9 c (• 0
0 .CPf 41
-O.lil60
C.C2211:

;• 1 1 r.
l .
c r r T r r 1 1-> i
. i 1 t* 1 i 1* 1 " •
c f. F T I f. L
0.09311-
-n «43oit
-O.r2i>12
-0.1F.171
-0.1^912
0 .210JJ3
-0.1^61?
0 .Ifc660
0 .17644
0 .^6022
-O.C4142
- 0 . ! 7 4 f, t
O.C614:.
-J.[HC33
C.2f>0i?0
0 . 0 4 d 6 5
0.2S305
0.25251
0 .244,04
n . p i c 6 1
0.12410
J. 21467
-0.04488
-0.1097fr
-0.17366
-0.30 OOP
0 .T5751
0 ..2201
0 . (.' 2 0 6 7
-O.OOC1C
-0.',-;752
- 1 . r r 4 1 r
0.0 19b3
-0 .cb743
r.cMOi


0"9,-

C . f «. i' 7 7
C . 4 '"' 4 5 0
C.VG4R3
0.73HQ6
0 .541- 15
0.5&404
0.8 C. 5 ? P
C .io'(l9
0 . 5 7 7 o 1
L .5.52!: 1
0 . 3 o 7, 3 7
0.86076
C . 7 0 c. 0 L'
0.1-075
0 .61096
0.55355
C. 60289
0.59996
C.f "631
C .7C2f 1
0 .56f"2f;
0. till 32
C.66 1"2
0 .f.£u<47
0 .67071
0 .76467
0'.70972
0.89115
0.77921
0.78691
0.6016 I
0 . 6 ; c 2 1
0.97200
0 . H b 8 3 0
0.9t776
L ! ' 1 1
L ! i T 1
p
1 ? . r t. L V

f
(.'.110
? . 6 <•;
0 . 0 I ?
0.5"^
0 . fa f. 1
0.7 5>
P . 2 r 5
O."r7
0.30J
1 . 1 h 2
0 . f ? 7
C.C^O
0 •C'' 7
C . C 3 3
1 .Ifec
0 . 0 3 P
1 . 3 ° I
1.09C
1.514
0 . 0 C 2
0.23C
l.Ol"
0.032
0 . 1 " 5
3.71C
1.TP3
1 . 1 7, 7
0.829
O.OC7
o.ccr
2.345
C .6?P
P .C P6
1.411
0.042

-------
11.   RY2 Equation

PRCDICllVE EQUATION REGRESSION

FILE   PREDICT  (CREATION DATE  =  02/17/7b)

******************  *  *  *  I

DEPENDENT VARIABLE..    RY2

VARIABLE(S) ENTERED ON STEP  MJCRER   5..
                                                " U L T I F L b
                                            PEAK
                                                                                        /1 7 / 7 -
                                                                              L: ?T
                                                                              L I', T
MULTIPLE R
R SQUARE
ADJUSTED R SQUARE
STANDARD ERRO-R
                    0.91390
                    0.83521
                    0.78943
                    0.33457
A'JALYSIS OF VARIA\CF
PEGRESSIOh
RESIDUAL
VARIABLE
HSECHG
PCOR
OZONED
RRMILE
PEAK
(CONSTANT)
B
4.376418
-5.479691
-34.05850
0.7322320
0.3645832D-01
0.2180154
                  VARIABLES  IN  THE  EQUATION 	

                              BETA      STD ERROR
                              0.70395
                             -0.48705
                             -0.43228
                              0.35431
                              0.26902
                                            0.68429
                                            1.25251
                                            8.21178
                                            0.20536
                                            0.01399
        40.903
        19.140
        17.202
        12.714
         6.795
OF SOI'
rj.
18.
VARIABLE
DUACPE
VACHSE
INCOME
VACOFF
UN'EMP
OFFJ06
UUJOf:
MA" JOB
JCLS
NONHH
VALUE
ROOMS
KIDS
UNIV
RELJOB
RELUU
RELOFF
RELHAN
GCVT
LAND
POPDEN
DISCED
ACCESS
INTPEN
TRANS
AIPFRT
VACANT
RZONED
CZONED
IZONED
LIMITS
POLICY
T L I * I T
TPCLIC
DRIVE
CF SQUARES
1 0 . _> 1 1 .; I
2.014P2
PETA I*
-0.12*56
-O.OM 43
-o.ieror
-0.01948
0.09182
-0.09540
-0.11675
- 0 . 1 1 B 1 8
-0.13060
0.0897
  0.70678
  T.87553
  O.fc°123
                                                                                                                  0.59812
                                                                                                                  t.56073
                                                                                                                  0.52848
                                                                                                                  0.63635
                                                                                                                  0.81047
                                                                                                                  0.95514
                                                                                                                  0.94348
                                                                                                                  0.96716
                                                                                                                  0.90357
                                                                                                                  0.76183
                                                                                                                  0.82320
                                                                                                                  0.75030
                                                                                                                  0.65692
                                                                                                                  0.80816
                                                                                                                  0.h45:'5
                                                                                                                  0.83137
                                                                                                                  0.70522
                                                                                                                  0.6=524
                                                                                                                  0 .86b95

                                                                                                                  C .80256
                                                                                                                  D .80549
                                                                                                                  O.S'-Ull
                                                                                                                  0 . 8 1 1 " 5
                                                                                                                  0.72629
1.4F9
p. OPC
2.535
0 . 0 3 U

O.P64
1.031
1 .11C
1.530
0.50*
0.005
0.705
3.380
0.5] H
0.432
0.224
0.3&0
0.43C
1.009
O.lCfe
1.312
0.150
1.796
0.012
O.lPt
0.257
0.475
1.373
0.486
0.269
0.336
1.714
0.007

1 .014

-------
12.  CY1 Equation

PREDICTIVE EQUATION REGRESSION

FILE   PREDICT  (CREATION DATE =  r.2/1i/7t)



DEPENDENT VARIABLE..    CY1

VARIABLE(S) ENTERED ON STEP NUMBER   5..     TLIMT
                                                  U  L  T  I  P  L
                                                                                        -.••••*. 1 '• I.L
                                                                                        (.,•• •Sill \
                                                                                  LIST  1
                                                                                  L U' T  j
MULTIPLE R
R SQUARE
ADJUSTED R SQUARE
STANDARD ERROR
                    0.85281
                    0.72729
                    0.68718
                    0.57838
                    ANALYSIS  OF  VARIANCE
                    REGRESSION
                    RESIDUAL
VARIABLE

VACOFF
LIMITS
CAPAC2
KIDS
TL1HIT
(CONSTANT)
                  VARIABLES IN THE EQUATION  	

                   B          BETA      STD ERROR  B
           -0.215*105
            0.5069719
           -0.2320196D-01
            0.15897710-01
           -0.6026162D-01
            -2.265219
-0.61369
 0.67519
-0.11751
 0.37027
-0.30139
0.03275
0.07750
0.00521
0.00415
0.02182
13.215
12.790
19.616
14.693
 7.630
DF SLK
b.
31.
VARIABLE
OUACRE
VACHSE
INCOME
UNt (• f
OFF JOB
WUJOC
iwAf^joe
JOfS
NONHH
POOR
VALUE
ROOMS
UNIV
PEL006
RCLWW
RILOFF
RELMAN
OOVT
LANG
POPOEN
OISCBO
RRMILE
ACCESS
INTDEN
TRANS
AIRPf T
VACANT
RZONED
CZONEO
ozorf o
I70NED
POLICY
TPOLIC
OR IVE
R1CET-
CF SQUARE?
i 0 . 3 3 d : 0
11.373nb
\/ADTAkirc
BETA U
-0.11611
-0.02212
-0 .00681
-O.OlObP
-0.13188
0.0103?
-O.lf til
-o.ii7f •:
0.093b7
0.139t8
-0.09852
-0.00568
-0.09690
-0.06567
-0.09265
-0.Of.fcS1>
-0.08361
0.09110
-0.06013
-0.13875
-O.l9b71
0.020f 1
-0.11712
-0.06502
-0.01726
-C .nc"CCn
0.05175
-0.05557
-0.1170fc
O.llb13
0.0b372
c.oc;r27
O.Obc fjO
-0.10197
-fj.1'0138
r- E /i r S
(• .
r.
luf^T Tfv Tut
NCI I f\ I n t
P iKTI AL
-C ,21f-35
-0 .01lOf
-0 .01259
-C. 01763
-0.23770
O.C1973
-0.29380
-O.Jr:520
0.16931
0.^2271
-0 .16907
-0.009ie
-0.13tl1
-C. 12387
-0.17273
-C. 130.23
-0.15852
0.15883
-0.10776
-0.216H
-0.3151b
0.0372f.
-0.11*751
-0 .1099h
-0 .02'.'95
-T .CP7P2
0.092Cl>
-0.088f C
-0.2115L
0.25213
C.G8';? T
O.OR137
L'.1025f
-0 .lf.f-.2f
-r . co 197
,LA«!-
066ST
.'3153
TCLERA..CE.
0.7r?H2
0.91150
0. 9235:2
0.7723L
D.8if-9t.
0.9921P
O.HbOOV
C .-(.(> 5 4
0 . b 8 7 b 1
0.69328
O.HC 323
0.70715
0.55bf>2
0.9C13P
0.9l7Cb
0.97178
0.97=.57
0.77205
0.1.17-90
O.P5P25
0.819HL
0.891^8
0.77180
0.779Hf
0.f2C>-l
0.71tc2
C.7f-l93
0.6<3-'2b
0.85^'fcl
0 . 8 2 1 7 r
0.7^6.-".
o .7^1:?
0 .6r..bJ
fj .^:. -1 It
r . 5 ?. v « 2
^
1 f . l J "< r. 1

F
2.1'..^
O.Cbf
0.101:
D.01L
1.97E
0.013
3 . ] 1 »•'
2.2U
0.971
1 ,72i
0.971
O.OOi
0.61 =
O.bll
1.015
0.5C9
0.851
O.d51
0.3B;.
2 . 1 2 1
1.173
O.OIb
1.310
0.111
0.030
0 . ? 3 *
0.2 0 7
0.26,1
1.551
2.21^
o . : ••- <•
c . c ^ 7
0 . ? b ]
C. V.-'.C
o. c! . :•

-------
13.  CY2 Equation

PREDICTIVE EQUATION REGRESSION

FILE   PREDICT  (CREATION DATE  =  02/14/78)



DEPENDENT VARIABLE..    CY2

VARIABLE(S) ENTERED ON  STEP  NUMBER   5..
                                                                                       ? /1 4 / 7 *
                                                MULTIPLE
                                      R  C  r-  IV  E  S  S  I  C I1
                                            VACANT
                                                                         V - f I 6 r L -. LIST
                                                                       R E o * E 5 S t C !\ LIST
MULTIPLE R
R SQUARE
ADJUSTED R SQUARE
STANDARD ERROR
                    0.80663
                    0.65065
                    0.59928
                    0.80603
                    ANALYSIS OF  VARIANCE
                    REGRESSION
                    RESIDUAL
VARIABLE

PHASE
TRANS
DISCBD
GOVT
VACANT
                  VARIABLES  IN  THE  EQUATION  	

                   B          BETA      STD  ERROR B
            -1.538799
           -0.4589325D-02
           -0.1779279D-C1
           -0.4169631D-02
             1.481964
-0.6111''
-0.48208
-0.42150
-0.29076
 0.25541
0.26515
0.00106
0.00461
0.00152
0.62051
33.60?
18.92C
14.872
 7.491
 5.727
(CONSTANT) -0.5360620
OF SU1-'
5.
34.
VARIABLE
DUACRE
VACHSE
INCOME
VACOFF
UNFMP
OFFJOB
UUJCB
MANJOB
JOBS
NONHH
POOR
VALUE
ROOMS
KIDS
UN IV
RELJOB
RELUU
RELCFF
RE L MAN
LAMP
POPDEN
RRWILE
ACCESS
INTDEN
AIRPRT
RZONED
CZCNED
OZONED
IZCNED
LIMITS
POLICY
TLI^'IT
TPOLIC
DRIVE
WIDET
CF SQUARES
41.11118
22.06952
BETA IN
P . 0 4 1 f • 0
-0.06327
-0.07297
-0.1C941
-0.17053
C. 14171
0.0614C
-0.12059
0.04676
0.13476
0.07962
0.02P92
-0.00608
-0.09082
0.1067?
-0.10005
0.03920
-0.144*6
-0.21026
0.03009
0.02348
0.11415
-0.04111
-0.05989
0.0632C
-0.02819
0.07065
-0.13881
-0.05335
0.20292
0.01421
o . o n 1 7 »
0.01234
-0. lifter
-o ,047e?
                                                            ' FA1*, SCUARF.
                                                                t .2?r?t
                                                                C.64S65
                                                                                                    NOT IN THE CliUATIO-
                                                                     1 2 . 6 6 1' ,'
 PARTIAL

 0 .057?*-
-O.C9540
-0.11904
-0.17081
-0.26649
 0.19066
 0.09586
-0.19289
 0.07013
 0 .21902
 0.11169
 0.03291
-0.00907
-0.12926
 0.12613
-0.12888
 0.04531
-0 .19766
-0.30112
 0.04512
 0.03301
 0.1869P
-0.06520
-0.08790
 0.09367
-0.043U
 0.11010
-0.21978
-0.07414
 0.31177
 C .02277
 T.01 '77
 0.01^82
-0.15150
-C .1)6650
                                                                                                                TOLERANCE
I). 79421
0.92956
0.85151
0.85310
0.63238
0.85149
0.89IC6
0.72261
0.92272
0.68399
0.86453
0.77612
0.70770
0.48792
0.57968
0.46686
0.65319
0.71651
0.78574
0.69068
0. 9?f>66
0.87877
0.75259
C.76744
0.61662
0.84H37
0.87576
0.67450
0.82471
0.89691
0 •(•(•? 14
0.50118
0.6C742
0.71477
0.109
0.303
0.474
0.992
2.52?
1.245
0.306
1.275
0.16?
1.66?
0.417
0.036
0.003
0.561
0.533
0.557
0.068
1.342
?.291
0.067
0.036

0.141
0.2^7
0.29?
O.CC1
0.4C5
1.675
0.162
3.55?
O.C17
                                                                                                                                  O.C13
                                                                                                                                  0.77?
                                                                                                                                  0 .1 •:•. (•

-------
    APPENDIX B



GLOSSARY OF TERMS
          B-l

-------
     A dependent variable is the output variable of a mathematical
function, which has as its arguments one or more input or independent variable^.

     Given a set of interrelated variables, an exogenous variable is  one  whose
value is determined by forces external to or outside the system of relationships
An endogenous variable is an internal variable of the system which is completely
determined (i.e., caused) by one or more exogenous and/or endogenous  variables.
The relationships of any one endogenous variable can be translated into mathe-
matical  form by designating it as the dependent variable of a function whose
independent variables are all related causal factors, whether they be endo-
genous or exogenous variables.

     Instrumental variables are exogenous variables which are introduced  into
nonrecursive systems of causal  relationships (i.e., those involving inter-
actions between endogenous variables) to allow the estimation of structural
coefficients.  A detailed list  of conditions associated with instrumental vari-
ables is given in Heise [171.

     A wastewater major project is defined as the construction or extension
of interceptor or collector sewer lines during the period 1958 to 1963 in a
community in the United States.  If construction date information is  not
available, then a grant funding date in the period 1956-1960 is acceptable.
The project had to affect an increase in absolute system collection capacity
of 1 MGD or more, and had to cost a minimum of $200,000 to construct. Phased
projects are considered if the first phase of construction on the collection
network was complete within 1958-1963 and the last phase of construction  was
complete by 1965.

     Induced development is defined as urban land use associated with, caused,
stimulated, or allowed by, or located because of the construction and operation
of a major project.

     Secondary development is induced development which has a direct causal
relationship with a major project.

     Tertiary development is induced development which has a direct causal
relationship with secondary development, and so is only indirectly related to
a major project.

     The drainage basin of a wastewater project is defined as the land area
which drains, by gravity, to any point of the collection network of the major
project.  In the case of essentially flat terrain (e.g., river deltas), this
area is restricted to the locus of points no greater than 1,000 feet from any
point along the interceptor line of the major project.
                                       B-2

-------
     The legal service area of a wastewater major project is  defined as
the drainage basin of the major project plus any additional  areas connected
to the collection network by pumping stations and force  mains.

     The area of analysis is defined as the legal service area  of a wastewater
major project in the base year.   It must be a minimum size of 5,000 acres  (or
approximately 8 square miles), and contain  significant amounts  of vacant de-
velopable land, some of which must be more  than 5,000 feet from the nearest
interceptor line in the base year.

"201" refers to Section 201 of the Federal  Water Pollution Control  Act
Amendments of 1972 (PL 92-500).   Section 201 calls for detailed planning for
the wastewater treatment facilities needed  to achieve the goals of the Act.

"208" refers to Section 208 of the Federal  Water Pollution Control  Act
Amendments of 1972 (PL 92-500).   Section 208 provides for the designation  of
state and areawide agencies for the purpose of developing effective water
quality management plans for areas that, because of "urban-industrial  con-
centrations" or other factors, have "substantial water quality  control problems."
The approach is aimed at integrating controls over municipal  and industrial
wastewater, storm sewer runoff,  nonpoint source pollutants and  land use.
                                      B-3

-------
          APPENDIX C



DEFINITION OF MODEL VARIABLES
              C-l

-------
     This appendix summarizes the model variables used in the GEMLUP-II model,
A complete discussion of variable selection is given in Section II.6 of the
first volume report [12].  Table C-l lists the definitions of endogenous
model variables, while Table C-2 summarizes the definitions of exogenous
model variables, both in English units.  The metric units of those variables
actually used in the predictive model and worksheets are listed in Table C-3.
                                     C-2

-------
                                                       TABLE C-l

                                     DEFINITION OF ENDOGENOUS MODEL VARIABLES
         Name
         Description
   Land Use Codes [106]
         RES
         COMM
        OFFICE
o
I
CO
        MANF


        WHOLE



        HIWAYS

        EDUC
 Number  of dwel 1 ing' units  per 10,000 acres of area
 of  analysis  in  1970.
 Commercial  land use  per 10,000 acres of area  of
 analysis  in  1970  in  1,000 square
 feet.

 Office-Professional  services land
 use  per 10,000 acres of area  of analysis
 in 1970 in  1,000  square feet.
Manufacturing land use per 10,000 acres
of area of analysis in 1970 in '1,000 sq. ft.
Wholesale-warehousing land use per 10,000
acres of area of analysis in 1970 in
10,000 square feet.

Non-expressway highway lane miles per 10,000
acres of area of analysis in 1970.
Educational  land use per 10,000 acres of area
of analysis in 1970 in 1,000 square feet.
   11   Household units
   12   Group quarters
   13   Residential hotels
   14   Mobile home parks or counts
   19   Other residential

 52-b9  Retail trade
   62   Personal services
   64   Repair services
   66   Contract construction services

   61   Finance ,insurance, and real
        estate services
631-636 Business services (excludes
638,639 warehousing)
     65 Professional  services
    692 Welfare  and  Charitable
        services
    699 Other services

  2,3   Manufacturing


   51   Wholesale trade
  637   Warehousing and storage services


   45   Highway and street right-of-way


  681   Nursery, primary, and secondary
        education

-------
                                                    TABLE C-l  (CONTINUED)

                                          DEFINITION OF ENDOGENOUS MODEL VARIABLES
     Name
          Description
 Land Use Codes  [106]
     REC
     OTHER
Active, outdoor recreational  land  use per 10,000
acres of area of analysis in 1970 in acres.

Other urban land uses per 10,000 acres of area of
analysis in 1970 in 1,000 square feet.
 73  Amusements
 74  Recreational activities
 75  Resorts and group ca;nps
 15  Transient lodgings
691  Religions activities
 71  Cultural  activities and
      nature exhibitions
 79  Other cultural,  entertain-
      ment, and recreational
o

-------
                                TACLE  C-2

                DEFINITION OF EXOGENOUS  MODEL  VARIABLES
Name
jescnpnon
Data Source
Service Area Base Year Characteristics:   Socioeconomic  Variables

DUACRE = Dwelling units per  mile2  in  area of  analysis in  1960.
       = (100*du60)/acre
         where:   du60 = 1960 census  tracts* housing  units in  100s   Census
                 acre = 1960 area  of  census tracts  in miles         Census

VACHSE = Percent vacant available  dwelling units  in  area  of
         analysis in 1960.                                          Census

INCOME - Relative medium income of families and  unrelated
         individuals in area of analysis  compared to county
         income  levels in 1960.
       = (10*inc)/median
         where:   inc     =  1960 median income for families in
                           census  tracts  in $10s
                 median  =  1960 median ir.come for families in
                           the county**

VACOFF = Vacancy rate of office buildings in  area of analysis
         in 1960.

UNEMP  = Unemployment rate  in area of analysis (census  tracts)
         in 1960.

OFFJOB = Office  employment per mile   in area  of  analysis  in  1960.
       =^ (100*smoff )/acre
         where:   smoff  = I960 office employment in  census          Census
                          tracts in  100s***.

WWJOB  = Warehouse and wholesale employment per  mile in  area
         of analysis in 1960.
       = (100*wwemp)/acre
         where:   wwemp = 1960 employment  in wholesale trade4"  in     Census
                         census tracts in 100s.
                                                            Census

                                                            Census
                                                            BOMA/PTanning
                                                            Agency

                                                            Census
  * Census tracts refers to those tracts  most closely approximating  the
    area of analysis in areal  extent.
 **
    County refers to the county containing most of the  legal  service  area.
*** Fire, Business Services,  Public Administration,  and Repair Services.
  + Trucking,  Warehousing,  and Wholesale  Trade.
                                     C-5

-------
                          TABLE C-2  (CONTINUED)

                DEFINITION OF EXOGENOUS MODEL VARIACLES
Name - Description
Data Source
MANJOB = Manufacturing employment per mile  in area of analysis
         in 1960.
       = (100*manemp)/acre
         where:  manemp = 1960 manufacturing employment in
                          census tracts in 100s.
JOBS   = Total employment per mile  in area of analysis in 1960,
         (100*totemp)/acre
         where:  totemp = 1960 total civilian employment in
                          census tracts in 100s.
                                          2
NONHH  = Non-household population per mile  in area of analysis
         in 1960.
         (100*nonh60)/acre
  Census
  Census

POOR
RENTS*
VALUE
ROOMS
KIDS
where: nonh60 = 1960 population in group quarters in
census tracts in 100s.
= Percent of total families with income below $3,000
in area of analysis (census tracts) in 1960.
= Percent of total housing units that are renter occu-
pied in area of analysis in 1960.
= Median value of housing units in area of analysis
(census tracts) in 1960.
=v Median number of rooms in housing units in area of
analysis (census tracts) in 1960.
= School age children per 100 households in area of
analysis in 1960.
= 100*sch60/du60
where: sch60 = 1960 population 0-14 years of age
in census tracts in 100s.
Census
Census
Census
Census
Census
Census
 *  Used  only  in  the disaggregation analysis of RES.
                                     C-6

-------
                            TABLE  C-2 (CONTINUED)

                DEFINITION OF EXOGENOUS MODEL  VARIABLES
Name - Description
                                                                 Data  Source
GOVT   = Total county government expenditures in 1962 in 10  $.

LAND   = Price of vacant land in area of analysis in 1960
         relative to median regional  income.
       = price/median
         where:  price = median price of one  acre of residential
                         vacant land ($) in area of analysis
                         in 1960.
                                                                   Planning
                                                                   Agency
                                                                   Census

                                                                   Census
UNIV   = Categorical  variable  to  indicate  the  presence  or
         absence of a college  or  university in the  area of
         analysis in  1960.
         where:   1  =  a college or university existed  in census
                     tracts  in 1960.
                 0  -  none existed.

RELJOB = Relative employment density  of the area  of analysis
         in 1960.
       = JOBS/(100*rempl/areal)
         where:   rempl =  1960  total employment in the SMSA
                         in  100s.                  2
                 areal =  1960  SMSA land area in mile  .

RELWW  = Relative warehouse  and wholesale  employment  density
         of the area  of analysis  in 1960.
       = WWJOB/(100*rwwemp/areal)
         where:   rwwemp =  1960 employment  in warehouse  and
                         wholesale trade  in the  SMSA in 100s.

RELOFF = Relative office employment density in the  area of
         analysis in 1960.
       = OFFJOB/(100*roffl/areal)
         where:   roffl = 1960  SMSA office  employment  in 100s.
        V
RELMAN = Relative manufacturing employment density  of the area
         of analysis  in 1960.
       = MANJOB/(100*rempl*manper/areal)
         where:   manper =  percent of  1960  total SMSA  employment
                          in manufacturing.
                                                                   Census
                                                                   Census
Census


Census
                                                                   Planning
                                                                   Agency
                                      C-7

-------
                            TACLE C-2 (CONTINUED)

                  DEFIMTIOIJ OF EXOGENOUS MODEL VARIABLES
Name - Description
                                                                 Data Source
POPDEN = Population density of the area of analysis in 1960
         in persons per mile2.
       = (100*pop)/acre
         where:  pop = total population of census tracts in
                       1960 in 100s.
                                                                 Census
SFDET  = Percent of housing that is
         area of analysis in 1970.
       = unitl/hse70
         where:  unit! =
                                   single family detached in
                        number of single family detached
                        units in census tracts in 1970 in 100s.
                hse70 = total number of housing units in
                        census tracts  in 1970 in 100s.
                                                Census

                                                Census
MF
         Percent  of
         analysis in
         (unit34+uni
         where:   unit34  =
   housing that is  multifamily in  area  of
    1970.
   t5+unit50)/hse70
                                          4 family housing
                                          tracts in 1970 in
         number of 3 and
         units in census
         100s.
units  = number of 5-49 family housing units
         in census tracts in 1970 in  100s.
unit50 = number of 50+ family housing units
         in census tracts in 1970 in  100s.
Census


Census
SFATT  =
         ((hse70-uni
         Percent of
         in  area of
   tl)/hse70)-MF
   housing that is single family attached
   analysis in 1970.
PCOMM1 = Percent of
         50,000 ft2
                     commercial  development  less than
                     in  floor  area.
       = comml/(comml
         where:   comml
       comm2 + comm3)
         total commercial floor space in
         1,000 ft2 in area of analysis in
         1970 for buildings with less than
         50,000 ft2.
commZ -  total commercial floor space in
         1,000 ft2 in 1970 in area of analysis
         for buildings with less than 100;000-ft:
         but greater than 50,000 ft2.
comm3 =  total commercial floor space in 1,000
         ft2 in 1970 in area of analysis for
         buildings with greater than 100,000 ft2.
                                                                  Planning
                                                                  Agency
                                        C-8

-------
                          TABLE C-2 (CONTINUED)

                 DEFINITION OF EXOGENOUS ''ODEL VARIABLES
-;a,T;-5 - Description
Data Source
PCOMM2 = Percent of commercial  development with
         floor area between 50,000 and 100,000 ft2.
       = comm2/(comml  + comm2 + commS)

PCOf'M3 = Percent of commercial  development with
         floor area greater than 100,000 ft2.
       = comm3/(comml  + comm2 + commS)


Service Area Base Year Characteristics:   Accessibility

DISCED = Distance in miles from centroid of area of  analysis
         to centroid of nearest CBD* in year (t + 0)**

RRMILE = Railroad mileage  per mile  in the area of analysis
         analysis in year  (t +  0)
       = (rail*640)/AREA
         where:   rail  = railroad mileage in area of  analysis
                        in year (t + 0)

ACCESS = Limited-access highway interchanges per mile  in
         area of analysis  in year (t + 5)
       = (intchg*640)/AREA
         where:   intchg =  number of limited access inter-
                          changes in area of analysis in
                          year (t + 5)

INTDEN = Relative limited-access highway interchange density
        vof the area of analysis in year (t + 5)
       = ACCESS/(ctyacc*640/county)
         where:   ctyacc =  number of limited access inter-
                          changes in the county in year (t +  5)
                 county =  area  of county in mile2

TRANS  = Number of transit stops (bus and commuter rail) in
         the area of analysis in year (t + 0)

AIRPRT = Distance in miles from centroid of area of  analysis
         to centroid of nearest commercial airport in the
         year (t + 0)
       USGS  topo-
       graphic  map
       USGS  topo-
       graphic  map

       USGS  topo-
       graphic  map/
       Planning
       Agency
       USGS topo-
       graphic map/
       Planning
       Agency
       Planning
       Agency

       USGS topo-
       graphic map
* Central Business District, defined as the center of the nearest urban
  area with population exceeding 100,000.

**t U cue year tne-wastewater major project was completed and became
"operational.-
                                       C-9

-------
                             TABLE C-2  (CONTINUED)

                   DEFINITION OF EXOGE.'i.'JS MODEL VARIABLES
liar.e - Description
Data Source
Service Area Base Year Characteristics:  Land Use Constraints

AREA   = Area of analysis in acres

VACANT = Percent vacant developable acreage in area of analysis
         in year (t + 0)
       = vacdev/(AREA-vacund)
         where:  vacdev = vacant developable acreage in area of
                          analysis in year (t + 0)
                 vacund = vacant undevelopable acreage in area
                          of analysis in year (t + 0)

RZONED = Percent of total acreage zoned for residential use in
         the area of analysis in year  (t + 0)
       = rzone/AREA
         where:  rzone = acres of land zoned for residential use
                         in the area of analysis in year (t + 0)

CZONED = Percent of total acreage zoned for commercial use in the
         area of analysis in year (t + 0)
       = czone/AREA
         where:  czone = acres of land zoned for commercial use
                         in the area of analysis in year (t + 0)

OZONED = Percent of total acreage zoned for office use in the
         area of analysis in year (t + 0)
       = vozone/AREA
         where:  ozone = acres of land zoned for office use in
                         the area of analysis in year (t + 0)

IZONED = Percent of total acreage zoned for industrial use in
         the area of analysis in year  (t + 0)
       = izone/AREA
         where:  izone = acres of land zoned for industrial use
                         in the area of ana-lysis in year (t + 0)

SOILS  = Percentage of total area of analysis having a "severe"
         soil type classification for  urban development suita-
         bility
       = soil/AREA
         where:  soil = Acreage in area of analysis with a
                        "severe" soil  type classification with
                        regard to suitability for urban devel-
                        opment.
Project Data
Planning
 Agency
Planning
 Agency
Planning
Agency
Planning
Agency
Planning
Agency
Planning
Agency

Planning
Agency/Soil
Conservation
Service (SCS)
                                       £-10

-------
                            TABLE  C-2  (CONTINUED)

                  DEFINITION OF  EXOGENOUS  HODEL  VARIABLES
r;are - Description
                                                                  Data Source
                                                       "severe"
ONLOT  = Percentage  of total  area  of  analysis  having a
         soil  type classification  for on-lot sewage disposal
         suitability.
       = onsite/AREA
         where:   onsite =  Acreage  in  area  of analysis with a
                          "severe"  soil  type classification for
                          on-lot sewage  disposal  suitability.

LIMITS = Categorical  variable to indicate  the  seventy of gov-
         ernmental  restrictions on  on-lot  sewage  disposal during
         the period  (t + 0)  to 1970.
                                     is  prohibited entirely.
                                     is  prohibited except on
         where:   4 =  on-lot  disposal
                 3 =  on-lot  disposal
                     large  lots.
                 2 -  on-lot  disposal  is  permitted  but  percola-
                     tion tests  are  required.
                 1 =  on-lot  disposal  is  permitted  but  package
                     plants  are  prohibited.
                 0 =  no restrictions.

POLICY = Categorical  variable  to indicate  the  presence or
         absence of governmental  policies  designed to  limit  the
         number of hookups  to  the sewerage system  in area  of
         analysis anytime during the  period (t + 0)  to 1970.
         Examples are sewer  moratoria and  rationed connections.
         where:   1 =  policies  on hookup  limitations existed
                 0 =  policies  did not exist.

TLIMIT = The number of years during  the  period of  analysis that
         on-site sewage disposal  was  limited
       = 1970-ylimit
         where:   ylimit = the  year on-site disposal limitations
                          went into  effect (or 1970 if no
                          restrictions).

TPOLIC = The number of years during  the  period of  analysis that
         sewer system hookup limitations were  in effect
       = 1970-ypolic
         where:   ypolic = the  year limitations on  sewage system
                          hookups went into effect (or 1970  if
                          no restrictions)
                                                                  Planning
                                                                  Agency/SCS
Planning
Agency/
Local
Government
                                                                  Planning
                                                                  Agency/
                                                                  Local
                                                                  Government
                                                                  Planning
                                                                  Agency/
                                                                  Local
                                                                  Government
                                                                  Planning
                                                                  Agency/
                                                                  Local
                                                                  Government
                                      C-ll

-------
                          TABLE C-2 (COiiTi:;UED)

                DEFINITION OF EXOGENOUS MODEL VARI-V,LES
Name - Description
                                                                 Data Source
                                                                   Census
Regional Growth Factors
ENERGY = Relative electrical energy cost factor in municipality
         compared to average U.S. commercial rate in 1960.
       = encost/$51.59
         where:  encost = cost of energy for commercial users
                          in 1960 in units of dollars per
                          1500 KWh.
DRIVE  =
       - 100s of workers who drive per mite2 in the county in 1960.
       = (autoeoj/county
         where:  autoco = workers who drive to work in county
                          in 1960 in 100s.
RIDET  = Workers who ride mass transit per mile  in the county
         in 1960.
       = massco/county
         where:  massco = workers who use mass transit in county
                          in 1960 in 100s.

PERCHG = Percent change in county population 1960-1970

DENCHG = Population density change  in the county 1960-1970.
       = (copop2/county -  copopl/county)
         where:  copopZ = county population in 1970.
                 copopl =  county population in 1970.

JOBCHG -'Change in total regional employment per mile  1960-1970.
       = 100*(remp2 -  rempl)/SMArea
         where:  rempl = 1960 total employment in SMSA in 100s.
                 rempZ = 1970 total employment in SMSA in 100s.
                 SMArea= area of SMSA in miles .

HSECHG = Percent change in total regional housing units 1960-1970.
       = (rhse2 - rhsel)/rhsel
         where:  rhsel = total housing units in SMSA in 1960
                         in 100s.
                 rhse2 = total housing units in SMSA in 1970
                         in 100s.
                                                                   Census
                                                                   Census


                                                                   Census
                                                                   Census
                                                                   Census
                                                                   Census
                                                                   Census
                                                                   Census
                                                                   Census

                                                                   Census
                                     C-12

-------
                            TABLE C-2  (CONTINUED)
                  DEFINITION  OF  EXOGENOUS  TCOEL  VARIACLES
Name - Description
Data Source
COMCHG = Percent change  in  total  regional  retail  trade*  employment
         1960-1970.
       = (rcom2 - rcoml)/rcoml
         where:  rcoml  = total  retail  trade  employment in  SMSA
                         in 1960  in  100s.
                 rcomZ  = total  retail  trade  employment in  SMSA
                         in 1970  in  ICOs.
POPDIF = Percent change  in regional  population  1960-1970.

MANCHG = Percent change  in regional  manufacturing  employment
         1960-1970.
       = (perman*remp2-manper*rempl)/(manper*rempl)
         where:   perman  = % of 1970  total  SMSA  employment  in
                          manufacturing
                 manper  = % of 1960  total  SMSA  employment  in
                          manufacturing

SERCHG = Percent change  in regional  services**  employment  1960-
         1970.
       = (rser2 - rserl)/rserl
         where:   rserl  = 1960 SMSA service employment in  100s.
                 rser2  = 1970 SMSA service employment in  100s.

HOSCHG = Percent change  in regional  hospital  employment 1960-1970.
       = (rhosp2 - rhospl)/rhospl
         where:   rhospl  = 1960 SMSA hospital  employment in 100s.
         v        rhosp2  = 1970 SMSA hospital  employment in 100s.

EDUCHG = Percent change  in regional  educational  employment
         1960-1970.
       = (red2 - redl)/redl
         where:   red!  =  1960 SMSA  educational employment  in 100s.
                 red2  =  1970 SMSA  educational employment  in 100s.

OFFCHG = Percent change  in regional  office*** employment  1960-
         1970.
       = (roffZ - roffl)/roffl
         where:   roffl  = 1960 SMSA office  employment in 100s.
                 roff2  - 1970 SMSA office  employment in 100s.
Census

Census


Census
Census

Census
Census
Census
Census
Census
Census
Census
Census
Census
*   Food and Dairy, Eating and Drinking,  and Other Retail.

**  Professional  and Related Son/ices, Other Personal Services, Entertainment
    and Welfare and Fraternal Organizations.

*** FIRE, (Finance, Insurance, and Real Estate) Business Services  Public
    Administration, Repair Services, and  Public Administration.

                                    C-13

-------
                            TABLE C-2  (CO;ri;;UED)

                  DEFINITION OF :x.occ;;oi'S "OCLL VARIABLES
Nane - LQScriotion
Data Source
PERCAP = Median far-rHy income in SMSA in 1970.                       Census

EMPOP  = Regional  (SMSA) employment to population ratio in 1970.     Census
       = remp2/rpop2
         where:  rpop2 = 1970 SMSA population in 100s.              Census

CAPCHG = Change in median family SMSA income 1960-1970.
       = PERCAP - per60
         where:  per60 = median family income in SMSA in 1960.       Census

STAY     Index of mobility - % of 1960 families who were in the     Census
         same house in 1955.

Wastewater Treatment Major Project Characteristics

TIME   = Number of years available for secondary growth to occur.
       = phyear-year
         where:  phyear = the year aerial photographs were taken    Planning
                          from which land use data was extracted.    Agency
                 year   = the year construction was completed on    Project
                          the major project on initial phase.        Data

PHASE  = Categorical variable to indicate the presence or           Project
         absence of phasing of the major project.                    Data
         where:  1 = phasing has occurred.
                 0 = phasing has not occurred.

COST   =>Normalized per capita local costs  ($) of the major
         project in area of analysis in year (t + 0).
       = 1,000* (totest/pil - fedcst/pi2)/popcom
         where:  totcst = total major project construction          Project
                          cost (1,000 $)                             Data
                 fedcst = federally funded  share of major           Project
                          project cost (1,000 $)                     Data
                 pi!    = consumer price index for "year"
                 pi2    = consumer price index for the year of
                          federally funding of the major project
                 popcorn = population served by major project        Project
                          facility in year  (t + 0)                   Data

LENGTH = Running length of interceptor sewer lines in miles         Project
         going through relatively undeveloped land (<.1 du.per       Data
         acre) in area of analysis in year  (t + 0).

-------
                            TABLE C-2 (CONTINUED)

                  DEFINITION OF EXOGENOUS  MODEL  VARIABLES
Name - Description
Data Source
CROSS  = Index of available  undeveloped  land  in  the  area  of
         analysis through  which  interceptor sewers go  through
         in t + 0.
       = 640*LENGTH/AREA

SERVED = Percent of the area of analysis easily  served by the
         major project in  year (t + 0)
       = area5k/AREA
         where:  areaSk =  acres  of land  within 5,000 feet of the
                          major project  interceptor  sewer in
                          area of analysis  in year  (t  + 0)

CAPAC1 = Total hydraulic design capacity of the  major  project
         wastewater collection system in million gallons  per
         day  (mgd) in year (t + 0) (or in 1965 if a  phased
         project)

CAPAC2 = Total hydraulic design capacity of the  major  project
         wastewater treatment plant in mgd  in year  (t  + 0)

PEAK   = Actual Peak flow  in the major project wastewater system
         in mgd in year (t + 0)

RECAP  = Reserve capacity  of the wastewater major project
         collection system in year (t +  0)
       = CAPAC1   PEAK

RECAP1 = Percent reserve capacity of the wastewater  major
         project collection  system in year  (t +  0)
       = 1QO*RECAP/PEAK

RECAP2 = Percent reserve capacity of the wastewater  major
         project treatment plant in year (t + 0)
       - 100*(CAPAC2 - PEAKJ/PEAK
 Project
 Data
 Project
 Data
 Project
 Data

 Project
 Data
                                     C-15

-------
                                TABLE C-3

 METRIC  UNITS OF  VARIABLES  USED  IN THE PREDICTIVE  EQUATIONS AND WORKSHEETS
Variable
Predictive
Equations
Dependent
RES +

COMM+

OFFICE +

MANF +

WHOLE+

HIWAYS+

EDUC +

REC +
Names
Worksheets

Residential

Commercial

Office-
Professional
Manufacturing

Wholesale

Highways

Education

Recreation
Metric Units



Dwelling units per 10,000 square meters of
area of analysis
Square meters of floor area per 10
meters of area of analysis
Square meters of floor area per 10
meters of area of analysis
Square meters of floor area per 10
meters of area of analysis
Square meters of floor area per 10
meters of area of analysis
Lane kilometers per 10,000 square
area of analysis
Square meters of floor area per 10
meters of area of analysis
Square meters of recreational land

,000 square

,000 square

,000 square

,000 square

meters of

,000 square

use per
10,000 square meters of area of analysis
OTHER +

SFDET
SFATT
MF
PCOMM1
PCOMM2
PCOMM3
Independent
VACANT
LAND
RECAP!

MANJOB

STAY
DRIVE

SERVED
KIDS
ACCESS

Other

% Single Family
% Two Family
% Mul ti family
% Small Comm'l
% Med Comm'l
% Large Comm'l

Vacant Land
Land Cost
% Collection
Reserve
Manufacturing
Density
Nonmobility
Driver
Density
Sewer Service
Kid Density
Interchanges

Square meters of floor area per 10
meters of area of analysis
*
*
*
*
*
*

*
*
*

Employees per square kilometer of
analysis
*
,000 square












area of


100s of drivers per square kilometer of area
analysis
*
*
Interchanges per square kilometer
of analysis



of area

                                                                           of
* Same as English units.
+ Units for Predictive Equation variables are per 10,000 acres (square meters)  of
  area of analysis, while units for Worksheet variables are totals for the entire
  area of analysis.
                                 C-16

-------
                           TABLE C-3 (CONTINUED)
 METRIC UNITS OF VARIABLES USED IN THE PREDICTIVE EQUATIONS AND WORKSHEETS

        Variable Names
Predictive
Equations
JOBCHG

RRMILE

OZONED
PEAK
DUACRE

IZONED
POPDIF
VACOFF
AIRPRT
OFFJOB

EMPOP
UNEMP
PERCAP
GOVT
RECAP
CROSS


POOR
LIMITS
PERCHG
COST
CAPAC2
VACHSE
INTDEN
RZONED
CAPCHG
HOSCHG
HSECHG
TLIMIT
PHASE
TRANS
DISCED
vacdev
vacund
AREA
price
median
Worksheets
Employment Growth

Railroads

Office Zoning
Peak Flow
House Density

Industrial Zoning
Population Growth
Office Vacancy
Airport Distance
Office Employees

Employee Ratio
Unemployment
Future Income
Government
Collection Reserve
Interceptor Density


Poverty
Onsite Restrictions
County Growth
Sewer Costs
Treatment Capacity
Vacant Houses
Interchange Density
Residential Zoning
Income Growth
Hospital Growth
Housing Growth
Restriction Years
Phasing
Transit Stops
CBD Distance
Vacant Developable
Vacant Undevelopable
Area of Analysis
Median Price
Median Income
Metric Units
Employees per square kilometer of area
of analysis
Kilometers of railroad track per
square kilometer of area of analysis
*
*
Dwelling units per square kilometer
of area of analysis
*
*
*
Kilometers
Employees per square kilometer of area
of analysis
*
*
*
*
*
Kilometers of interceptor pipe per
10,000 square meters of area of
analysis
*
*
*
*
*
*
*
*
*
*
*
*
*
*
Kilometers
Square meters
Square meters
Square meters
*
*
  Same as English units.
                                     C-17

-------
                           TABLE C-3 (CONTINUED)

 METRIC UNITS OF VARIABLES USED IN THE PREDICTIVE EQUATIONS AND WORKSHEETS
        Variable Names
Predictive
Equations         Worksheets
                                      Metric Units
CAPAC1
manemp+
acre
autoco
county
areask
sch60-t-
du60
intchg
rempl +
remp2+
SMArea
rail
ozone
izone
smoff+
rpop2+
LENGTH
to test
fedcst
pil
Pi 2
popcorn
ctyacc
rzone
per60
rhospl+
rhosp2+
rhsel+
rhse2+
Collection Capacity
Manufacturing Workers
Tract Area
Drivers
County Area
Sewered Land
School Kids
Dwelling Units
Limited -Access
Current Employment
Future Employment
SMSA Area
Track
Zoned Office
Zoned Industrial
Office Workers
Future Population
Interceptors
Project Cost
Federal Funds
Index One
Index Two
Population Served
County Interchanges
Zoned Residential
Current Income
Current Medicals
Future Medicals
Current Houses
Future Houses
Square kilometers
*
Square kilometers
Square meters
*
*
*
*
Square kilometers
Kilometers
Square meters
Square meters
*
*
Kilometers
*
*
*
*
Square meters
*
*
*
*
*
*Same  as  English  units.
+Units for  Predictive  Equation  variable  are  in  100s while units for
 Worksheet  variables are  in  Is.
                                    C-18

-------
                APPENDIX D

GRAPHS OF ACTUAL VERSUS PREDICTED LAND USE
    FOR THE CROSS-VALIDATION ANALYSIS
                    D-l

-------
 1.  RES  Equation
                                                               TSP
                                                                                      flGUSN
                                                                                                  P A ii L
                                    PLOT  OF ACTUAL<*>  AMD  FITTEC<+> VALUES
                                                                                           PLOT  CF  f-.ESIt UALS(O)
10
8
9
11
15
16
17
19
20
22
t-j 23
25
28
31
33
35
36
37
ACTUAL
5911.
9706.
.1197E+05
.2357E+OS
6072.
3745.
9709.
5100.
7540.
5017.
7372.
5903.
4626.
7772.
6516.
.1358E+05
.1050E+05
369.0
.1322E+05
FITTED
6919. * +
7983. + *
.1547E+05 * +
.1374E+05 +
6033. +
.1114E+05 * +
b938. + *
.1205E+05 * +
6407. + *
1461. + *
3406. + *
469.3 + *
3351. «• *
5713. + *
6803. *+
.1344E+05 +
.1412E*05 * +
4711. * +
.1461E+05 * +
RESIDUAL
0.0
-.101E+04 . 0.
.172t+04 . . 0 .
-.35CL+04 0
* .9B2E+04
39.4 . 0
-.740F+04 0 . .
771. . .0
-.6-95E + 04 0
.113E+04 . . C .
,3b6t+04 . . 0.
.397T+04 . . C
.54-+04 . . . 0
.127^04 . . 0 .
.206L+04 . . 0 .
-287. . 0
134. .0 .
-.362L+04 C .
-.434E+04 C.
-.1391+04 . 0 .
39
5375.
3670.
.171E+T4

-------
    2.  COMM Equation
    LI'":
    1
TC'
                                                                T~p
                                                           V!
1 f f   CF  61 MS 1 •
                                            PLOT  Ct-  A C T L A L ( * >  A > D  F I T T j. L ( * )  V A LIt S
                                                                               'LC T  fF  .''
                                                                                                                         MM," )
TO
1

*
8
9
11
15
16
17
19
20
22
23
25
28
31
33
34
35
36
37
ACTUAL

892.4
1794.
1206.
3H62.
111?..
1104.
2027.
701. "
?776.
1897.
1640.
774. C
722.8
1197.
807.8
1071.
.1127E+05
245.7
1510.
FITT£D

1 438. * +
2fcS7.
250. P. + *
1 9 1 . t + *
llc.9. * +
2251. * +
251.4 + *
4336. * +
56^3. * +
178.2 + *
1870. *+
2373. * +
1379. * +
2596. * *
537.4 +*
-133. « + *
6266. *
-37.09 * *
1950. * +
f c c. lot A I

-tOL- .
-873.
c.i L 5 .
.3f,7r + C4
-71.2
-•119L+G4
.178!:+04
-.?«-.?E*C* C
-.2V2L+C4
.172E+01
-270.
-.1COL+04
-656.
-.140E+04
270.
.120F. + 04
* .bOOE+04
?P3.
-441 .

0.0
C.
• * • •
0
• • •
0
. c .
. . 0.
1
• • •
0 .
• • [. •
0. " .
.0 . .
. c .
.0 . .
.0 •
. I .
• • •
. .0 .
.0. .
  39
609.2

-------
CD
J=-
        3.  OFFICE Equation

                            PRINCETON UNIVERSITY
    ID
LI

4
8
9
11
15
16
17
19
20
22
23
25
28
31
33
34
35
36
37
NE 15
ACTUAL
86.40
288.5
555.1
687.5
935.9
403.4
822.3
995.9
230.9
177.8
34.00
94.70
144.6
577.2
385.5
296.6
2111.
94.60
489.7
PRII
FITTED
468.1
914.7
1009.
-214.4
648.9
647.7
161.0
1000.
844.7
287.5
210.7
655.8
393.9
681.8
515.2
1065.
1312.
64.15
681.7
                                                        TSP
                                           PLOT  OF ACTUAL(*) AND  FITTEDC+)  VALUES
                                                                     OF  fUOUSTi  I?*-"?
                                HAGL
                                                                             RESIDUAL

                                                                              -7H2.

                                                                              -62f .
                                                                              -454.

                                                                               S02.

                                                                               287.
                                                                              -244.
                                                                               661.

                                                                              -4.31
                                                                              -614.

                                                                              -110.

                                                                              -177.

                                                                              -561.

                                                                              -249.

                                                                              -105.

                                                                              -130.

                                                                              -768.
                                                                               799.
                                                                               30.4
                                                                              -192.
                                                                                           PLOT Ch RESIt OALS«j )
                                                                                                                0.0
                                                                                                              0   .
                                                                              0.
                                                                                                                0.
                                                                                                                     0
                                                                                                                        .  0
      39
98.60
110.7
-12.1

-------
UNIVERSITY
                                              7 Si
                                                                              VETS I CK  OF  i'cL l.'ST
                                                                                                                             ' r-E.    1
    PLC7  OF AC7UAL<*) t"D  FITTLr<+>  VALLLS
                                                                                   f:LOT OF  F05
335.5
4542.
FITTER

1717. +*
1090. +
2irft. * +
lie;. * *
1205. **
3332. +
435.1 *
4755. * *
4363. * *
156.4 + *
177.2 •• *
2727. * +
344.9 * *
1313. +
1848. * +
5971. * *
3865. +
51.18 +*
2259. + *
Fu" S 1 C U A L

f-7.9
-39.4
-.159E+04
-366.
-134.
-6£.8
lit.
-.333F+04
496.
.111F+04
.219E+04
-.237E*04
700.
-158.
-r37.
-.545E*04 0
* .68CE+C4
284.
.228E+04

C .0
0
c
. c .
0.
0.
0
0
0 . .
.0
. 0
» •
0
.0
0.
0.
* •
• •
0
• *


•
•
•
•
•
•
•
•
*
•
c
.
•
•
•
•
•
•
0
39
140. P
-363.1
                                                        504.
                                                                                                                 .0

-------
5.  HIWAYS Equation

LINE   17           PRINCETON1 UNIVERSITY
                                                                  TSF
                                                                               VEkSICr. Of-  AUGUST,
                                                                                                                     21
                                         PLOT OF ACTUAK*) AND FITTED(+) VALUES
                                                                                               PLOT CF FESICUALS(O)
   ID
cr>
 8
 9

11

15
16
17

19
20

22

23

25

28

31

33

34
35
36
37

39
ACTUAL

 15.40

 47.00
 84.00

 57.50

 56.60
 42.3C
 25.30

 3.000
 15.20

 45.60

 13.40

 30.30

 30.80

 48.30

 34.50

 24.80
 48.20
 14.00
 36.50

 15.50
                 FITTED                                                         RESIDUAL

                  8.726        + *                                                6.67

                  48.93                         *+                            •    -1.93
                  114.9                                         *              +  -30.9

                  38.66                     +        *                             18.R

                  53.74                            +*                              2.8f
                  68.18                      *            •»                       -25.9
                  37.00               *    *                                     -11.7

                  62.94     *                          +                          -59.9
                  flO.OO          *                             +                 -64.8

                  31.15                 *       *                                  14.4

                  19.50          * +                                             -6.10

                  70.58                 *                  +                      -40.3

                  40.50                 *   *                                    -5.70

                  53.30                         *  +                              -5.00

                  76.19                   *                  >                    -41.7

                  105.9               *                                   +      -Fl.l     0
                  64.83                         *      +                         -16.fc
                  21.67          *  «•                                            -7.67
                  41.14                    * *                                   -4.64

                  5.484      +    *                                               10.4
                                                                                                                   c.r
                                                                                                                    .  C
                                                                                                                        0

                                                                                                                    .0
        0.

 0.

       0 .

        0.

0 .
                                                                                                                  0.
                                                                                                                  0.
                                                                                                                      0    .

-------
   6.  EDUC Equation
L]
. ID
*
8
9
11
15
16
17
19
20
22
23
25
V
•Nl 28
31
33
34
35
36
37
It.? 18
ACTUAL
454.0
635.8
724.0
984.8
350.8
499.5
982.5
368.6
1279.
707.3
655.0
359.2
715.5
662.5
643.7
1037.
457.6
95.50
615.5
PRII
FITTED
488.2
694.8
747.4
21*. 8
652.5
408.0
1108.
440.9
896.0
489.7
320.7
456.4
863.8
690.1
630.7
895.5
460.7
-41. b6
804.4
                       PRINCETON UNIVERSITY
                                                                            VEF-Sin-  OF-  AulUSTi
                                      PLOT  OF- ACTUALt*)  AND FTTTLC<+)  VALUE.S
                                                                                             PLCT CF
RESIDUAL
* + -34.2
* + -59.0
*+ -23.4
+ * 766.
* * -302.
+ * 91.5
* + -126.
* + -72.3
+ * 383.
* * 218.
+ * 334.
* * -97.2
* + -148.
*+ -27.6
+ 13.0
•» * 141 .
* -3.11
« 137.
* + -189.
c.c
C .
• 0 . .
. 0 . .
• • . <
0. . .
. . 0 .
. 0 . .
. 0 . .
. • .
0
• • .
. 0 • .
. 0 .
. 0 . .
C
. 0 .
0
. 0 .
. 0 . .
39
11.60
109.6
-98.0
0 .

-------
7.  REG Equation
LINE   19
                     PRINCETON UNIVERSITY
                                                        TSP
                                                                     CF  AUGUST,
                                    PLOT OF ACTUALC*)  AND FITTEDC + ) VALUE'S
ID
4
8
9
11
15
16
17
19
20
22
V "
00 25
28
31
33
35
36
37
ACTUAL
28.00
37.00
437.0
500.0
13.00
103.0
106.0
444.0
289.0
65*00
23D.O
382.0
62.00
247.0
98.00
193.0
203.0
19.00
215.0
FITTED
108.1 * +
184.5 * +
930.6 *
547.9 * +
292.8 * +
263.1 * +
193.2 * +
362.9 *
401.6 * +
-35.56 + *
129.3 + *
432.1 * +
93.51 * +
292.9 * +
229.2 * +
314.9 * +
680.7 * +
140.4 * +
76.13 + *
RESIDUAL
-tO.l
-147.
+ -494.
-47.9
-2eo.
-160.
fcl.l
-113.
101.
101.
-50.1
-31.5
-45.9
-131.
-122.
-478.
-121.
139.
                                                                                          PLOT  OF  RESIDUAL^C)
                                                                                                               0.0
                                                                                                          .0
                                                                                                          .  0
                                                                                                               0.

                                                                                                               0.

                                                                                                               0.
                                                                                                          .  0

                                                                                                          .  0
                                                                                                          •
                                                                                                          .  0
                                                                                                                     0 .
39
129.0
60.81
                                                                                   f 8.2

-------
         8.  WHOLE  Equation
CD
I
ID
4
8
9
11
15
16
17
19
20
22
23
25
28
31
33
34
35
36
37
IN"7 20
ACTUAL
74.30
182.3
937.6
312.5
333.7
432.9
.0
425.0
1871.
1399.
300.6
646.0
361.4
570.4
671.0
156. 5
256.5
9.300
732.1
I1 P. I!*
FITTED
389.2
17.65
1156.
-165.6
-11.39
563.2
108.5
652.7
1099.
672.5
576.9
322.7
511.5
53. P9
953.5
382.3
769.9
257.0
547.2
                              I1 P. I \CFTu;. UNIVERSITY
                                                             TiF
                                                                                v L « S T CI  f F A I: 'J I S T ,  1'.
                                              PLOT  OF  ACTUAL(*> Al>0  FITTf.C( + >  VALUES
                                                                                                        CF
+ * Ifa5.
«• -219.
+ * 478.
+ * 345.
* + -130.
* + -10V.
* + -226.
+ * 7 7 2 .
+ * 726.
* * -?7f .
* * 3?3.
* + -150.
+ * 517.
+ -282.
* + -22f..
* + -513.
+ -248.
* * 165. .
0
.
•
. 0
.
. 0
.
.
.
. 0
0
0 .
. 0
*
                                                                                                                       I .C
                                                                                                                        .   C
                                                                                                                      0 .
                                                                                                                                . 0
       39
10.90
-500.fi

-------
    9.  OTHER Equation
    LINE   21
                PRINCETON UNIVERSITY
                                           TSP
VERSION1 OF AUGUST, 19b9
                                      PLOT  OF  ACTUALC*)  A\P FITTEC(*> VALULS
                                                                                          f-LOT Cf  RFSILUALS(O)
ID
   8
   9

  11

  15
  16
  IT

  19
  20

  22

  23

  25

  28

  31

  33

  34
  35
  36
  37

  39
ACTUAL

 486.1

 308.8
 763.6

 866.2
 377.0
 268.7
 558.5

 532.6

 82.20

 389.8

 274.6

 414.0

 282.9

 146.9
 316.7
 58.30
 1016.

 45.40
FITTED                                                         RESIDUAL

 100.1           +        *                                       3P6.

 116.4            *   *                                           152.
 680.0                      **                                   -7b.3

 446.1                   •»      *                                  317.

 588.9                      +     *                                ?77.
 842.2                  *        *                                -465.
 133.8            +      *                                        ?01.

 1174.               *                  +                        -905.
 1510.                     *                 +                  -552.

 159.9            +       *                                       373.

 117.1           *+                                             -34.9

 1233.                  *                +                       -M3.

 510.2                *    +                                      -236.

 940.8                  *          +                              -527.

 605.6                *      *                                    -223.

 2343.            *                                          +  -.220E+-04 0
 724.3                *       *                                  -tCtt.
-227.4     «•    *                                                286.
 220.0               +               *                            796.

 61.86          *+                                              -16.5
                           .0
0.0
 •

 .0
0.

 . 0

 . 0
 •
 . 0
                                                                                                                    0 .
                                0.

-------
                                APPENDIX E

                         SUPPLEMENTARY INFORMATION
Thomas McCurdy
EPA Project Officer
                                       E-l

-------
I.  General
     The EPA Technical  Committee mentioned in the Acknowledgements on p.  iii
included the following  personnel:
     Martha Burke              Office of Transportation and Land Use Policy
     J. David Foster           Control Programs Development Division, OAQPS
     Walter Issac              Municipal Construction Division, OWPO
     Thomas H. Pierce          Office of Land Use Coordination
     John L. Robson            Strategies and Air Standards Division, OAQPS
     David Sanchez             Control Programs Development Division, OAQPS
     Carol Wegrzynowicz        Municipal Construction Division, OWPO
OAQPS stands for Office of Air Quality Planning and Standards and OWPO means
Office of Water Program Operations.
     A number of Technical Committee members reviewed a draft of this report,
A consensus comment was that a number of items were not fully explained in
the contractor's text.   Thus, the purpose of this appendix, written by the
EPA Project Officer, is to supply needed supplementary information.  The
material  is organized around the worksheets.   Information  is presented on
the following topics:
     1. how to obtain input variables for the  predictive model  (Worksheet  1)
     2. how to estimate the sensitivity of the model to changes in  input
        variables  (related to Worksheet 1).
     3. how to distribute predicted  land use within the area of analysis
         (related to Worksheet 2).
     4. how to estimate residential  and commercial breakdowns  (Worksheets
        3 and 4).
                                   E - 2

-------
II.   Simplifying the Analysis  and Obtaining  Input Data
     To use the nine predictive (Worksheet 2)  and six disaggregated (Worksheet 4)
 equations requires the user to come  up  with  values  for 52  variables.   (Additional
 input data are required for VMT and  emissions prediction equations).  Some datum
 will  be easy to obtain because it is available from the wastewater treatment
 facility owner or operator; this is  particularly true of "major project"
 variables, such as collection system capacity and peak flow.   Other data may
 be  difficult to obtain, especially non-OBERS  future year socio-economic data.
 It  is the purpose of this  Section to provide  some help on  how a user can obtain
 needed input data.  A simplifying procedure  is also provided  to reduce user
 effort to a minimum.  This procedure,  however, correspondingly reduces the
 amount of confidence that  can be placed in  GEMLUP predictions.  Using the
 full-blown analysis allows the user  to say that there is a 90 percent pro-
 ability that the predicted land uses will  be within ± (x)  percent of their
 true value (the percentages being obtained from optional  Worksheet 5 as a
 confidence interval).  If any short-cut procedure is used, on the other
 hand, nothing can be said about the  confidence  level of a prediction.  The
 tradeoff  in using a simplified procedure is between time and  rigor.   Only  a
 user can  decide which  is more important in his or her situation.
     Because many input variables are used  in more than one predictive or
 disaggregation equation,  the  number  of input variables required varies with the
 combination of predictive equations  used.   Although there are  numerous  ways
 to combine the 9  land  use predictive equations, there is one  way that is
 preferred.  The preferred order of  land uses  is:
                                 E - a

-------
     1.   residential                            6.   education
     2.   manufacturing                         7-   wholesale
     3.   commercial                             8.   other
     4.   office-professional                    9.   recreation
     5.   highways

    Something should probably be said about highways being placed fifth in the
list.  The causal analysis (Volume I) confirmed that Highways is an important
land use in inducing other development.  A question may  arise, then, in the
reader's mind concerning the reason why Highways is not ranked higher than
fifth.  The reason is that the above listing strikes a balance between
theoretical importance and air pollution emissions importance -- for purposes
of GEMLUP, highways do not have any direct emissions.  Motor vehicular
emissions due to vehicles using highways are accounted for  in the VMT model
and  are not tied directly to the amount of highway land use.  Therefore, from
the  viewpoint of an inventory for direct emissions, highways are an  important
land use.
     To use the residential predictive equation requires that 12 input vari-
ables be obtained; of these, nine are considered to be sensitive and the
remaining 3 variables are considered to be very sensitive.   (The relative
sensitivity of a variable depends upon how much it affects  predicted output
vis-a-vis other variables.  This is discussed  below under "sensitivity analysis."
Note that all residential variables are defined to be sensitive because of the
critical position that residential predictions play in subsequent analyses).
If manufacturing predictions are made after residential predictions, only four
additional variables  are required.  One of these is sensitive and another is
very sensitive.  Five of the variables obtained for residential predictions
are  also used in the manufacturing equation.
                              E  -  4

-------
    Marginal new information requirements for each equation, in the order,
listed above, are summarized in Table E-l.   Also included are the disaggregated
equations, listed in the only order that is logical.   The first five land
uses listed in E-l  are considered to be "core" uses that should be predicted
for each application.   The remaining land uses are considered to be of secondary
importance both from an emissions (practical) and theoretical point of view.
    Given the extensive data requirements of the secondary and disaggregated
equations and the relative insensitiveness  of their predictive land uses to air
pollution emissions, it probably is not worth the effort to obtain data needed
to use the equations if the analysis is being done solely to estimate area of
analysis emissions.   In this instance, a user can save a lot of time by just
predicting  core  land uses and "scaling" secondary land uses from them by simple
ratios.  The ratios, found below in Table E-2, are developed from raw data
obtained during GEMLUP case studies.  They are based on mean values for the various
land uses.  "Redundant" ratios are provided so that more than one  core  land
use can be used to estimate a particular secondary land use.  The differing
estimates could then be averaged to obtain a representative value.
    To provide an example, suppose the residential predictive equation was used
and it estimated that the area of analysis would have 10,000 dwelling units ten
years    after the waterwater facility of interest is constructed.  Using
                                                         2
the default ratios of Table E-2, approximately 710,000 ft  of educational
building area might be expected (0.071 x 10,000 x 1,000).  Suppose also that
                                                                      2
the manufacturing equation was used and it predicted that 2,000,000 ft  of
manufacturing would exist in the area of analysis by the end of the 10 year
period.  Using the education-to-manufacturing default ratio of Table E-2
                                   E-5

-------
                                 Table E-l
      MARGINAL INPUT INFORMATION REQUIREMENTS TO RUN THE PREDICTIVE
                     AND DISAGGRATED EQUATIONS
        Land Use Type
CORE
   Residential
   Manufacturing
   Commercial
   Office - Professional
   Highways

SECONDARY*

   Education
   Wholesale
   Other
   Recreation

DISAGGREGATED+

   % Single Family
   % Multi-Family
   % Two-Family
   % Large Commercial
   % Medium Commercial
   % Small Commercial
# of New
Input
Variables
Required
   12
    4
    7
    2
    2
    8
    4
    5
    4
    4
    3
    0
    4
    3
    0
# of New
Sensitive or
Very Sensitive
Variables
   12
    2
    3
    0
    0
    1
    3
    4
    3
    1
    0
    0
    0
    0
    0
# of
Previously Used
Input Variables
Required
    0
    5
    5
    6
    6
    4
    5
    3
    4
    6
    4
    0
    2
    4
    0
Notes:  *New variables are defined to be those not included in one or more
         of the core land use equations, regardless of whether or not the
         variable appears in a secondary land use equation.

        +New variables are defined to be those not included either in one or
         more of the core land use equations or in one or more of the preceding
         disaggregated equations within the same class (i.e., residential and
         commercial).
                                 E-6

-------
                                  Table E-2

      DEFAULT RATIOS TO RELATE SECONDARY LAND USES TO CORE LAND USE TYPES
Variables                             Ratio

.  .  .  with RESIDENTIAL

      Education                       0.071
      Wholesale                       0.054
      Other                           0.052
      Recreation                      0.024

      Manufacturing                   0.207
      Commercial                      0.212
      Office - Professional            0.058
      Highways                        0.005

.  .  .  with MANUFACTURING

      Education                       0.367
      Wholesale                       0.264
      Other                           0.252
      Recreation                      0.115

...  with COMMERCIAL

      Education                       0.355
      Wholesale                       0.258
      Other                           0.246
      Recreation                      0.112

.  .  .  with OFFICE - PROFESSIONAL

      Education                       1.317
      Wholesale                       0.948
      Other                           0.903
      Recreation                      0.413
Units1
1,000 fr/du
acres/du

1,000 ft2/du


lane miles/du
1,000 ft2/!,000 ft2
acres/1,000 ft'
1,000 ft2/!,000 ft2
acres/1,000 ft'
HOOO ft2/1,000 ft2
acres/1,000 ft'
Note:  Abbreviations used in the units are:
        p
      ft  = square feet
      du =  dwelling unit (housing unit)

      Note that the unit given as 1,000 ft2/!,000 ft2 could be stated as
      ft?/ft2.  The first choice is shown because it has the same dimensions as
      the land use prediction equations (Worksheet 3).
                                  E-7

-------
                                    p
results in an estimate of 734,000 ft  of educational building area (0.367 x
2,000 x 1,000).  Averaging the two estimates gives a "best-guess" prediction of
722,000 ft2 of educational building area.
    Default value percentages for the two disaggregated  land uses, residential
and commercial, are presented in Table E-3.  While the estimates represent
mean values for the 40 cases used in GEMLUP, local data  should be used
wherever possible.  Since the data are relatively easy to obtain, particularly
for residential, the effort probably will be well worth  it.
    To use the disaggregated estimates, simply multiply  the decimal equivalent
of the percentage value shown in Table E-3 by the predicted value "output"
by the commercial equation of Worksheet 2 multiplied by  the area of analysis/
10,000 acres factor needed to convert density figures to totals.  Suppose, for
                                                                 •5   2
example, that the Worksheet 2 commercial prediction is 2,000 x 10  ft 710,000
acres and the area of analysis  is 20,000 acres.  The factor is therefore 20,000
acres divided by  10,000 or 2  (forget  its units:  the figure essentially
represents the number of  10,000 acres  parcelsper area of analysis).  Multiplying
2,000 x 103 ft  by 2 gives 4,000,000  ft  of commercial development building
area within the  area of  analysis  as  the  predicted value. Using  the default
percentages of Table E-3  results  in  the  following breakdown by size of
                                      2                                2
commercial establishment.  428,000 ft of  large  commercial, 420,000 ft  of
medium commercial, and 3,152,000  of  small  commercial  (4  x  10  times 0.788).
    If the  user  is  going  to predict with  core equations and derive secondary
and disaggregated  land uses from  them, only the  variables  listed in Table E-4
will have  to  be  obtained  as  inputs to Worksheet  1.  The  variables appear  in  the
order shown on Table 2-5  (p.  2-20).   There are now  27 variables  that have to be
obtained instead  of the  original  52  listed in Table 2-5. That table should  be
consulted  for  a full description  of  the  variables.
                                   E-8

-------
                                Table E-3

               DEFAULT PERCENTAGES FOR DISAGGRATED LAND USES
                                              % of Total Land
Land Use                                      Use in the Category

Single Family Detached                             73.5
Multi-Family                                       10.6
Two Family^                                        15.9

     Subtotal                                     100.0
                2
Large Commercial -                                 10.7
Medium Commercial                                  10.5
Small Commercial4                                  78.8

     Subtotal                                     100.0
Notes:    Included are single-family attached, single family quadplexes,
          and the like

        2                                                    2
          Commercial development with floor area > 100,000 ft

        3                                                                     2
          Commercial development with floor area between 50,000 and 100,000 ft
        4                                                    2
          Commercial development with floor area  < 50,000 ft
                                     E-9

-------
                                    Table E-4
          LIST OF VARIABLES NEEDED TO PREDICT CORE LAND USES
            Variable
          Equation(s)
County Area
Vacant Developable
Vacant Undevelopable
Zoned Office
Zoned Industrial
Limited Access
Population Growth
Office Vacancy
Median Price
Future Employment
Airport Distance
Track
Area of Analysis
Sewered Land
Interceptors
Collection Capacity
Peak Flow

SMSA Area
Tract Area
Dwelling Units
School Kids
Nonmobility
Median Income
Government
Current Employment
Manufacturing Workers
Drivers
Residential
Residential, Manufacturing, Commercial
Residential, Manufacturing, Commercial
Manufacturing, Office-Professional
Office-Professional
Commercial
Office-Professional
Manufacturing
Residential, Highways
Commercial
Manufacturing
Manufacturing, Office, Highways
All Core Equations
Commercial
Hiahways
Residential, Commercial, Highways
Residential, Commercial, Office -
Professional, Highways
Commercial
Residential, Manufacturing, Commercial
Commercial, Office-Professional
Commercial
Residential
Residential, Highways
Highways
Commercial
Residential, Manufacturing
Residential
                                 E-10

-------
III.  Obtaining Needed Input Variables for Worksheet 1
      Input data must be obtained for variables listed  in Table 2-5 in order
to use all of the predictive equations.  Basically, data are needed for three
time periods (time of wastewater facility completion, t, and five and ten years after
that ( t+5 & t+10 )  and four geographical area (area of anlysis, census tract(s)
most representative of the area of analysis, county, and SMSA).  Because t
itself will most likely be in the future, there is a problem of getting values
for these supposedly current time period variables.  The material to follow is
designed to help solve this problem.  The areal issue will be discussed first.
      The area of analysis is the legal service area of the wastewater treatment
facility.  It generally includes only one watershed, but may include contiguous
areas whose sewage is conveyed to the natural drainage basin by some mechanical
means.  Data needed for the area of analysis is entirely "physical" such as
geographical area, vacant undevelopable land, and amount of zoned residential
land.  These data mostly come from maps.
      Census tract socio-economic data are required.  These data come from the
U.S.  Bureau of Census's Census of Population and Housing;Census Tract series.
Updating this information will have to be done; some guidance on doing it appears
below.  If the area of analysis is not tracted, municipality-level data are
substituted for  census tract data.  In that case, the data come from the Census
of  Population and the Census of Housing:-  Metropolitan Housing Statistics.  Again,
the data will have to be updated.
      The user will have to decide what census tract(s) are to be used.  A rule-of-
thumb is to use  census tracts that represents the area of analysis most closely.
This  simple  rule is difficult to apply in practice, as there are a large number
of circumstances possible, such as the area of analysis is:
                                   E-ll

-------
      1.  fully tracted (in one or more municipalities).
      2.  partially tracted, with the remaining area in another municipality
          (which itself may be tracted).
      3.  untracted in an incorporated area.
      4.  untracted in an unincorporated area.
      By far the most common circumstance encountered during GEMLUP data gathering
is the first case -- the area of analysis was fully tracted.  Often, however, a
number of municipalities were involved (or one municipality and tracted areas in
the county).  This situation' is depicted in Figure E-l.
      Required census tract socio-economic data would definitely be obtained
for the following census tracts:  4 to 6 and 8.  Less than half of census tracts
1 and 9 are within the area of analysis, so they would not be used.  Whether data
for census tract 2 would be used or not depends upon the situation.  If most of
the people living in the tract are located within the area of analysis or_ if
population is distributed evenly throughout the census tract, the tract should
be used.
      Note that just about  all socio-economic variables are used on a per unit
area basis (or are relative variables to begin with; i.e., are percentage variables),
This is true for all census tract variables, which means that non-relative census
tract data are divided by tract area to derive the per unit area (density) variable
used in GEMLUP predictive equations.  Examples are:
      1.  Manufacturing Density, or manufacturing employment divided by census
          tract area.
      2.  House Density, or the number of housing units divided by census
          tract area.
      Census tract area will have to be obtained directly off of the maps included
in the SMSA Census Tract report (with a planimeter preferably) if the data are
                                   E-12

-------
                        Figure   E-l
HYPOTHETICAL EXAMPLE OF A  FULLY-TRACTED AREA OF ANALYSIS
r
                        1
                            Municipality A
                         MMM» * •MVM* *» ••• • J ^ *M
                  $%
                  It;-':';:1

                 *.v;.;:'.:
                                   ..'.'.»
                                   '•.V'v.v'.vV*
II
ill
             <—t
                                                    Census Tract
                              ^^_i^ ^b ^^i

                                   ^
  Municipality  B
                  Of
        W$%  Analysis

        *i
                                                County

                                                Census

                                                Tracts
                                E-13

-------
not already available.  Many city and regional planning agencies have census
tract area data because of its use in transportation and housing plan develop-
ment.  These agencies should be checked first for the area data.
      Area information is also required for the municiaplity (but only if census
tracts are not used), the county, and the SMSA.  Area information for these
jurisdictions are most easily found in the latest County and City Data Book.
The information is also contained in the Census of Population.
      We turn now to the problem of obtaining estimates for variables in the
future.  As mentioned in the opening paragraph of this section, basically three
time periods are involved:  t, when the project "opens,"  t+5,  five years after t,
and  t+10, or ten years after the project opens.  As we shall discuss below,
there probably will not be a good source of data  (certcintly no single source) for all
variables for the needed time period, since t and t + 10 could  be any year  and
projected data are often only done for decennial  or, at best, quinquennial  periods.
Consequently, data available will have to be manipulated to obtain needed values.
Some assistance follows on how this can be done for each variable listed  in Table 2-5.
      V.  County Area:  area of the county in square miles for  time t
          (and t + 5).  This variable does not change, except for minor,
          infrequent  boundary adjustments.  Use the value in the latest
          County and  City Data Book or other Census material, or obtain
          it from the county or regional planning agency.  This variable
          is required  in the Residential equation, so it must be obtained
          even if the  simplified predictive approach is used.   It is a
          "sensitive" variable.
      2.  Vacant Developable:  vacant developable acreage in the area of
          analysis  (area) in time t.  This variable and the next, vacant
          Undevelopable, are related by the following simple relation:
                 (1)            (2)            (3)            (4)
            Total  Area  =    Developed    +    Vacant      +  Vacant
            Acreages         Area            Developable    Undevelopable
            (Area of        Acreaget        Area          Area
            Analysis)                       Acreaget       Acreaget
                                    E-14

-------
     Because it is probably easier to get estimates for variables
1,2 and 4, vacant developable can be obtained as a residual.  The
fact that time t will be in the future and that variables 1 and 4
do not change over time, means that vacant developable is most easily
obtained by estimating developed area acreage (2) for time t and
substracting it from the quanity  [(1) - (4)] .  (How 1 and 4 are
obtained is explained later.)

     Developed area acreage for time t is estimated from developed
area coverage for a current or past time period (t1) updated by some
proportionality factor.  Development in time t1 is obtained from
aerial photographs (as was done in GEMLUP case studies), or from
existing land use maps or other planning studies.  The portionality
factor applied to development in time t1 that intuitively makes a
lot of sense is a population/density ratio of the form:
         population.
         (estimated)
               population^,

               (known)
         family size
         (estimated)
               housing unit density.
               (estimated)
This ratio is called the "factor" in subsequent discussion,
simple formula for developed acreage becomes:
                                       The
      Developed
      Area
      Acreage^
Developed
Area
Acreage^,
    Factor
+   (Acres)
     The factor is trend related as it includes future residential
density and family size (in time t), which might change from the base
year (t').  However, since the time difference between t and t1 will
probably never be greater than 10 years, most likely there will not
be a big change in the variables in question for the t1—> t time period.
     Population of the area of analysis in times t and t1 must be
obtained.  Because the area of analysis will probably not be the
same as a census area, population data must be apportioned somehow.
(Population-^ will probably be provided in the facility plan documentation,
but is assumed not to be available for this discussion.)  There are a
number of ways to apportion population data and a number of places
                          E-15

-------
where the information could come from.  Census block and
tract data could be apportioned to the area of analysis, if
available, to estimate population in time t'.  The past annual
population change rate could then be used to estimate population
in time t.  Traffic zone data may also be used for this purpose,
as may watershed-based pouplation projections, if available.

     Because there are a lot of possibilities, it is not possible
to cover them all here.  Local and regional planners deal with this
issue a lot; they should be able to provide assistance in popula-
tion apportionment.

     A numercial example using the factor should clarify the
procedure and tie loose ends together.  Population in time t1 is
estimated to be 10,000.  It was apportioned from census tract and
block data.  Developed land associated with this population is
about 2,500 acres (@ 4 housing units per acre).  Population^ is
estimated to be 20,000; this figure comes from the facility plan.
Family sizet is thought to be 3.1 persons per family (actually
3.1 persons per housing unit), which is the latest estimate of family
size available from the regional planning agency.  It is not expected
to change between now and t.  Housing unit densityt is expected to be
about 4.5 units per acre based on recent development trends,

     Substituting the above into the formula for developed area
acreage^ gives:
Developed
Area
Acreage^
2,500 acres  +
(20,000 - 10,000)  people
                     3.1 people
                          unit
                4.5 units
                     acre
              =  2,500  acres  +  717  acres

              =  3,217  acres
     This value would  be  subtracted  from  the  known  quantity  of  (area
of analysis) minus  (vacant  undevelopable  area)  to obtain  Vacant
Developable.
     A totally different  tack  could  be  taken  to come  up with estimates
of  vacant developable  land. This  method  is  also based on estimating
developed land in some past or current  time  t1, but is more  direct.   It
simply entails determining  how much  land  is  being subdivided annually
in  the area of analysis and multiplying it by the time period (t  - t1).
                           E-16

-------
    For example, say that 23 acres of land are being developed  annually
    in the area of analysis -- on average, and this rate is expected to
    remain pretty constant.  For an 8 year period,  this means that 184
    acres will be developed.  This has to be added  to  the already
    developed acreage to get total developed land in time t.  The subtractions
    mentioned above would be required to get Vacant Developable.
         Vacant Developable is used in the Residential  and two other "core"
    predictive equations.   It is a sensitive variable.

3.  Vacant Undevelopable:   vacant and undevelopable acreage in the area of
    analysis in time t.  The variable appears in the same 3 core equations as
    Vacant Developable.   It is a very sensitive variable, so care should be
    used in estimating its value.

         Vacant Undevelopable is a semi-fixed physical  variable.  It should
    include flood plains,  steep slopes, quarries, and publically owned land
    that won't be permanently developed.  Datum can be  obtained from planning
    agency studies of natural constraints (or physical  features) or from U.S.
    Soil Service county soil surveys.  The data may have to be scaled off of
    maps.

4.  Zoned Residential:  acres of land zoned for residential use (all types)
    in the area of analysis in year t.  It is a sensitive variable but does not
    appear in any of the core equations.

         The variable can  be estimated from a current (time t1) zoning map, or
    maps if more than one  municipality is involved.  The datum probably will
    have to be scaled off  from the map(s).  The municipality comprehensive
    plan(s) should be checked to determine if there is  more land on the plan
    designated as residential than currently zoned for.  If so, the situation
    should be discussed with local or regional planners to determine what the
    probable trend in zoning will be between t1 and t.

5.  Zoned Office;  acres of land in the area of analysis zoned for office,
    including professional office  development in year t.  The variable apepars
    in two core predictive equations, and it is a sensitive variable.  It is
    obtained in the same manner as Zoned Residential.

6.  Zoned Industrial:  acres of land in the area of analysis zoned for
    industrial development in year t.  It appears in one core equation; it
    is not a sensitive variable.  Again, it is obtained in the same way as Zoned
    Residential.

7.  Onsite Restrictions:  a coded variable to indicate what governmental
    restrictions are expected to be placed on on-lot sewage disposal between
    the time t and t + 10.  This variable is not sensitive nor does it appear
    in a core equation.
                                E-17

-------
         The datum may be  available  in the  208 Plan  (Areawide  Wastewater
    Management  Plan).  The  director  of a  local or regional sewer  authority
    may  hazard  a  guess as  to what  the future will bring  in the way of
    on-lot  disposal restrictions.  Knowledge of past restriction policies
    may  help to ascertain  what  the future will bring.  For example,  if
    the  policy  is or was that all  on-lot  disposal is prohibited if housing
    densities reach 2 per  acre,  and  future  densities are expected to be
    greater, then the variable  would be coded as 4.

         A  problem may arise  if more than one policy is  expected in  an
    area of analysis.  The coding  rule that should  be  used in  this and
    other ambiguous cases  is:   use the most restrictive  category applicable
    to most of  the area  of analysis  most  of the time.Also, severity of
    restriction has precedence  over  time  but not area.  Suppose, for example,
    that most  ( >50%) of an area of  analysis will not  be allowed to  have
    on-lot  disposal for  six of  the 10 years.  This  is  a  straightforward  case,
    and  Onsite  Restrictions is  coded 4  (see Table 2-5).  If  the disposal
    restriction will only  apply for  two years on most  of the land, the
    variable is still coded 4  (Restriction  Years goes  from 6 to 2, however.)
    Finally,  if most of  the area will have  no restrictions during the ten
    year period,  Onsite  Restriction  is coded 0  - even  if it  had more severe
    restrictions  for some  of the area of  analysis some or all  of the time.

 8.  Restrictions  Years:   the number  of years of time between time t  and
    t +  10  that the  coded  Onsite Restrictions condition  is expected  to exist.
    The  variable  is  not  sensitive  nor  is  it in  one  of  the core equations.
     It is obtained  in the  same  manner as  Onsite Restrictions.

 9.  County  Interchanges:  the  number of  limited access interchanges  (regardless
    of functional classification of  the  roadway) expected in t + 5 in the
    county  containing most of  the  area of analysis.  (If the area of analysis
     is split evenly  among  two  or more counties, then all counties should be
    used.   Other  county  level  variables  must reflect the multi-county situa-
    tion also,  however.)

          County Interchanges is not  in  a core  land  use predictive equation.
     It is  not a sensitive  variable either.

         Datum  for the variable will usually come from the regional
     planning agency.   If the area  of analysis  is  in an urbanized  transportation
     planning area having a so-called 3C  planning  process, the  figure will  be
     available as  part of the long-range  programming element.  If  not, the  state
    transportation  agency should be  consulted.  Municipality comprehensive
    plans,  or transportation planning  programs, may also be  of help  in  some
     areas.   The City Engineer will be  of help  in  those cases.

10.   Limited Access:  the  number of  limited access  interchanges  (regardless
     of functional classification of  the  roadway)  expected  in t +  5  in  the
     area of analysis.
                                 E-18

-------
          This variable is not sensitive,  but does  appear in the
     Commercial  core equation.  It  is  obtained in  similar fashion to
     County Interchanges.

11.  Transit Stops:   the number of  bus and commuter rail  transit stops expected
     in t + 5 in the area  of analysis.  The variable is  not sensitive nor does
     it appear in a  core land use equation.

          Transit Stops may be obtained from a regional  planning agency,  or
     other agency, doing the 3C Plan because a transit element is part of
     the planning process.  Probably,  however, the  bus or rail company will
     have to be  contacted.  This was the way the datum was obtained most  often
     in the case study phase of GEMLUP.

          It will be difficult to get  an estimate  of transit stops for the
     future.  If the transit company or authority  does not know what will
     happen in t + 5, use  data for  the present adjusted  by what the recent
     trend in line closings has been  in the area of analysis.  If there is
     no service  in the area now, try to find out if any  is being seriously
     discussed.   If  existing housing densities are  low,  probably there would
     not be any service by t + 5 even  if a transit  company is operating in near-
     by locales.

12.  County Growth:   the percent change in county  population projected for
     the time period t to  t + 10.  This variable is sensitive but is not  used
     in a core equation.

          Datum for  the county containing  most of  the area of analysis is used.
     If the area is  fairly evenly split between two or more counties, all
     should be used.  (If percentage  changes are used, a weighted average must
     be derived.  See a standard statistical textboook for information on how
     to compute a weighted average. The problem can be  avoided by adding
     together the actual population estimates for  the two years and computing
     the percentage  change for the  total.)

          County population estimates  are  available from many sources.  The
     best might be the regional planning agency, but county planning departments
     and the state community affairs department are other possibilities.   The
     estimates should be those "officially recognized" for air quality maintenance
     planning purposes or 208 planning.  (If there are differences between these
     two estimates,  take an average.)

          Probably decennial or quinquennial projections will have to be
     manipulated to  obtain population  estimates for t and t + 10.  A linear
     interpolation between years is probably adequate, unless population  levels
     are expected to change dramatically in the counties.  In those instances,
     the projections should be plotted and a smooth curve drawn to connect the
     point estimates.  Year t and t +  10 projections can then be read off the curve,
                                  E-19

-------
13.  Future Population:   projected SMSA population for the year t + 10.
     If the area of analysis is not in a SMSA, use projected county popula-
     tion (see County Growth).  The variable is very sensitive, but does not
     appear in any core equation.

          Probably the best source of Future Population is OBERS.  The
     information is also often available from a regional planning agency.
     If OBERS is used, the data will have to be interpolated to obtain year
     t + 10 projections (see County Growth).

14.  Population Growth:  the projected percentage change in SMSA population
     between t and t + 10.  Population Growth is obviously related to Future
     Population:

     Population    _    Future Population.  lr,   -  SMSA Population.
     Growth        "                     l+lu                     r
                                  SMSA Populationt



          Probably the best place to obtain an estimate for SMSA Population^
     is OBERS.  Thus both input variables for Population Growth could come
     from the same place, which has some advantages, such as:  fewer sources
     to locate, fewer assumptions that must be considered, and compatability
     of inputs and format.

          The variable is not sensitive, but does appear in one core equation.

15.  Office Vacancy:  percent of office space located in the area of analysis
     that is expected to be vacant in year t.  Office Vacancy is in one care land
     use equation.   It is not a sensitive variable.

          Obtaining  an estimate of relative office vacancy for a future time
     period will be  very difficult, unless there  is a sophicated economic
     planning effort underway in the region.  Since this is uncommon, probably
     the best that can be done is call a local commercial realator, or building
     manager, and determine what the approximate  vacancy rate is now and what
     the recent trend has been.  Because office vacancy is negatively correlated
     with median income, perhaps a current vacancy rate could be combined with an
     income proportionality factor to estimate time t office vacancy.  However,
     the effort to do so may not be worth the increase in precision of the
     variable estimate.

16.  Future Houses:  the projected number of housing units expected in the SMSA
     in time t+10.   If the area of analysis is not in an SMSA, the projected
     number of housing units expected in the county in t+10 will be used.   (If
     more than one county is involved, proceed as discussed in  12.^  County
     Growth.  If more than one SMSA is involved,  an unlikely situation, use the
     SMSA that most  of the area of analysis is in.)
                                   E-20

-------
          Future Houses  is  not  a sensitive  variable  and  does  not  appear
     in a core predictive equation.

          Datum for the  variable may be  in  a regional  or local  housing
     plan.  If so,  interpolation will  most  certainly have to  be done
     for the year t+10.  The  datum could also be  obtained by  using  the
     OBERS projection of SMSA population and multiplying it by  a  housing
     unit per population factor (the reciprocal of average family size).
     (See also the  discussion of family  size in   2., Vacant Developable.)

17.   Median Price:   the  projected median price of one acre of vacant,  "raw"
     residential land in the  area of analysis in  year t.  This  is a sensitive
     variable and it appears  in two  core land use equations,  including
     Residential.

          It will be very difficult  to get  an estimate for Median Price.
     Possible sources are asking a local developer what  the going price of
     raw land is now and how  he thinks it will change by time t.   A current
     land price estimate could  be "inflated" by  a fixed  annual  percentage also,
     if the user can buy the  argument that  land  values go along with (but do
     not rise faster than or  lag behind) the inflation rate.  The projected
     inflation rate itself  could come from  a general U.S. projection put out
     by the Treasury Department or other federal  government agency.

          There is  another  possible  solution. Median Price is  not used in
     isolation in any predictive equation -- it  is part  of the  Land Cost
     variable, which is  Median  Price divided by  Median Income (see below).
     If the user wants to make  the assumption that these two  items vary at
     the same rate, which  is  almost  identical to  assuming that  Median Price
     increases as does inflation, then a current  or past Median Price to
     Median Income  ratio could  be used for  time  t also.   In that  case, the
     user would first get Median Income  estimate  for a specific year and
     then get a Median Price  estimate for the same year.  The ratio, which
     was 0.537 on average  and varied between 0.058 and 2.546  for  the 40 GEMLUP
     cases, would then be  used  as Land Cost for  time t.   If land  cost is
     changing differently than  income in the area of analysis,  however, this
     approach should not be used.

18.   Future Income:  the projected median family income in the  SMSA in year t+10,
     Unrelated individuals  are  excluded  from this definition.  The variable is
     not sensitive and does not appear in a core  land use equation.

          Future Income estimates may be available from a regional planning
     agency.  It can be obtained from OBERS's per capita income estimate by
     multiplying the estimate by expected average family size (persons per
     family).  This is the  preferred method.  Interpolation of data for the
     correct year  (t+10) probably will have to be done.
                                E-21

-------
           If the area of analysis is not in an SMSA, then a county-level
      income estimate must be obtained.  This may be available from a regional
      or State planning agency.

19.   Future Employment: projected total employment in the SMSA in t+10.  If
      the area of analysis is not in a SMSA use projected county employment
      instead.  The variable is a very sensitive one; it is used in one core
      land use equation.

           OBERS is a good source of total employment estimates.  Interpolation
      will probably have to be done.

20.   Future Medicals: projected SMSA hospital employment in year t+10.  The
      variable does not appear in any core equation and is not particularly
      sensitive.

           Future Medicals should probably be obtained from OBERS estimates
      of future employment or population.  This is done by developing a hospital
      workers-to-total employment or hospital workers-to-population ratio from
      the most recent Census data, and applying it to OBERS projections.  The
      second ratio was relatively constant for GEMLUP case studies, and is
      preferred to the first.  Whichever ratio  is used, the user must assume
      that it will not change over time.

21.   CBD Distance:  the distance in air miles  (straight-line)  from the approximate
      centroid of the area of analysis to the approximate centroid of the nearest
      central business district in a city of 100,000 population or more in year
      t.  This variable is not a sensitive one, nor is it used  in a core equation.

           The nearest 100,000  city in time t must first be identified and its
      CBD located.  The distance between centroids is then scaled off of a map
      with a known scale.  A state highway map  is generally suitable for this
      purpose.

22.   Airport Distance:  the distance  in air miles  (straight-line) between  the
      approximate centroids of the area of analysis and nearest commercial
      airport in year t.  A commercial airport  is an airport with at least one
      in and one out regularly scheduled flight available to the public on a
      fee basis.  Sight-seeing flights are not  included in this definition.

           After identifying such an airport for time period t, scale off the
      distance between centroids on a  road map.  (Perhaps the regional planning
      agency should be checked to determine  if  a new commercial airport will be
      built before time t that may be  closer to the area of analysis than the
      existing airport.)

           The variable is in one core equation (manufacturing), but is not a
      sensitive quantity.
                                    E-22

-------
23.  Track:  miles of railroad track in the area of analysis in time t.
     Note that parallel tracks are counted as two tracks.  Track is a very
     sensitive variable and appears in three of the five core land use
     equations.

          Track is easily obtained from USGS maps.  A local or regional  planning
     agency should be consulted to determine what changes have occured since
     the USGS maps were published and what might be expected by the year t.
     Note that abandoned rail lines that go someplace (i.e., are connected to
     a mainline somewhere) are counted even if they are not currently used.

24.  Area of Analysis:  the area of analysis is the legal service area of the
     wastewater treatment project being studied.  It includes the drainage basin
     for the stream receiving wastes from the project plus other areas connected
     to the treatment plant by mechanical means.  Unit of the variable is acres.
     Datum will be available in the facility plan provided by the project
     developer or owner.

          The variable appears in all land use predictive equations, both core
     and secondary.  It is a very sensitive variable and must be accurate.

25.  Sewered Land:  land area in acres within 5,000 linear feet of the major
     project interceptor sewer(s), if any, in the area of analysis in year t.
     The 5,000 feet distance is omni-directional, but is only within the area
     of analysis.  The variable is sensitive and appears in one core equation.

          The  location and extend of all new interceptor sewers will be in the
     facility plan.  If maps provided with the plan are large enough, Sewered
     Land may be scaled directly from them.  If not, the new sewers will have
     to be plotted onto adequate maps and Sewered Land scaled off of those
     new maps.

26.  Interceptors:  running length in miles of the major project interceptor
     sewer(s), if any, going through relatively undeveloped land in the
     area of analysis  in year t.  Relatively undeveloped is defined to be
     less than one housing unit per acre, and the relatively undeveloped area
     should be at least 1000 feet deep  (i.e., from the interceptor sewer).   In
     other words, if only land adjacent to the pipe is relatively undeveloped
     and the rest is subdivided and developed, there is no relatively undeveloped
     land for  purposes of computing  Interceptors.

          Datum  is obtained by plotting the interceptor sewer onto aerial
     photographs or land use maps and scaling off the running length of pipe
     going through   <1  housing unit/acre land.  This variable  is most easily
     obtained  in conjunction with Sewered Land.

          Interceptors  is in one core equation, but is not  a sensitive variable.
                                  E-23

-------
27.  Collection Capacity:  hydraulic design capacity of an interceptor
     sewer(s) major project in million gallons per day (MGD) in time t.
     If the interceptor sewer(s) is(are) phased, the design capacity
     may be that for the year when the last phase is completed; however,
     all phasing must occur before t+5.  If phasing of interceptor sewers
     occurs within the area of analysis beyond t+5, the GEMLUP model
     should not be used.

          The datum will be available in the facility plan.  The variable
     is sensitive and it appears in three core equations, including Residential.

          If Collection Capacity is not available in the facility plan for
     some reason, it can be estimated by summing hydraulic flow of the major
     project interceptor sewer(s), using nomographs of Manning's formula
     (based on pipe size and slope; see Metcalf and Eddy, Inc.  Wastewater
     Engineering.  New York:  'McGraw Hill Book Company, 1972.).  If slope is
     not known, average values for hydraulic flow based on pipe diameter
     alone can be used (see Facilities Requirements Branch.  Guidelines for
     1976 Update of Needs for Municipal Wastewater Facilities"  Washington,
     D.C.:  U.S. Environmental Protection Agency,1976).

28.  Peak Flow:  anticipated peak flow in the interceptor sewer(s) major project
     in MGD in year t.  If phasing occurs, Peak Flow is for the last phase.
     Again, all phasing must be complete by t+5.

          Peak Flow should be available in the facility plan.  If it is not,
     it can be estimated by scaling known average flow by a population-dependent
     peaking factor.  Such a factor appears as Figure E-2.

          The variable is sensitive and it appears in four core equations,
     including Residential.

29.  Treatment Capacity;  hydraulic design capacity of the wastewater treatment
     plant in MGD in year t.  This information will be available in the facility
     plan.  The variable is not sensitive nor does it appear in a core equation.

30.  Population Served:  the population served by the major project facility in
     year 5.  This information will be in the facility plan.  The variable is
     not sensitive.  It does not appear in any core land use equation.

31.  Project Cost:  the total cost of major project construction (in thousands
     of dollars).  This value will be in the facility plan.  The variable is not
     sensitive, and it is not in a core equation.

32.  Federal Funds:  the federal share of major project cost (in thousands of
     dollars).TTTis non-sensitive, non-core equation variable will be part of
     the facility plan.
                                   E-24

-------
                 Figure  E-2




RATIO OF MINIMUM AND PEAK FLOWS TO AVERAGE DAILY FLOWS
                        40
60
80
100
            Population in Thousands
                         E-25

-------
33.  Phasing:  a nominal variable that indicates if multi-year construction
     phasing of interceptor sewers will occur in the area of analysis.   This
     variable will be in the facility plan.  It is not sensitive and is not
     in a core equation.

34.  SMSA Area:  area of the SMSA in year t in square miles.  If the area of
     analysis is not in a SMSA, county area should be used.  Data for the
     variable is found in the County and City Data Book.  A regional planning
     agency would have the information also.

         SMSA Area appears in one core equation, but it is not a particularly
     sensitive variable.

35.  Tract Area:  area of the census tracts for which socio-economic data are
     obtained.  Units are in square miles.  The data can be obtained by scaling
     the census tract maps, but it may be available from the regional planning
     agency also.

         If the area of analysis  is not tracted, then Tract Area becomes area
     of the municipality or other civil division used for socio-economic data
     gathering.  Municipality area is found in the County and City Data Book.

         Tract Area  is  a very sensitive variable because of its use in many
     GEMLUP socio-economic variables  (as a divisor).  It is in three core
     equations.

36.  Dwelling Units:  number of dwelling units in the census tracts in year t,
     rounded off to  the nearest 100 units.  (I.e., if the value is 5,327 then
     53 is used.)  The  variable is sensitive and appears in two core land use
     equations.

         Values for  the variable  for  a current or past  time period  (t1) can be
     obtained from census tract data.  The time t1 variable can be updated to
     time t using a population ratio of populationt/population^i.  The number
     of dwelling units in t would therefore be obtained thusly:


                Dwelling   =   Dwelling        Popu1ationt
                Unitst         UnitSf    '    Populationt-


     Since both t1 variables are known, only Population^ is required.  A
     regional or city planning agency would be the source for census tract
     population projections.
          This method assumes that the variable of interest (Dwelling Units)
     is linearily related to population.  If that is not true, then a cor-
     rection to the ratio has to be made.  This can simply be done by
     applying another ratio to the population ratio  to  rectify  the  non-
     linearity.  For example, suppose  that  average family  size  has.changed
     radically between  t' and t.  A family  size ratio should  then be used
                                 E-26

-------
    along with the population ratio to estimate Dwelling Units.

             Dwelling  .  Dwelling   .   P°Pu1ationt    .   Family Sizet.
             Unitst       Units^1       Population^       Family Sizet

          An example should elucidate both problem and solution.  Input data
    are as follows;  DU^ would not be known in practice.
            Time               t]_         t.       Units

            DU                100       (150)     Dwelling Units
            Family Size       3.00       2.67     People/Dwelling Units
            Population        300        400      People
    If just the population ratio was used to estimate DUt, Dwelling Units
    would be underestimated by about 17 units.


                      DU.    »   100   .    400   =   133
                        r                 300
    Using the correction factor of family size produces the correct answer.
                 DU.   =    100    .   400    .   3.0CL  =    150
                    T                 300
          The principle of using a correction factor to adjust for non-
    linearity in the dependent variable-to-population relation is used in
    computing many of the variables that follow.


37.  Current Houses:   number of dwelling (housing) units in the SMSA in time
     t:   The variable is not sensitive and does  not appear in any core equa-
     tion.  Current Houses should be obtained in the same manner as the pre-
     vious variable.   Time t1  values would come  from the County and City
     Data Book,  while  t  values would be obtained from regional  planning
     agency projections (its housing element).

38.  School  Kids:   the number of people less than 15 years of age living
     1n  the census  tracts in year t.   The variable  is sensitive and
     appears in  one core equation.

          The  variable is obtained  in  the same manner as variable 36.   Time
    t variables will  have to  come  from regional  or local  planning agency


                                    E-27

-------
projections.  The population ratio factor described earlier she-Id be used,
supplemented by other ratios of family size or the relative proportion of
children in the total population (or both) changes.

     For example, if the age structure of the population is changing, a for-
mula for computing school kids might be:


  School   _   School       Populationt      Relative Proportion of Kidst
  Kids.    "   Kids.,
                            Population^'     Relative Proportion of Kidst.

Using the same numerical example used in 36 above, say the proportion of
school aged children (0-14) in the total population changed from 0.367
in t' to 0.263 in t.  This means that the population is aging.  Multiplying
these proportions by total population results in 110 Schools Kids in t' and
105 in t.  Plugging numbers into the formula gives the correct answer.
          School       110       400       0.263   =   105 children
                                        '   0.367
     The alert reader may question why the time period indices of the cor-
rection factors are different in 36 and 38.  In 36 the time t1 correction
factor is  in the numerator, while in 38 it is in the denominator.  The
reason is  that time indices must cancel as well as variable units.  In 36,
the Family Size factor has units of:


                     Populationf      Dwelling Unitt
                      Dwelling  Unitf    Population^

 (The numerator  is multiplied by the reciprocal of its denominator.)  The
 units now  cancel throughout the original equation.

     In  38,  Relative  Proportion of Kids has  units of:

                      School Kidst        Populationf
                     Populationt     '   School Kidst1


 These units also now cancel  with  the original  equation.


 39.   Vacant Houses:   the  percentage  of  vacant  available  dwelling  (housing)
      units  in  the census  tracts in time t.   The  percentage  is  coded  as a
      decimal;  i.e.,  10% is coded  as  0.1 and  17.3%  is  coded  as  0.173.  Datum
      for  time  t1  would  come  from  census tract  information.  This  could be
      used for  time t, or  an  appropriate planning agency  could  be  contacted
      for  a  more  recent  estimate.  While the  variable  does not  appear  in a
      core equation,  it  is a  very  sensitive variable in those equations
      where  it  does appear.
                                      E-28

-------
40.  Nonmobility:  the percentage of families in time t living in the same
     house as they did in t-5.   The geographical area is the census tracts,
     and the datum comes from Census of Population and Housing:   Census
     Tract books.  The variable appears in the residential  core  equation and
     is sensitive.  Percentages are coded as decimal  equivalents.

          The most recent census value could be used  as a surrogate for
     Nonmobility.  If a comprehensive housing element planning effort exists
     in the region, the variable could probably be estimated from it.  The
     variable will most likely fall between 45 and 65 percent and is thought
     to be relatively constant in the short run.

41.  Median Income:  the median income of families living in the county or
     counties containing most of the area of analysis.   The time index is
     year t.

          A time t' value can be obtained from the County and City Data Book.
     A more recent estimate will probably be available from a socio-economic
     profile done by the regional planning agency.

          Median Income is only used in conjunction with Median  Price (17).
     See the discussion of that variable for more information.  Median Income
     is sensitive and appears in the residential and  one other core land use
     equation.

42.  Current Income:   the median family income of people living  in the SMSA
     containing most of the area of analysis (in time t).  If the area is
     not located in a SMSA, county data are used, which means that Median
     Income  =  Current Income.  The variable is not  sensitive nor does it
     appear in a core equation; in fact, it is only used in disaggregating
     residential predictions into the single- and multiple-family categories.

          If the area of analysis is in a SMSA, then  time t data can come
     from OBERS by multiplying  the per capita income  estimate by average
     family size.  Because OBERS data are in terms of some  specified base
     year  dollars, Current Income should be updated  to account  for infla-
     tion.

          Another source of information can be a regional planning agency
     socio-economic profile.

43.  Poverty:  the percentage of total families in the census tracts with
     income below the H. E. W..poverty level in time  t.  The percentage is
     coded as a decimal.  The variable is not sensitive and is not used in
     a core equation.

          Time t1 values can be gotten from census tract data.  Because the
     variable does not change rapidly, the time t1 figure should be adequate
     for predictive, purposes.  If a more recent estimate is desired, then a
     socio-economic profile done by a public planning agency should be con-
     sulted.
                                      E-29

-------
44.  Index One:  the Consumer Price Index (CPI) for year t.   The Index is
     published for selected major cities in CPI-Detailed Report (Bureau of
     Labor Statistics, U.S. Department of Labor).Use the  value for the
     nearest major city.

          The variable is only used in combination with Index Two.  An
     estimate for time t if it is in the future will not be  available.
     The recommended way to proceed is to use the current CPI for whichever
     Index (One or Two) applies to the later year and go backward in time
     for the other index.  There is no need to update the two indices with
     an inflation rate, because each index would be changed  by the same per-
     centage and the net result would be the original difference.

          The variable is not sensitive and is not used in any core land
     use equation.

45.  Index Two:  the CPJ for the year of federal funding; the year should
     be less than time t.  The variable is not sensitive nor is it. in a
     core equation.  It should be obtained in the manner in 44, above.

46.  Government:  the total amount of county government expenditures in
     millions of dollars in year t.  (Round off $7,761,000 to 7.76, for
     instance).  Use the county or counties containing most of the area
     of analysis.  The variable appears in one core land use equation; it
     is not sensitive.

          It will be hard to obtain data for this variable unless the
     county projects expenditures in advance.  Perhaps the regional plan-
     ning agency might be doing fiscal analyses of future expenditures.
     Probably the only way to get an estimate is to obtain recent county expen-
     diture audits, determine what the yearly trend is by plotting values
     on a graph, and project ahead using the same rate-of-change.  This
     trend analysis will only be approximately valid,  but the estimate
     should be accurate enough for the use to which it is put.

47.  Current Employment:  total SMSA employment in year t.   If the area of
     analysis  is not in a SMSA, use county employment data.  The variable
     appears in one core equation, but is not sensitive.

          SMSA total employment projections are available in OBERS.  Some
     interperlation between OBERS projection years will have to be done,
     however.  If the area  is not in a SMSA, perhaps the information can
     come from a county-level socio-economic profile maintained by a county
     or regional planning agency (such information will be available in a
     3C transportation plan, for instance).  If not, use past census estimates
     projected ahead by a percentage change number used by state or regional
     economic/industrial development organizations.

48.  Unemployment:  the percentage of unemployed workers in. the census tracts
     in year t. (Percentage is coded as a decimal).  This variable is sensitive,
     but does not appear in a core equation.
                                      E-30

-------
          An estimate for Unemployment might be contained in a socio-economic
     profile done by a county or regional  planning agency.  Since it is a
     relative variable;  the current value  could be used by assuming no change.
     A proportion could  also be established between unemployment percentages
     for the area of analysis and some (larger) municipality for which a time
     t estimate is available.  For example, say the census tracts containing
     most of the area of analyses has an unemployment rate of 6% in 1980,
     the latest census year.  The county's unemployment rate was 10% in 1980.
     Year t is 1984, and the state has projected an unemployment rate of 8%
     for the county in 1984.  An estimate  of Unemployment, then is:
       Unemployment  =  Unemployment^
                                              County Unemployment
                                              County
                          (0.06)
              0.08
              0.10
0.048
     Unemployment,
0.05
49.  Office Employment:   the number of employed office workers in the census
     tracts in time t.   The variable is sensitive,  but is  not in a core
     equation.

          Values for time t will  be difficult to come by.   Office Employment
     should be obtained  for a current or recent time period t1 and adjusted
     by a population ratio (see 36).

50.  Manufacturing Workers:  the  number of manufacturing workers living in
     the census tracts  in time t.  This variable appears in two core equa-
     tions, including residential.   It is also a sensitive variable.  Prob-
     ably the easiest way to obtain a value for this variable is to get the
     most recent U. S.  Census estimate and apply a  population ratio to it
     (see 36).

51.  Current Medicals:   the number  of hospital workers living in the SMSA--
     if there is one—in year t.   If the area is not in a  SMSA, use county
     figures.  Time t1  values come  from the U. S. Census,  and a projection
     to time t can be made using  a  population ratio procedure.  See item
     36.  The variable  is not sensitive nor is it in a core equation.

52.  Drivers:  the number of people living in the county who drive to work,
     either as a 'driver or a passenger, in year t.   The value should be
     rounded and coded to the nearest one hundred people;  i.e., 1,371 is
     coded as 14.  The variable appears in the residential core equation;
     it is not sensitive.
                                      E-31

-------
   Drivers may be available in a regional 3C plan.  Probably a
population ratio will have to be used to update U.S. Census infor-
mation.  Exactness is not required, as evidenced by rounding off of
the datum.
                        E - 32

-------
III.  Sensitivity Analysis
     The impact that a variable has on a particular equation depends
upon its contribution to total variance explained in the dependent vari-
able by the equation.  I call this impact "sensitivity."  A sensitivity
analysis was performed for all predictive  equations.* The analysis took
the form of answering the question:  "What is the percentage change in the
dependent variable when all other independent variables are set to their
mean value?"  The higher the resultant change, the greater a particular
variable's impact is on the resultant, and the greater is its sensitivity.
There are other ways to do sensitivity analyses, and one alternative method
is used for the Residential equation for illustrative purposes.
     The user should be careful to obtain good data for sensitive variables
because of their importance.  Assuming that carefulness is related to time
spent in obtaining data and that user time is a scarce resource,  a user should
spend most of his constrained time on important variables and relatively less
on the others.
     Note that this sensitivity analysis has nothing to do with model vali-
dation, which is discussed in Section II.B of the Report.  In particular, this
sensitivity analysis should not be confused with coefficient stability analy-
sis described in that section, which it superficially resembles.
     Results of the sensitivity analyses follow.  If the percentage change
value has a negative sign, it means that a positive change in the independent
variable causes a negative change  in the dependent variable  (an inverse re-
lationship, in other words).
Residential Equation
     The Residential equation  is most sensitive to changes  in Nonmobility,
and is fairly sensitive to all other variables except Driver Density.  The
 *In English units only; results of the analysis in metric units would probably
  be different.
                              E - 33

-------
percent change in Residential (in terms of housing units per acre)  due
to a 10.0% increase in each independent variable, with all  other indepen-
dent variables held constant at their mean value, is:
                        Nonmobility             -9.7
                        Vacant Land             -6.1
                        % Collection Reserve     3.3
                        Land Cost                3.3
                        Manufacturing Density    2.6
                        Driver Density           1.0
     Another way to look at variable sensitivity is to set the variable at
its extremes (low, usually zero, and high, the maximum value recorded
during GEMLUP field data gathering effort) and see what happens to  the de-
pendent variable.  This was done for the Residential equation; results are
presented in Table E-5.
Office-Professional
     The Office-Professional equation is most sensitive to changes  in Office
Zoning, which may be a surprise to jaundiced observers of the local zoning
process in this Country.  No other variable has nearly as important an im-
pact on Office-Professional as does Zoning.  The next most important variable,
Railroads (or the density of railroad track in the area of analysis), is odd.
Its importance is undoubtedly due to its being correlated to a factor that is
also highly correlated with office space, perhaps size or age of city (them-
selves highly correlated), a high order of central place functions, or indus-
trialization.
     The percentage change in Office-Professional due to a 10.0% increase in
each independent variable, with all other independent variables held con-
stant at their mean value, is:
                        Office Zoning           7.9
                        Railroads               2.1
                        House Density           1.5
                        Peak Flow               1.4
                        Population Growth       0.8
                        Industrial Zoning      -0.7
                               E - 34

-------
                              Table  E-5
                PERCENTAGE CHANGE IN RESIDENTIAL DUE
           TO EACH INDEPENDENT VARIABLE SET TO ITS EXTREME
              VALUES & ALL OTHER INDEPENDENT VARIABLES
                       SET AT THEIR MEAN VALUE

   Independent             Range of Dependent             Difference in the
    Variable                 Variable (%)                    Range (%}

   Land Cost                 33.0 - 233.5                       200.5
   % Collection Reserve      65.8 - 201.1                       135.3
   Nonmobility               63.3 - 197.1                       133.8
   Vacant Land               63.8 - 160.5                        96.7
   Manufacturing Density     73.6 - 157.9                        73.6
   Driver Density            90.4 - 161.4                        71.0
As can be seen by comparing data in the text with Table E-5, a variable's
rank order with respect to sensitivity varies from one method to another.
The reason for this is that the relationship between mean and extreme values
varies greatly among independent variables.   Variables widely scattered
around their mean or having a skewed distribution will have a potentially
large impact on the dependent variable if they are misspecified.
  Recreation

       Recreation, in units of acres of land per area of analysis, is most

  sensitive to Vacant Houses (negative relation) and the ubiquitous Railroads

  The percentage change in Recreation because of a 10.0% increase in each

  independent variable, with all other variables held constant at their mean

  value, is:

                             Railroads            6.1
                             Vacant Houses       -6.1
                             Office Density       4.6
                             County Growth        4.0
                             Industrial Zoning   -1.7
                             Treatment Capacity   1.3
                                 E - 35

-------
Manufacturing
     Manufacturing, in units of floor area per area of analysis, is most
sensitive to Office Zoning and Vacant Land.  The second relation is logical
but the first is not.  Apparently there is a factor that affects both in-
dustrial and office land uses (but not industrial zoning).  This factor
could probably be identified by analyzing causal relations found in  Volume
I of GEMLUP (EPA-450/3-78-014a), but this has not been done because of time
constraints.
     The percentage change in Manufacturing due to a 10.0% increase in each
independent variable with all other independent variables held constant at
their mean value is:
                        Office Zoning           7.0
                        Vacant Land            -5.2
                        Railroads               2.7
                        Office Vacancy          1.6
                        Manufacturing Density   1.3
                        Airport Distance        1.2
Wholesale
     Wholesale land use, in terms of floor area per area of analysis is most
sensitive to Employee Ratio and Income.  Of all the independent variables
investigated so far, they are the only ones that produce a greater than one-
 to-one  relative change in a  dependent variable.   Compared with preceeding
equations, Wholesale  is relatively  sensitive  to  all its  independent  vari-
ables, but Employee  Ratio and  Income will  have a much  greater  impact than any
of the others.
     The percentage  change in Wholesale due to a 10.0%  positive change  in
each independent variable, with all other  variables held  constant  at their
mean value, is:            Employee Ratio           47.6
                            Income                  -21.7
                           Unemployment             7.1
                           Driver Density           6.0
                           Office Vacancy           5.7
                           Office Workers           5.3
                              E - 36

-------
 Highways
      Highways,  in units of non-expressway lane miles per area of analysis,
 is the most sensitive to Railroads,  which has to be a surrogate for some
 other variable.  The percentage change in Highways because of a 10.0% in-
 crease in each  independent variable, with all other variables held constant at
 their mean value, is:
                         Railroads               3.9
                         Land Cost               2.9
                         Government             -2.0
                         Collection Reserve      1.3
                         Interchange Density     1.1
 Education
      Education, in terms of floor area per area of analysis, is most sen-
 sitive to Sewer Service.  The percentage change in Education due to a 10.0%
 increase in each independent variable, with all other variables held constant,
 is:
                         Sewer Service           3.7
                         Land Cost               2.0
                         On-site Restrictions    2.0
                         County Growth          -0.9
                         Sewer Costs            -0.5
                         Poverty                  *
 * Less than a -0.05% change.
Commercial Equation
     The Commercial equation is equally sensitive to changes in Kid Density,
Vacant Land, and Sewer Service.  The percentage increase in Commercial (in
terms of floor area per area of analysis) due to a 10.0% change in each
independent variable, with all other variables held constant (at their mean
value) is:
                           Kid Density          -3.9
                           Vacant Land          -3.7
                           Sewer Service         3.6
                           Employment Growth    -0.9
                           Interchanges          0.8
                           % Collection Reserve -0.6
                                 E - 37

-------
Other
     Other land uses include transient lodgings (hotels, motels, and rooming
houses), churches and other religious structures, cultural activities (museums,
theaters, auditoria, and libraries), and indoor recreational facilities
(separate from schools).  Other does not include hospitals, prisons, or
colleges.  The units of Other are in square feet of floor area per area of
analysis.  The percentage change in Other because of a 10.0% increase in each
independent variable, with all other variables held constant, is:
                           Income               -116.2
                          •Vacant Houses        - 18.2
                           Railroads              13.9
                           County Growth           6.6
                           Residential Zoning      5.4
                           Interchange Density    -2.1
     The highly sensitive negative relationship between Income and Other is
puzzling at first glance because, it is commonly said, cultural activities
are directly related to the income and educational level of residents.  An
explanation for the inverse relation can be obtained from the zero-order
correlation coefficient.  Other is significantly positively correlated with
population density, which implies that cultural activities are most often
found in highly developed areas.  This is certainly true of many museums and
auditoria, although theaters are more evenly dispersed throughout a metropoli-
tan area (espcially movie theaters).  Motels and hotels are also found  down-
town, particularly in large cities.  Next, there is a negative, significant
correlation between population density and per capita SMSA income, which
Income measures.  These two factors taken together would result in the negative
relationship between Income and Other that is observed above.
                                E  -  38

-------
Summary
     The preceding analyses of sensitivity were used in developing a list
of very sensitive and sensitive variables.  Variables not on the list were
labeled insensitive or not sensitive variables.  The designations correspond
to those used in Section III of this Appendix.

 Postscript
      Peter  Guldberg,  principal  author  of  the main  report  reviewed  this
 Appendix  and made an  important  point regarding this  section.   He says that
 really nothing  can  be said  about  causal relations  in  predictive equations
 because individual  regression  coefficients  are biased due to  interaction
 among the independent variables.   This can  be  seen by changing the order
 that variables  enter  a regression equation;  variable  coefficients  change as
 their order of  entry  changes.   Thus, comments  concerning  the  relative causal
 effect of a variable  vis-a-vis  others  cannot be made.
     I make  these types of comments in  Office-Professional when discussing
 the  Office  Zoning variable  and  in Other when describing the Income variable.
 Please disregard any  implications of causality in  these discussions. What
 I say about relative  sensitivity  (as defined here) of the variables and
 the  pragmatic reasons for their relative  sensitivity still hold, however,
 but  only within the abstract context of the regression equation itself.  See
 Volume I  for an extended discussion of "real-world" causal implications of
 GEMLUP equations.
                                E - 39

-------
IV.  Distribution of Predicted Land Uses Within the Area of Analysis
     The quantities of land uses predicted to occur within ten years of
increased sewage capacity apply to the area of analysis as a whole.  There
is no spatial resolution to the predictions, in other words.  Nor was any
ever intended.
     Spatial disaggregation was not attempted for three reasons:  (1) it
would have complicated the modeling exercise immensely, (2) it would have
ignored local knowledge of an area, and (3) it was unnecessary from
an air pollution emissions estimation viewpoint.
     Our desire to provide relatively simple, easy-to-use predictive equations
would have been obviated if disaggregated predictions were made.  Doing this
would require that some sort of internal grid system be used, that land uses
by grid cell be obtained, and that cell-specific predictive equations be
generated.  The user would then have to grid off his area of analysis and
use nine predictive equations for each cell.  The user would also have to
allocate existing land uses to each cell to ascertain what new development
was predicted to occur within a particular cell.
     We wanted not only to avoid doing cell-by-cell analysis but to use
local knowledge of existing and proposed development.  Local and regional
planners know where things are  and where  and what kind of  development is
about to go  in next.  They also know what the zoning is and where  streets
and highways will be constructed or improved.   It was felt  that  this local
knowledge would be more accurate in internally  allocating  development than
using a numerical technique based on case study  information.  Not  only
that, but if all development followed a deterministic pattern there would
                              E - 40

-------
not be a need for planning at all.  Phrased another way, the community-
represented by its planning agency—should be able to decide where develop-
ment should go given the fact that development is coming.  Zoning is
obviously based on that premise, and we wanted to work within that framework.
     Finally, for most air quality impact analyses, it is immaterial where
development is within a region as only total regional emissions are used in
the impact assessment.  This is particularly true for most land use types,
whose individual emissions are so low that they are lumped together as
"area source" emissions anyway.  In other words, air pollution modelers
do not need individual source emission estimates for their diffusion work
for most land uses in a community.  They only require specific estimates for
large "stationary sources" emitting approximately 100 tons or more of
pollution per year; these sources are large ("heavy") industrial firms,
power plants, refineries, and some extractive manufacturers.  And it should
be noted that specific estimates may not even be required for these large
sources for all pollutants.  Hydrocarbon emissions, for instance, are usually
only needed for an urbanized area on SMSA, because the analytical methods
used to convert hydrocarbon emissions into ambient ozone concentrations are
aspatial.  In fact, the air quality analyses are usually done on a multi-
county area larger than an SMSA  (called an AQCR, or Air Quality Control
Region).
     Distribution of predicted land uses within the area of analysis, then,
is done by the user using personal knowledge of the area, data in the
comprehensive plan, the zoning map, and information on development trends
                                  E - 41

-------
obtained from personal contacts, subdivision proposals and building permits.
It is intuitive to some extent, obviously, but should be suitable for most
projection purposes.  If disaggregated land use predictions are required
for some reason, the user will have to use another methodology; GEMLUP cannot
help.
                                 E - 42

-------
V.  Disaggregating Residential and Commercial  Predictions

     Worksheet 4 contains formula to disaggregate residential  and commercial

land use predictions.  (Worksheet 3 contains row entries for them, but data

for the rows come from Worksheet 4.)  As explained in Section  I.A.Z.b., the

disaggregation equations were developed using logit analysis,  where percentage

breakdowns of the classes of residential and commercial land uses were fitted

to a logistic S-shaped curve.  The equations were transformed  into regression

equations for the user.

     The disaggregation equations require that the following variables be

defined.  (If a variable is derived from input variables—those a user must

obtain data for—the input variables are shown in parentheses.)

Residential Disaggregation Equation

     Kid Density (school kids, dwelling units)
     Income Growth (future income, current income)
     Sewer Service (Sewered land, area of analysis)
     Office Zoning (zoned office, area of analysis)
     Nonmobility
     Hospital Growth (future medicals, current medicals)
     Housing Growth (future houses, current houses)
     Poverty
     Railroads (track, area of analysis)
     Peak Flow



Commercial Disaggregation Equation

     Office Vacancy
     Onsite Restrictions
     Treatment Capacity
     Kid Density (school kids, dwelling units)
     Restriction Years
     Phasing
     Transit Stops
     CBD Distance
     Government
     Vacant Land (vacant developable, area of analysis)
                               E - 43

-------
Of the 24 different input variables needed to use the disaggregation equations,
11 are used in core equations and only 4 are used in the residential equation.
Consequently, if a user opts for the simplified approach discussed in
Section I of this Appendix, he or she would not be able to disaggregate
residential and commercial predictions without getting more information.
This defeats the intent of simplification.
     How the user gets around this problem should depend upon what he or she
does with the predictions.  For instance, if the user is primarily interested
in using GEMLUP results to estimate area emissions, the commercial disaggre-
gation need not be done since the per floor area emission rate is the same
for all commercial categories.  A residential disaggregation should be done in
this case, however, because per unit area emission rates do vary for residential
categories for some, but not all, pollutants.
     Two logical ways to disaggregate commercial and residential predictions
are (1) use the existing local breakdown for the future estimates, or (2) use
the breakdown found in the GEMLUP case studies.  The existing local breakdown
for the residential categories can be obtained from U.S. Census information.
The existing local breakdown for commercial land uses may be available from
a local or regional planning agency, but probably will not be.  In that case,
the only option is to use GEMLUP case study information.
      The residential  and commercial  breakdowns found in the GEMLUP cases
 appear in Table E-6.   They could be  entered directly onto Worksheet 3
 (as decimal  equivalents)  and multiplied by the predicted quantities on
 the left to  come up with estimates of disaggregated land use.  Again, local
 knowledge should be used to determine if the GEMLUP breakdowns are reasonably
                              E - 44

-------
consistent with the existing development pattern of the area being inves-

tigated.  The case studies obviously had a lot of single family detached

housing and small (usually strip commercial and neighborhood shopping center)

commercial development.  If the user's area of analysis has a greatly different

development pattern, the disaggregated equations should be used instead of

case study breakdown.
                               Table E-6

                 RESIDENTIAL AND COMMERCIAL BREAKDOWN
                        FOR GEMLUP CASE STUDIES
Disaggregated
  Category

Residential
  Single Family
  Two-Family
  Multi-Family

Commercial
  Large
  Medium
  Small
Mean Value
 (Percent]

   100.0
    83.2
     4.8
    12.0

   100.0
    10.7
    10.5
    78.8
Range of Values
   (Percent)
  33.7 - 95.1
   0.0 - 27.0
   0.0 - 39.3
   0.0 - 54.5
   0.0 - 49.2
  29.6 -100.0
                                 E - 45

-------
                                   TECHNICAL REPORT DATA
                           (Please read Instructions on the reverse before completing)
1. REPORT
    3fpRAT-4%0/3-78-014b
                             2.
                                                          3. RECIPIENT'S ACCESSION-NO.
                                                          5. REPORT DA
                                                              'ORT DATE
                                                              May,  1978
4. TITLE AND SUBTITLE
    Growth Effects of Major  Land Use Projects (Wastewater
    Facilities); Volume  II:   Summary, Predictive Equati or|s PERFORMING ORGANIZATION CODE
    and Worksheets
7. AUTHOR(S)

    Peter H. Guldberg,  Ralph B.  D'Agostino
                                                          8. PERFORMING ORGANIZATION REPORT NO.
                                                               C-921
9. PERFORMING ORGANIZATION NAME AND ADDRESS

    Walden Division  of  Abcor, Inc.
    850 Main Street
    Wilmington, MA 01887
                                                           10. PROGRAM ELEMENT NO.
                                                           11. CONTRACT/GRANT NO.

                                                               68-02-2594
12. SPONSORING AGENCY NAME ANO ADDRESS
    Environmental  Protection Agency
    Office of Air  Quality Planning and Standards
    Strategies and Air Standards Division MD-12
    Research Triangle  Park, NC 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  in  a research program whose goal  is  to
    develop methodologies to predict the  total  air pollutant emissions resulting from
    the construction  and operation of major  land  use projects.  Emissions are quanti-
    fied from the major project, from land use  induced by the major project, from
    secondary activity occurring off-site  (e.g.,  electrical generating stations),  and
    from motor vehicle traffic associated with  both the major project and its induced
    land uses.

    This report  documents the development  of predictive equations for the induced  land
    use from wastewater major projects.   The predictive equations are included  in  an
    impact assessment procedure that estimates  the total air pollutant emissions as-
    sociated with  induced development from a wastewater major project.  This procedure
    is formalized  in  a set of easy-to-use worksheets, which serve as an operational
    tool for environmental engineers and  planners.
17.
                               KEY WORDS AND DOCUMENT ANALYSIS
a.
                  DESCRIPTORS
                                              b.lDENTIFIERS/OPEN ENDED TERMS  C.  COS AT I Field/Group
              Land  Use
              Planning
              Sewage  Treatment
                                Plants
Path Analysis
Causal Analysis
Secondary Effects
Induced Land Use
18. DISTRIBUTION STATEMENT

               Unlimited
                                              19. SECURITY CLASS (This Report)
                                                  Unclassified
                       21. NO. OF PAGES
                           207
                                              20. SECURITY CLASS (Thispage)
                                                  Unclassified
                                                                        22. PRICE
EPA Form 2220-1 (9-73)

-------