To my wife and parents. 
EMPLOYEE TRAVEL AT THE DALLAS/FORT WORTH 
REGIONAL AIRPORT 

by 
WALDO ANTONIO ZAMBRANO 
B. S. 
in Civil Engineering 
THESIS 
Presented to the Faculty of the Graduate School of 
The Texas at Austin
University of in Partial Fulfillment 
of the Requirements for the 
Degree of 
MASTER OF SCIENCE IN ENGINEERING 
THE UNIVERSITY OF TEXAS AT AUSTIN 
May 1976 
ACKNOWLEDGEMENTS 
I gratefully acknowledge the invaluable contributions of Dr. William of
J. Dunlay, Jr., as my supervisor in the preparation this thesis, and 
Dr. Clyde Lee as co-supervisor, who responded enthusiastically to all my requests for advice. 
The The Council for
Special recognition to University of Zulia, Advanced Ronald Tom
Transportation Studies, Linehan, Lyndon Henry, 
Caffery, The Delgado Family, Rita Tamayo and all those persons or entities 
who in one way or another aided in the completion of this report. Special thanks to my wife, my great support, and my parents. 
ABSTRACT 
This thesis presents an analysis of data that were obtained from 
the Employee Travel Survey made at the Dallas/Fort Worth Regional Airport (DFW) in May 1975 and a methodology for estimating DFW employee vehicular volumes arriving at or leaving the airport in a given time 
interval. From the survey information, an analysis is employee
made of 
characteristics for all the DFW employees. This is followed by a com
parison of employee characteristics according to whether or not they previously worked at Love Field Airport in Dallas. 
Theoretical distributions are developed for the period between the 
times that the work shifts start and end and the actual time
employees' that employees arrive at or leave the airport relative to those starting and ending times. Different theoretical distributions are obtained for different periods of the day. Finally, a model is developed for estimvehicular volumes from the work shift times or
ating employee (starting ending) and the number of employees on each work shift. Estimates from 
the model were found to compare favorably with actual counts of employee vehicles made during the survey. 
IV 
TABLE OF CONTENTS 
iv
ABSTRACT 
vi
LIST OF TABLES 
viii
LIST OF FIGURES 1CHAPTER I. INTRODUCTION 1
Objectives of Study 
1
and Limitations
Scopes 
3
CHAPTER 11. PREVIOUS RESEARCH 
5
CHAPTER 111. EMPLOYEE TRAVEL SURVEY 
The 5
Survey Form 
9
Survey Method Problems Encountered in Employee Survey 13 Sample Size 14 
17
CHAPTER IV. ANALYSIS OF DFW EMPLOYEE DATA 
I
Total DFW
Sampled Employees 7 
DFW to Previous Work Place
Employees According 30 
CHAPTER V. ARRIVAL AND DEPARTURE DISTRIBUTIONS OF DFW EMPLOYEES 47 
Determination of the Periods 47 Determination of the Theoretical Distributions 50 
of Day for Analysis 
CHAPTER VI. MODEL DEVELOPMENT 59 
Fundamental and Theoretical Concepts 59 Computation of Vehicular Volumes 64 Input Data 
66 Output 
68 Calibration of the Model 71 
CHAPTER VII. CONCLUSIONS AND RECOMMENDATIONS 76 
APPENDICES 
A Appendix B 89 Appendix 80 
REFERENCES 
100 
V 
LIST OF TABLES 
PAGE 
3.1. 	EMPLOYEE BY CATEGORY OF EMPLOYER 6 
3.2. 	EMPLOYEE SAMPLE BY TYPE OF EMPLOYER CATEGORY 15 
4.1A. 	DISTRIBUTION OF DFW EMPLOYEE RESIDENTIAL LOCATION BY CITIES INSIDE THE INTENSIVE STUDY AREA 21 
4.18. 	DISTRIBUTION OF EMPLOYEE RESIDENTIAL LOCATION BY CITIES OUTSIDE THE INTENSIVE STUDY AREA 22 
4.2. 	ZONE NUMBERS FOR THOSE EMPLOYEES WHO DID NOT GIVE A COMPLETE ADDRESS BUT DID GIVE THE CITY 25 
4.3. 	MODE COMBINATION USED BY DFW EMPLOYEES IN THEIR WORK TRIP. 27 
. 
4.4. 	PERCENT OF DFW EMPLOYEES BY TYPE OF WORK 28 
4.5. 	DFW EMPLOYEES BY AGE 31 
4.6. 	DFW EMPLOYEES BY SEX 32 
4.7A. DISTRIBUTION OF DFW EMPLOYEE RESIDENTIAL LOCATION BY CITIES INSIDE THE INTENSIVE STUDY AREA ACCORDING TO PREVIOUS WORK PLACE 35 
4.78. 	DISTRIBUTION OF DFW EMPLOYEE RESIDENTIAL LOCATION BY CITIES OUTSIDE THE INTENSIVE STUDY AREA ACCORDING TO PREVIOUS WORK PLACE 39 
4.8. 	MODE COMBINATIONS USED BY DFW EMPLOYEES IN THEIR WORK TRIPS ACCORDING TO PREVIOUS WORK PLACE 41 
4.9. 	DISTRIBUTION OF DFW EMPLOYEES BY OCCUPATION ACCORDING 42 TO PREVIOUS WORK PLACE 
4.10. 	DFW EMPLOYEES BY AGE ACCORDING TO PREVIOUS WORK PLACE. 45
.. 
4.11. 	DFW EMPLOYEES BY SEX ACCORDING TO PREVIOUS WORK PLACE. 45
. . 
5.1. 	LIMITS OF PERIODS OF STARTING AND ENDING WORK SHIFTS 51 
... 
5.2. CHARACTERISTICS OF PERIODS OF DFW EMPLOYEES STARTING THEIR WORK SHIFTS 
VI 
TABLE 
PERIODS OF DFW EMPLOYEES
5.3. CHARACTERISTICS OF 
57
ENDING THEIR WORK SHIFTS 
6.1 COMPARISON OF DFW EMPLOYEE VOLUME FROM DIFFERENT 
72
SOURCES (ARRIVING AT THE AIRPORT) 
6.2 COMPARISON OF DFW EMPLOYEE VOLUME FROM DIFFERENT 
73
SOURCES (LEAVING THE AIRPORT) 
VII 
LIST OF FIGURES 
FIGURE  PAGE  
3.1  EMPLOYEE  SURVEY  FORM  7  
3.1  (CONTINUED)  8  
3.2  INTRODUCTORY  LETTER  TO  EMPLOYERS  10  
3.3  INSTRUCTIONS  TO  EMPLOYERS  11  
3.4  INSTRUCTIONS  TO  SUPERVISORS  12  
4.1  REGIONAL  ANALYSIS  AREA  DESIGNATIONS  18  
4.2  CITIES  INSIDE  THE  INTENSIVE  STUDY  AREA  19  
4.3  ZONAL  DISTRIBUTION  OF  DFW  EMPLOYEES  20  
4.4  DISTRIBUTION OF DFW THE INTENSIVE STUDY  EMPLOYEES AREA  BY  CITIES  INSIDE  23  
4.5  DFW  EMPLOYEES  WORK  TRAVEL  MODE  26  
4.6  DFW  EMPLOYEES  BY  FAMILY  INCOME  29  
4.7  ZONAL DISTRIBUTION WORK AT LOVE FIELD  OF DFW EMPLOYEES AIRPORT  WHO  USED  TO  33  
4.8  ZONAL DISTRIBUTION WORK AT LOVE FIELD  OF DFW EMPLOYEES AIRPORT  WHO DID  NOT  34  
4.9  DISTRIBUTION OF DFW EMPLOYEES WHO FIELD AIRPORT BY CITIES INSIDE THE  USED TO WORK AT INTENSIVE STUDY  LOVE AREA.  .  .  36  
4.10  DISTRIBUTION OF DFW EMPLOYEES FIELD AIRPORT BY CITIES INSIDE  WHO THE  DID NOT WORK AT LOVE INTENSIVE STUDY AREA.  .  .  37  
4.11  DFW EMPLOYEES PREVIOUS WORK  WORK TRAVEL PLACE  MODE  ACCORDING  TO  40  
4.12  DFW EMPLOYEES WORK PLACE  BY  INCOME  LEVEL  ACCORDING  TO  PREVIOUS  44  
5.1  DISTRIBUTION OF DFW EMPLOYEES  WORK.  SHIFT  STARTING  TIMES  OF  48  
VIII  

FIGURE 	PAGE 
5.2 	DISTRIBUTION OF WORK SHIFT ENDING TIMES OF 
DFW EMPLOYEES 

49 
5.3 	TYPICAL DISTRIBUTION OF TIME DIFFERENCE DATA 53 
5.4 	PROCEDURE FOR FITTING TIME DIFFERENCE DISTRIBUTION 54 
5.5 	SAMPLE DISTRIBUTION FOR DFW EMPLOYEES ARRIVING AT THE DFW AIRPORT 58 
5.6 	SAMPLE DISTRIBUTION FOR DFW EMPLOYEES LEAVING THE DFW AIRPORT 58 
6.1 	EMPLOYEE VEHICLES ARRIVAL PATTERNS (SINGLE WORK SHIFT TIME) 60 
6.2 	EMPLOYEE VEHICLES DEPARTURE PATTERNS (SINGLE WORK SHIFT TIME) 60 
6.3 	EMPLOYEE VEHICLES ARRIVAL PATTERNS (OVERLAPPING OF CURVES) 63 
6.4 	EMPLOYEE VEHICLES DEPARTURE PATTERNS (OVERLAPPING OF CURVES) 63 
6.5 	FLOW CHART OF EMPLOYEE VEHICLE VOLUME COMPUTATION 65 
6.6 	INPUT DATA STRUCTURE 67 
6.7 	WORK SHIFT INFORMATION 69 
6.8 	PERIODS INFORMATION 69 
6.9 	CALCULATED DFW EMPLOYEE VOLUMES 70 
6.10 	CALCULATED AND COUNTED DFW EMPLOYEE VOLUMES (ARRIVING AT THE AIRPORT) 75 
6.11 	CALCULATED AND COUNTED DFW EMPLOYEE VOLUMES 
(LEAVING THE AIRPORT) 

75 
ix 
CHAPTER I. INTRODUCTION 
The work presented in this thesis is part of a project undertaken under the auspices of the Council for Advanced Transportation Studies at 
University of at 	project sponsored by the
The 	Texas Austin. The is 
Office 	of the of
of University Research U. S. Department Transportation. The study staff consisted of two principal investigators, Dr. William Dunlay, Jr. of Civil Engineering of Geography,
J. 	and Dr. Pat Burnett 
plus various research 	students (of which
associates, graduate research the author was and
one), undergraduate students. 
Objectives of Study 
The 	the research described in this thesis is to
objective of analyze part of the data obtained from the Dallas-Fort Worth 
Airport (DFW) employee survey in May (Ref. 4). analysis will consist
conducted 1975 	The 
of a comparison of DFW employees according to whether or not they previously worked at Love Field Airport in Dallas and an analysis of the 
times that employees arrived at or left the airport relative to work 
shift starting or ending times. Based on this latter analysis, a method
ology is developed for estimating 	employee vehicles
the number of DFW 
that be to enter and leave the 	time
can expected airport in any given interval. 
Scope and Limitations 
In the comparison of DFW employees according to whether or not they used to work at Love Field, a general analysis is made of all those 
characteristics for which both subsets of employees presented enough information in the above mentioned survey. In particular, the same anal
ysis is made of both subsets and a comparison is made of the DFW employees who previously worked at Love Field and the current DFW employees who did 
not. 
In 	the analysis of the time that employees arrive at or leave the relative to the of the
airport 	starting and ending their work shift, 
1 
distributions of the interval between those two times are tested 
against known theoretical distributions. The number of candidate theoretical 
and the
distributions is five, goodness-of-fit test used is the 
Test and For the methodology developed to estimate the number of employees Kolmogorov-Smirnov (Refs. 1,2, 5 13). 
entering and leaving the airport in a given time interval, the required input includes the periods of the day to be considered, parameters of the above 
theoretical distributions for each period, the starting or ending 
time  of  each  work  shift, and  the  respective  number  of employees.  The  
output  is the  number  of  employee  vehicles  entering  or  leaving  the  air 
port  for  any  given  time  period.  

CHAPTER 11. PREVIOUS RESEARCH 
The results of past airport employee travel surveys have not been widely published. References which discuss this matter are quite 
limited. In most of the references concerned with the 
study of airport access traffic, the influence of the airport employees is mentioned, but how this influence is distributed in time is almost described
never 
modeled.
or 
John Robinson and Peter Nordie 
(Ref. 12) present an origin/destination survey of Washington National Airport. In the employee survey part of that study the following procedures were followed: 
contacted
1. 	Organizations having more that 75 employees were 
From 	of
personally. an alphabetical listing of all employees 
these 	a
organizations sample consisting of every eighth person and
was selected, a weekly record of travel patterns was obtained for that sample. 
2. 	For organizations with less than 75 employees, questionnaires were sent to the organizations by mail with the instructions to 
be 	followed 
in selecting employees for the sample. 
3. The information obtained from the employee survey included: 
a) 	Work hours 
b) 	Home address 
c) 	Type of vehicle 
d) 	Automobile occupancy 
e) 	Trip time (from and to the airport) 
Attitude toward 	about the
f) 	trip time, i.e., feeling trip 
time. 
Only summary comments about the results of the questions asked in the employee survey were presented in the reference 12. (Ref. 3) presents study
Merrit Chance 	a as to how the different 
users (of whom employees are one category) of airport access highways 
create ground transportation problems. In the section called "Daily 
Distribution of 	and Travel
Air Passengers, Employees Related Visitors," several graphs that give some indication of the daily movement of people 
3 
at six airports, which are San Francisco, Washington National, Dulles, Friendship, Los Angeles, and London Heathrow,are presented. It is quite 
obvious that in
the percentage of volume represented by daily employees this study was deterministically from starting and ending
derived the work shift times. No references about the actual time distribution of airport employees arriving at leaving the airport were
or found. 
CHAPTER 111. EMPLOYEE TRAVEL SURVEY 
In order to the travel habits of at the DFW
investigate employees airport, a travel survey was conducted by Dunlay, et al. (Ref. 4). In 
forms for
this survey, the written, self-executed, type of questionnaire were distributed to more than 13,000 employees. The rate of response of 
completed questionnaires was 24 percent. Presently there are 13,368 employees at the Dallas/Fort Worth Airport. Table 3.1 shows the division of these employees by employer category. 
The Survey Form 
The form consisted of short
employee survey a introductory para
graph followed by eleven questions (Fig. 3.1). The first question was to determine each employee's present home address, allowing the option of answering either by street address or by the nearest intersection. 
This  option  was  designed  to  allow  the  response  to  be  less personal  and  
to  encourage  a  response  from  persons  who  might  be  reluctant  to  give their  
address.  

The next two questions dealt with how an employee perceives the distance and travel time between his home and DFW, and also between his home and Dallas Love Field. The answers to these questions can be compared with the true distances and travel times (Ref. 11). Question number four asked the employee to indicate the type of vehicle he uses in his work trip. This was straightforward, and all reasonable alternatives were included in the check list. The next was whether the
topic of interest, question five, employee worked at Love Field before DFW opened. If a change of residence occurred because of the shift to the DFW airport, this was noted in Part Bof old address also
the question. The respondent’s was obtained in Part Cof the same question. It was hypothesized that the decision to move should be related to the actual travel time and distance between DFW and the respondent’s old home. The employee was also asked to indicate the mode of transportation he used in his work trips to Love Field. 
5 
TABLE 3.1. EMPLOYEES BY CATEGORY OF EMPLOYER 
EMPLOYER 	CATEGORY NO. OF EMPLOYEES 
Airlines 9,126 
Air Cargo (1) 

382 
Air Mail Facility 350 
Airport Marina Hotel 250 

Allied Aviation Company 95 
APCOA (Airport Parking Control) 135 
Board

DFW Airport 480 
Federal Aviation Administration 135 

Food Service 1,406 
General Telephone Company 43 

L.T.V. Airtrans 	135 
Maintenance (2) 284 
Rent-A-Car Firms 268 
Security (2) 

120 
Other 	159 
TOTAL 	13,368 
Source: 	Central Offices at DFW of the Employers and the DFW Airport Board. 
(1) 
Excluding Airlines with Airfreight 

(2) 
Excluding DFW Airport Board (Airport Police) 


7 
Form 
Survey 

Employee 

3.1. 
Figure 

30 >4 38*2 46
22 	U49
i 	53 57
*1 45
29 3337 
17 18 48□50 51 54 

46
J
□□2021[ 	36 *0 44 □m52□56
28 32 
U 
25 cur cm (TTT 	LLLi 58 
47
ra123 m
1
19 31 3943 	53
?7 35 
to 
shift 
your
Regional 
of 
(Zip) (Zip)
Worth 
because A.M. 	one) 
chant* 	check


Dallas-Fort specify) 	*32,000
to 
(Please *32,000
plan lease A.M. P.M. A.M. P.M. A.M. F.M. 	Labor 
9 




food/alrllna/custodlan (Continued)
the 	bus ( 
mv you Town) Town) 	*26,000 Ov.r 
of do oror 
or address? Taxi Public Other 	Saturday Other Service 
opening (City (City airport.
street tha

tha 
raaldaoca TODAY? 	*20,000 *26,000 3.1. 
No 
TODAY? 
of 	TODAY? at 
before previous alroort TODAY? work Thursday Friday Foreman one) 
placa Yaa Name) Field: etart airport 55-64 65 one) qcck *13,000 *20,000

Flail your 	WORK: the work yeu Ov.r 

Craftsman, Technician/Operator Maintenance 	Figure
Leva your was (Strsat 	Love someone shift off the week check 
to LEAK at the r«.i. (please

EfflCWtNT: Yaa changed No what 	Intersection) vehicle by work one) -

arrive get leave 
off WO 	of
Dallas 	taken carpool 35-4* (please IS: -

own you your you you TRAVEL: Tuesday Wednesday check *5-5* 
a
AIRPORT at you moved, 	Street ARRIVE days *13,000
did did
work hav# my In

airport? No.) vehicle droDDed will will YUUR the h»i« (puase 	INCCft *6,500
have 
of YOU time time time time CF 21

you new
PREVIOUS Dll Airport? HUES*the you (Strsst (Nearest Type Driving tiding Beirut THAT What What What What check Sunday Mooday .ARE: AGE: Under 21-3* OCCUPATION: Professional Clerical Sslaa FAMILY Under *6,500 VQJR Tift FREQUENCY Please YOUR 
If 	YOUR YOUR
YOU
A.B. Ot 	C. A.B.C. D. 
11.
S. 	* »■». 10. 
two number intended
The next questions, six and seven, were to provide a of the in of the
sample employee’s '’normal” work day terms time when he began the shift, the number of hours worked, and the number of days per week that he worked. The sample size was assumed to be large 
effect
enough to offset the of asking for starting and quitting times 
on one specific day. 
The survey form ended with some requests for personal data. These 
can be used to relate the to
travel behavior of employees standard, identifiable demographic characteristics, similar to those gathered in the U. S. Census. The age group brackets were designed to encompass 
standard phases of the personal and family life cycle. The occupational breakdown the guidelines of census data, but was simplified
followed 
somewhat for the convenience of the The income
respondent. family survey
brackets shown on the form are the same as those used in the 
previous survey at Love Field by Alan M. Vorhees, Inc. in 1969, but were adjusted upward for inflation using the increase in the region’s Consumer 
Price Index. 
Survey Method 
The distribution and collection of the forms
employee survey proved 
to be a time-consuming task as seventy-one airport-related employers 
The vast
had to be contacted. forms were distributed
majority of survey through the mail. A letter of introduction was included which explained 
the purpose of the survey (Fig. 3.2). In addition, a set of detailed 
instructions was compiled for the employers, which suggested a particular distribution and collection procedure (Fig. 3.3). A few 
announcements 
of the study suitable were also included in 
the 
information 

for posting packet of 
sent to employers. The survey forms were distributed 
through employee supervisors. Distribution and collection instruc
the 
tions for the supervisors were printed on the envelope which contained 
the forms (Fig. 3.4). 
Figure 3.2. Introductory Letter to Employers. 
Figure 3.3. Instructions to Employers. Figure 3.4. Instructions to Supervisors. 
Problems Encountered in Employee Surve 
From the results examined so far, it appears that the overall design 
of the survey form was good, and that usable data were obtained. There 
was little or no confusion on most of the asked. the
questions However, wording of some questions could be improved, and suggestions for this are given 
below. Problems with the form itself involved wording, length, and the fact that the Airport Board had recently conducted a ot its own.
survey Thus, the fact that some employees might have been irritated by the necessity to execute another survey form may have lowered the response 
rate. 
The initial confusion in the employee survey was due to delays in delivering the packages of forms to the employers by the postal service. In the of forms delivered until the
some cases, packages were not day 
they were supposed to be filled out. This allowed no time for employers 
to their distribution effort. In
adequately organize addition, delays 
in channels resulted

some intra-airport communication in management personnel the forms late the
receiving as as following Monday or Tuesday, May Since the instructions requested that employers
19 and 20. and 
supervisors "...distribute the enclosed survey forms to the employees under your supervision on Friday, May 16..." some employers who received their forms after May 16 assumed that it was too late to distribute them 
The project staff subsequently had to contact these employers and encourage them to distribute the forms to their employees. In the code of the address
question one zip was requested as part 
This proved to be a valuable piece of information, as some respondents 
left out their city name but included their zip code. 
A small number 
of respondents misinterpreted question two as asking 
for a round trip distance. This could have been avoided by specifying 
distance.
a one-way Also, a few respondents may have misinterpreted 
question 2B as asking for the distance between DFW and Love Field since 
they were filling out the form at DFW. It was possible to spot-check these errors by the locations of the two airports relative to their homes. 
Question four could be improved by asking for the vehicle taken "most often" or "usually," as several multiple responses were encoun tered. 
Another troublesome question was the one that requested employees to classify themselves by occupation (professional, clerical, sales, craftsman/foreman, technician/operation, maintenance, labor, ser
other 
It deemed to check list for
vice). was preferable give the respondent a 
this purpose, to avoid nebulous and illegible answers which would be difficult to that
interpret by the project staff. However, it turns out the of such a list also be conducive to
wording may misinterpretation by the respondent. In addition, a question of this type actually solicits the respondent's perceived self-classification. That this can produce 
problems recognized by incongruities responses
has been the found between 
to to related
this question vis-a-vis responses questions, e.g., a 
-
"clerical" worker with income "$26,000 $32,000." The occupational breakdown used was selected in consultation with the North Central Texas 
Council of Governments which was doing a study of basic and non-basic industry in the Dallas/Fort Worth Region. With regard to questions six and seven, some employees with rotating shifts only unspecifically days and vary.
stated that work shift time 
Others specified that days and work shift time vary and included the work 
of the Still others checked all
days and shift times at the time survey. 
the days of the week. 
created
two new
Concerning question number 10, job categories were 
n
because of employee responses: MRent-a-carand "Hotel employees." 
Also, the percent of employees checking "Other Labor" was relatively 
high. 
Sample Size 
A total of forms late forms have been
3,157 employee plus 84 returned, which constitutes a 24 of return of the 13,368 forms distri
percent rate buted. Table 3.2 shows the number and percent of survey forms returned by type of employer. General Telephone Company, Airport Parking Control, 
and DFW Airport Board were the employer categories with the highest 
TABLE 3.2. EMPLOYEE SAMPLE BY TYPE OF EMPLOYER CATEGORY 
PERCENT 
EMPLOYER 	CATEGORY SAMPLE 
RETURNED 
Airlines 	1,211 
13.3 Air Cargo (1) 70 18.3 Air Mail Facility 200 57.1 Airport Marina Hotel 171 48.4 
Allied Aviation Company 0 0.0 
APCOA (Airport Parking Control) 87 64.4 
DFW Airport Board 294 61.2 
Federal Aviation Administration 57 42.2 
Food Service 283 20.1 
General Telephone Company 31 72.1 

L.T.V. Airtrans 29 21.5 Maintenance 94 33.1 
Rent-a-Car Firms 126 47.0 Security (2) 19 15.83 Other 26 16.3 Unidentified 594 4.4 
TOTAL 	3,241 24.2 
Source: 	Central Offices at DFW of the Employers and the DFW Airport Board. 
(1) 
Excluding Airlines with Airfreight. 

(2) 
Excluding DFW Airport (Airport Police) 


Board
rates of return (72.1%, 64.4%, and 61.2% respectively). On the other 
hand,  Allied  Aviation Company,  the  airlines  ,  and  security  were  the  
employer categories  with the  lowest  percent  rate  of  return  (0%,  13.3%  
and  15.8% respectively).  

CHAPTER IV. ANALYSIS OF DFW EMPLOYEE DATA 
is divided into two sections. The first section
This chapter major 
deals with the total sampled DFW employees and their residential location broken down by zone, city, type of vehicle used for work trip, occupation, 
income, age and sex. In the second section of the chapter, the same analysis and a comparison are made of the two subsets of employees at DFW, namely, those that used to work at Love Field airport before the 
and
opening of DFW, those that did not. 
The study area considered in this analysis is divided into zones designated by the North Central Texas Council of Governments (NCTCOG), 
to which
giving emphasis the so-called Intensive Study Area (ISA), covers 
primarily Tarrant and Dallas counties (Fig. 4.1). Later in this analysis, the ISA zones are grouped at the city level (Fig. 4.2). 
Total Sampled by DFW Employees 
DFW Employees by Zone. Figure 4.3 shows the distribution of residential location of DFW employees disaggregated by zones inside of 
the Table shows the the
study area. A.l in Appendix A frequency and percent of the in each of the
sample zone study area. There are two pertinent observations to be made. The first is the 
wide dispersion of small percentages of employees on the one hand, and the single concentration of 14.4 percent in five zones near the airport (zones 350, 368, 374, 375 380)
and on the other. Note that the maximum 
percent in any zone is only 5. Therefore, the sample distribution by 
zone is considered to be too small and an aggregation of the zones into bigger areas is desirable. 
DFW Employee Distribution by City. One way to aggregate the zones is at the city level. This produces a clear idea of how the employees are distribtued in the study area. Table 4.1 A and Table 4.18 show the frequency and percent of the sample of employees for the cities inside 
and outside the intensive study area, respectively. Figure 4.4 shows 
graphically that the city of Dallas has the largest single portion of 
17 
18 


Designations. 

Area 
Analysis 
Regional 

4.1. 
Figure 

19 
Area. 
Study 



Intensive 

the 
Inside 
Cities 

4.2. 

Figure 

20 


Employees 

DFW 
of 

Distribution 

Zonal 
4.3. 

Figure 

TABLE 4.1A. DISTRIBUTION OF DFW 
EMPLOYEE’ RESIDENTIAL LOCATION 
BY CITIES INSIDE 
THE INTENSIVE STUDY AREA 
SAMPLE 
CITY PERCENT 
FREQUENCY 
Addison 7 
0.22 
Arlington 233 7.19 Azle 3 
0.09 Balch 4
Springs 0.12 Bedford 93 
2.87 Benbrook 3 
0.09 Blue Mound 4 
0.12 
Carrolton 74 2.28 Cedar Hill 5 
0.15 Colleyville 28 0.86 Coppell 8 0.25 Dallas 801 24.71 
Dalworthington Gardens 1 0.03 De Soto 13 0.40 Cuncanville 13 0.40 Evenaan 3 0.09 Euless 215 6. 63 Farmers Branch 56 1.73 Fort Worth 257 7.93 Forest Hill 3 0.09 Garland 49 1.51 Grand Praire 88 2.72 Grapevine 93 2.87 Haltom City 20 0.62 Highland Park 2 0.06 
Hurst 138 4.26 
Hutchins 1 0.03 
Irving 404 12.47 
Keller 30 0.93 
Kennedale 1 0.03 
Lancaster 4 0.12 
Mansfiled 3 0.09 

Mesquite 25 0.77 
North Richland Hills 59 1.82 
Pantego 1 0.03 
Richardson 33 1.02 
0.59 
River Oaks 2 0.06 
Sachse 1 0.03 

Richland Hills 19 
Saginaw 2 0.06 Seagoville 1 
0.03 
0.46
Snithville 15 
0.28
9
South Lake 0.56
University Park 18 
0. 37
Watauga 12 
0.09
White Settlement 3 
0.03
1
Wilmer 
88.18
TOTAL 2,858 
TABLE 4.18. DISTRIBUTION OF EMPLOYEE RESIDENTIAL LOCATION BY CITIES OUTSIDE THE INTENSIVE STUDY AREA 
SAMPLE  7. OF  
CITY  
FREQUENCY  TOTAL  
Allen  2  0.06  
Alvord  1  0.03  
Argyle  7  0.22  
Aubrey  l  0.C3  
Blue Ridge  2  0.06  
Bonham  1  0.03  
Bowie  1  0.03  
Boyd  6  0.19  
Bridgeport  4  0.12  
Celina  2  0.o6  
Celeste  1  0.03  
Cleburne  1  0.03  
Clifton  1  0.03  
Collinsville  1  0.03  
Conroe  1  0.03  
Decature  1,  0.03  
Denton  31  0.96  
Elmo  2  0.06  
Ennis  2  0.06  
Fairfield  1  0.03  
Fannesville  1  0.03  
Ferris  2  0.06  
Flower Mound  2  0.06  
Frisco  8  0.25  
Gainsville  1  0.03  
Granbury  1  0.03  
Gordon  1  0.03  
Greenville  1  0.03  
Josbua  2  0.06  
Justin  4  0.12  
Highland Village  1  0.03  
Kerns  1  0.03  
Lake Dallas  6  0.19  
Little Elm  1  0.03  
Lewisville  113  3.49  
Mabank  '■ 3  0.09  
McKinney  9  0.28  
Midlothian  1  0.03  
Nevada  1  0.03  
Nocona  1  0.03  
Paradise  1  0.03  
Plano  26  0.80  
Pilot Point  2  0.06  
Ponder  2  0.06  
Poolville  1  0.03  
Quinlan  1  0.03  
RedOak  2  0.06  
Rhome  1  0.03  
Roanoke  30  0.92  
Rockwall  3  ‘0.09  
San Marcos  1  0.03  
Sanger  3  0.09  
Sunset  1  0.03  
Tioga  1  0.03  
Valley View  2  0.06  
Weatherford  3  0.09  
Waxahachie  5  0.16  
Wills Point  X  0.03  
Whitewright  1  0.03  
Wylie  4  0.12  
TOTAL  320  9.87  

23 
Area. 
Study 



Intensive 

the 
Inside 
Cities 

by 


Employees 

DFW 
of 

Distribution 

4.4. 
Figure 

DFW employees with 24.7 percent. In comparision, Irving has 12.5 percent 
of the total employees sampled—half that of Dallas despite its smaller 
population. Of the cities outside the ISA, Lewisville, Denton, 
Roanoke 
and Plano are the ones with the highest percent of employee residences. Two percent of the employees sampled either did not give their address or it was not possible to locate the given address on the reference maps 
(Ref. 8).
7 and 
Although some addresses were not identifiable by zone, it was 
possible to indicate the location of those DFW employees for which only 
the 
city name was given by assigning them a dummy zone number (see Table 4.2). explains the presence dummy zone
This of such numbers in Table 
A.l and later in Tables A.
A. 2 and A. 3 of Appendix 
DFW the
Employees by Type of Vehicle. "Driving my own vehicle," was that employees gave 84.6 percent they were
answer of the time when asked 
about the type of vehicle taken to and from work. Figure 4.5 shows the 
Data
distribution of DFW employees by type of vehicle. on employees who 
indicated they use more than one type of vehicle in their work trip are presented in Table 4.3, which shows only the five most frequently given 
combinations. 
DFW Employees by Occupation. Table 4.4 presents the percent of employees occupational category. account
in each Professionals for 30.5 
followed service airlines 
sales (12 percent) and other labor (11.5 percent). The last category, 
other labor, probably reflects the presence of a large variety of work 
types which could not be covered by the 10 categories named on the sur
percent of the responses, by (15.1 percent), 
form.
vey DFW Level of Income. The 4.6 shows
Employees by histogram in Fig. 
the level of

breakdown of employees by income. Although this question 
is 8 of the returned forms did

quite personal, only percent survey not mention the family income. 
DFW The of 21
Employees by Age and Sex. range to 44 years accounts for the 
highest percent of employees (75 percent) with 49.3 percent 
TABLE 4.2. ZONE NUMBERS FOR THOSE EMPLOYEES WHO DID 
NOT GIVE A COMPLETE ADDRESS BUT DID GIVE THE CITY 
CITY ZONE 
Addison 729 Arlington 704 Bedford 714 Carrolton 708 
Dallas 700 Denton 
715 De Soto 737 Duncanville 713 Ennis 735 Euless 710 Farmers Branch 709 Fort Worth 701 Garland 706 Grand Prairie 
703 Grapevine 712 Hurst 711 
Irving 702 Keller 740 Lewisville 730 McKinney 734 Mesquite 707 Mesquite 732 North Dallas 728 Plano 733 Richardson 705 
Sherman 
736 West Side Fort Worth 731 
26 
Mode. 
Travel 

Work 
Employees 

DFW 
4.5. 
Figure 

TOTAL 
OF 


EMPLOYEES 
1.97 1.14 2.59 0.49 0.31 4.50
DFW 
DFW 
BY PERCENT 
USED TRIP. 
643719 1610
COMBINATION WORK SAMPLE FREQUENCY
THEIR 146 
IN 
MODE 
EMPLOYEES Carpool by carpool
4.3. someone
a
a off
in in 
by
TABLE 
off
riding Surtran dropped other riding 
ororor oror 

dropped
car car car car car


COMBINATION someone ‘
or 
Own Own Own Own Own TOTAL 
TYPE OF WORK  PERCENT OF TOTAL DFW EMPLOYEES  30.5  12.0  5.6  2.2  3.5  7.2  11.5  5.8  15.1  1.1  0.9  1.0  3.5  100.00  
BY  
EMPLOYEES  
PERCENT OF DFW  SAMPLE FREQUENCY  990  389  180  72  113  233  374  188  491  37  29  32  113  3,241  
TABLE 4.4.  OCCUPATION  Professional  Clerical  Sales  Craftsman, Foreman  Technician/Operator  Maintenance  Other Labor  Service (Food)  Service (Airline)  Service (Custodian)  Rent-A-Car  Hotel Employee  No Response  TOTAL  

29 
Income. 
Family 

by 
Employees 

DFW 
4.6. 
Fiqure 

between 21 and 34 years of age (see Table 4.5). The distribution of 
employees by sex turned out to be 63.3 percent males and 34.1 females. 
The remaining 2.6 percent did not answer the question (see Table 4.6). 
DFW Employees According to Previous Work Place 
Because Love Field in Dallas was the major air carrier airport 
before the opening of DFW, a significant percentage of former Love Field employees airport when shift of operations
transferred to DFW the 
occurred 1974. In the survey, 57 the
in January, percent of employees sampled formerly worked at Love Field. This section will compare DFW employees who previously worked at Love Field with those who did not. The tables will be broken down by both subsets: "Former Love Field Worker" and "Non-Love Field Worker." Each subset is divided into three columns: the the "Percent of Total DFW Employees,"
"Sampled Frequency;" 
which 
tabulates the sample frequency of each category label as a percentage of the total DFW employee sample of 3,241; and "Percent of DFW Employees Within Category," sample frequency of
which tabulates the each 
category as a percentage of the total DFW employee sample in the same category. 
Distribution by Zone, Figures 4.7 and 4.8 and Table A. 2 of Appendix 
A, show the DFW employees by zone of residence broken down by whether 
they used to work at Love Field or not. It can be noted that those 
employees that used to work at Love Field are predominantly spread over 
the zones which correspond to the mid-cities area of the ISA and to 
Dallas County. On the other hand the DFW employees that did not work at Love Field are distributed most frequently over Tarrant County. 
Distribution by City. Table 4.7 A and Figs. 4.9 and 4.10 more clearly indicate the observations stated above. Cities like Carrolton, Dallas, and Irving, located in Dallas County and having relatively sig
nificant percentages of DFW employees, exhibit higher percentages of 
former Love Field than current DFW employees who did not work

employees at Love Field. On the other hand, cities like Arlington, Bedford, Hurst, and Fort Worth, located in Tarrant County and having a relatively high 
OF TOTAL EMPLOYEES  
PERCENT  DFW  7.7  49.3  23.9  13.3  4.3  .2  1.3  100.0  
AGE  
BY  
EMPLOYEES  
DFW  SAMPLE FREQUENCY  251  1,599  773  432  138  7  41  3,241  
4.5.  
TABLE  
CATEGORY  Under 21  21-34  35-44  45-54  55-64  Over 65  No Response  TOTAL  

SEX 
BY 

EMPLOYEES 

DFW 
4.6. 
TABLE 

63.3 34.1
PERCENT 2.6 SAMPLE FREQUENCY 2,052 1,105 84 
Response
CATEGORY Female No
Male 
\ < •:-i 
Airport.
3» 
-
r 
* 
IBm 

: 
:jx 
; 
v:; 
-
/ 
/ ?-.yv
bv.r-/"
1 


7^-■K

xX'Vyi
b'XX • 
/ 
-
!: 
A
frWjxX • 
1 gtpxrftx<T
r 
tl\7' b"i— 
j 
"'’
’x
f* ■**
bn-■ 
' 
1 
X~rrix [EPfV’T
xxcl.■p-V^X” ip
a’A*\ /■ \ ')-11 a :
v<5 '^ 
Vl 
!•;./*. r 
' 
«H
«^i2j— XhA;.-J...K' 
-
j2 ]-J_* 
/ 
—
A\_lv"X-pu « xr.;,,,jhdG3;i;>tL-( >r— Jry 
l„ 
/ 
-
~
>X
* 
F* 3 
(~j-~~-f-j
;, *VV^y jxxxj~r/y~~C|frW 
-
r 
• 
U-:
• 
X
• 
j• 
•tyXisSI kjyjyQyc rVS'K'*'^{Jm«"7? 
' 
-' 
' 
i 
\ 
Field Love 
at 
Work 
to 
Used Who 
Employees 
DFW 
of 
Distribution 
Zonal
5.03
i 
234 
< «£• *£ 
4.7.
<%<%<% C% <% 
o 23 4
EM3 mmi E22 EM Esa Figure 
I 
34 
Airport. 

Field 
Love 

at 
Work Not Did 
Who 
Employees 

DFW 
of 
Distribution 

Zonal 
4.8. 
Figure 

TABLE 4.7A. DISTRIBUTION OF DFW EMPLOYEE RESIDENTIAL LOCATION BY CITIES INSIDE THE INTENSIVE STUDY AREA ACCORDING TO PREVIOUS WORK PLACE 
FIELD WORKER NON-LOVE PERCENT OF PERCENT OF DFW FORMER LOVE FIELD WORKER 
PERCENT OF PERCENT OF DFW 
cm SAMPLE SAMPLE TOTAL DFW EMPLOYEES WITH-TOTAL DFW EMPLOYEES FREQUENCY FREQUENCY
EMPLOYEES IN CATEGORY EMPLOYEES WITHIN CATEGORY 
Addison 6 0.19 35.17 1 0.03 14.29 Arlington 71 2.19 30.47 158 4.36 67.31 Azle 3 0.09 100.00 Balch Springs 4 0.12 100.00 Bedford 43 1.33 1.48
46.24 48 51.61 
33.33 2
Benbrook 1 0.03 0.06 66.67 Blue Mound 3 0.C9 0.03
75.00 1 25.00 
62 1.91 83.73 10 0.31 13.51
Carroltoa 
Hill 2 0.06 40.00 2 0.06 40.00 Collewille 14 0.43 50.00 14 0.43 50.00 Coppell 4 0.12 50.00 3 0.09 37.50 Dallas 622 19.19 77.65 175 5.40 21.84 
Dalworthington 
(Gardens) 1 0.03 100.00 De Soto 9 0.23 69.23 4 0.12 30.77 Duncanville 8 0.25 61.54 5 0.16 38.46 
Cedar 
1 0.03 33.33 2 0.06 66.67 Euless 93 2.87 43.26 113 3.49 52.56 
Everman 
Branch 45 1.39 80.36 11 0.34 19.64 Fort Worth 53 1.64 20.62 197 6.08 76.65 Foresc Hill 1 0.03 33.33 
Fanners 2 0.06 66.67 
40 1.23 0.28 18.37
Garland 81.63 9 
30 1.20 44.32 49 1.51 55.68
Grand Praire 
50.54
45 1.39 48.39 47 1.45
Grapevine 
5 0.16 25.00 14 0.43 70.00 Highland Park 1 0.03 50.00 1 0.03 50.00 Hurst 62 1.91 44.93 75 2.31 54.35 Hutchins 1 0.03 100.00 Irving 230 7.10 56.93 169 5. 2\ 41.83 Keller 17 0.52 56.67 13 0.40 
Raltom City 43.33 
Kennedale Lancaster 0.12 100.00
4 
Mans filed 3 0.09 100.00 Mesquite 17 0.52 63.00 7 0.28 28.00 no. Richland Hills 15 0.46 25.42 42 1.30 71.19 Pantego 1 0.03 100.00 Richardson 24 0.74 72.72 9 0.28 
27.27 
47.37
Richland Hills 10 0.31 52.63 9 0.03 
50.00
River Oaks 1 0.03 50.00 1 0.03 
Sachse 1 0.03 100.00 Saginaw 1 0.03 50.00 
1 0.03 50.00 
Seagoville 1 0.03 50.00 Smithfield 5 0.16 
33.33
9 0.28 60.00 
0.16 55.55
South Lake 4 0.12 44.44 5 0.06 11.11
University Park 15 0.46 83.33 2 
0.25 66.67
Watauga 4 0.12 33.33 8 
0.09 100.00 
Willmei' 

White Settlement 3 
1 
0.03 100.00 
42.76
TOTAL 1,592 49.09 55.70 1,222 37.70 
36 
Field 
Love 
at 
work 
to 
used 
who 
Employees 

DFW 
of 
Distribution 

4.9. 
Figure 

Area. 
Study 

Intensive 

the 
Inside 
Cities 

by 
Airport 

37 
At 
Work not Did Who 
Employees 

DFW 
of 
Distribution 

4.10. 

Figure 

Area-
Study 

Intensive 

the 
Inside 
Cities 

by 
Airport 

Field 
Love 

of DFW exhibit
percentage employees, a higher percentage of DFW employees who did not work at Love Field than DFW employees who did. 
As can be seen in Table 4.1A, the City of Dallas has the highest sample frequency of DFW employees inside the ISA (801 DFW employees). Table that 77.7
4.7 A shows percent of DFW employees (622 employees) who 
live in Dallas indicated that, they used to work at Love Field Airport. This corresponds to 19.2 percent of the total employees surveyed at DFW to the total DFW
airport. Applying similar procedures employees who live inside the ISA, it is found that 55.7 percent of the total DFW employees that live inside the ISA indicated they used to work at Love Field Airport, which corresponds to the 49.1 percent of the total DFW employees surveyed. With regard to ISA, city with
cities outside the Lewisville is the the highest percent of DFW employees (see Table 4.IB). Table 4.78 shows that 65.5 the who live in Lewisville indicated
percent of DFW employees 
they used to work at Love Field, and this corresponds to 2.3 percent of 
the total employees surveyed. Also it can be noted that most of the DFW live outside the ISA used to work at Love Field. That
employees who 
is to say, although 68.4 percent of the DFW employees who live outside 
the ISA indicated they used to work at Love Field, this represents only 
the total DFW
6.8 percent of employees surveyed. 
Distribution by Type of Vehicle Figure 4.11 shows the distributions of both subsets of DFW employees by type of vehicle. It should be noted that current DFW who did not work at Love Field use the
employees 
"carpool" mode more frequently than former Love Field employees. Table 

4.8 shows the most frequent combinations of type of vehicle. "Own car or carpool" was the most frequent combination in both cases. 
Distribution By Type of Work, As seen in Table 4.9 which compares the "Percent of DFW Employees Within Category" for both subsets, Professional, Sales, Service (Airline), and Rent-a-Car, are the catein which the used work
gories percentages of DFW employees who to at Love Field are significantly higher than the percentages of DFW employees that did not work at Love Field Airport. On the other hand Service 
(food), 
Service 
(custodian) and Hotel Employee are the categories in which DFW 
TABLE 4.78. 
DISTRIBUTION OF DFW EMPLOYEE RESIDENTIAL LOCATION 
BY CITIES OUTSIDE THE INTENSIVE 
STUDY AREA ACCORDING TO PREVIOUS WORK PLACE 
FORMER LOVE FIELD 
WORKER N0N-L0VE FIELD WORKER 
PERCENT 0? DFW 
PERCENT OF DFW
CITY 
SAMPLE 7. OF TOTAL SAMPLE % OF TOTAL
EMPLOYEES 
EMPLOYEES
FREQUENCY DFW EMPLOYEES FREQUENCY DFW EMPLOYEES
WITHIN CATEGORY: 

WITHIN CATEGORY 
Allen  2  0.06  100.00  
Alvord  1  0.03  100.00  
ArgyLe Aubrey  6 1  0.19 0.03  85.71 100.00  6  0.03  14.29  
Blue Ridge  2  0.06  100.00  
Bonham  1  0.03  100.00  
Bowie  1  0.03  100.00  
Hoyd  3  0.09  50.00  3  0.09  50.00  
Bridgeport  2  0.06  50.00  2  0.06  50.00  
Celina  2  0.06  100.00  
Celeste  1  0.03  100.00  
Cleburne  1  0.03  100.00  
Clifton  1  0.03  100.00  
Collinsville  1  0.03  100.00  
Conroe  1  0.03  100.00  
Decatur  1  0.03  100.00  
Denton  24  0.74  77.42  7  0.22  22.58  
Elmo  2  0.06  100.00  
Ennis  2  0.06  100.00  
Fairfield  1  0.03  100.00  
Farmersville  1  0.03  100.00  
Ferris  1  0.03  50.00  1  0.03  50.00  
Flower Mound  1  0.03  50.00  1  0.03  50.00  
Frisco  5  0.16  62.50  3  0.09  37.50  
Gainsville  1  0.03  100.00  
Granbury  1  0.03  100.00  
Gordon  1  0.03  100.00  
Greenville  1  0.03  100.00  
Joshua  1  0.03  50.00  1  0.03  50.00  
Justin  1  0.03  25.00  3  0.09  75.00  
Highland Village  1  0.03  100.00  
Kerns  1  0.03  100.00  
Lake Dallas  4  0.12  66.67  1  0.03  16.67  
Little Elm  1  0.03  100.00  
Lewisville  74  0.23  65.49  36  1.11  31.85  
Mabank  2  0.06  66.67  1  0.03  33.33  
McKinney  7  0.22  77.77  2  0.06  22.22  
Midlothian  1  0.03  100.00  
Nevada  1  0.03  100.00  
Nocona  
Paradise  1  0.03  100.00  
Plano  18  0.55  69.23  8  0.25  30.77  
Pilot Point  2  0.06  100.00  
Ponder  1  0.03  50.00  ' 1  0.03  50.00  
Poolville  1  0.03  100.00  
Quinlan  1  0.03  100.00  
Red Oak  2  0.06  100.00  
Rhone  1  0.03  100.00  
Roanoke  20  0.61  66.67  10  0.31  33.33  
Rockwall  1  0.03  33.33  
San Marcos  1  0.03  100.00  
Sange  2  0.06  66.67  1  0.03  33.33  
Sunset  1  0.03  100.00  
Tioga  1  0.03  100.00  
Valleyview  2  0.06  100.00  
Weatherford  1  0.03  33.33  2  0.06  66.67  
Waxahachie  4  0.12  80.00  1  0.03  20.00  
Wills Point  1  0.03  100.00  
Whitewright  i  1  0.03  100.00  
Wylie  0.12  100.00  i  
|  
TOTAL  219  6.76  68.44  i  94  2.90  20.39  
—  

40 
Place. 

Work 
Previous 

to 
According 

Mode 
Travel 

Work 
Employees 

DFW 
4.11. 

Figure 

DFW CATEGORY
OF 
42.19 27.03 31.58 56.25 60.00 39.73
PERCENT EMPLOYEES WITHIN
WORKER
TRIPS WORK FIELD OF DFW 
PERCENT EMPLOYEES 0,83 0.31 0.19 0.28 0.19 1.79
TOTAL
THEIR -LOVE 
IN NON
27 10696 58
PLACE SAMPLE FREQUENCY
EMPLOYEES WORK 
DFW DFW 
CATEGORY
BY OF
PREVIOUS 
57.81 72.97 63.16 43.75 40.00 59.59
USED 
TO 
PERCENT EMPLOYEES WITHIN
WORKER 
OF
COMBINATIONS ACCORDING 
FIELD DFW 
.12
1.14 0.83 0.37 0.22 2.68 
MODE LOVE PERCENT TOTAL EMPLOYEES 
4.8. FORMER 
37 27 1274 87
TABLE SAMPLE FREQUENCY 
orororor or 
off or off
COMBINATION Vehicle Carpool Vehicle Surtran Vehicle Dropped Vehicle Other Vehicle Carpool Dropped
Own Own Own Own Own TOTAL 
TO PREVIOUS WORK PLACE  LOVE FIELD WORKER  PERCENT OF TOTAL DFW EMPLOYEES PERCENT OF DFW EMPLOYEES WITHIN CATEGORY  10.24 33.50  5.58 46.53  1.60 38.89  0.74 33.33  2.06 59.29  3.70 51.50  6.75 58.56  3.51 60.64  3.27 21.59  0.89 78.38  0.33 37.93  0.89 90.62  1.67 47.79  41.28 41.28  
ACCORDING  NON SAMPLE FREQUENCY  332  181  52  24  67  120  219  114  106  29  11  29  54  1,338  
EMPLOYEES BY OCCUPATION  WORKER  PERCENT OF DFW EMPLOYEES WITHIN CATEGORY  65.35  52.19  70.56  59.72  38.94  45.92  40.37  36.17  78.21  21.62  62.07  0  44.25  '  57.08  
OF DFW  LOVE FIELD  PERCENT OF TOTAL DFW EMPLOYEES  19.96  6.26  3.92  1.33  1.36  3.30  4.66  2.10  11.85  0.25  0.56  0  1.54  57.08  
DISTRIBUTION  FORMER  SAMPLE FREQUENCY  647  203  127  43  44  107  151  68  384  8  18  0  50  1,850  
4.9.  
TABLE  OCCUPATION  Professional  Clerical  Sales  Craftsman Foreman  Technician  Operator  Maintenance  Other Labor  Service (Food)  Service(Airline)  Service(Custodian)  Rent-a-Car  Hotel Employee  No Response  TOTAL  

employees that did not work at Love Field 
exhibit significantly higher percentages Love employees.
than former Field 
It can be noted that those in the "professional" category who 
indicated that they formerly worked at Love Field constitute the higher 
"Percent of Total DFW Employees" (19.96%). 

Distribution by Level of Income. Figure 4,12 shows each DFW em
ployee subset by level of income. As expectly, most of the DFW employees 
with high income levels previously worked at Love Field Airport. Also, 
it should be noted that the former Love Field employees have a signifi
cantly higher percent in the $13,000 $20,000 income level range than
-
the DFW who did not work at Love Field. On the other
employees hand, 
DFW employees who did not work at Love Field present a significantly 
level than DFW

higher percent of the "under $6,000" income range employ-
who did.
ees 
Distribution by Age and Sex. Table 4.10 shows the distribution of 
the subsets of DFW employees by age. Comparing the two subsets with re-
be that the
gard to percent of t>;fw Employees Within Category”, it can seen 
category "Under 21” persents a higher percent of DFW employees in the 
"Non-Love Field Worker” subset than in the "Former Love Field Workers” 
Field
category. In the remaining categories, the subset "Former Love 
Worker" shows higher percent than "Non-Love Field Worker." 
Table 4.11 shows the distribution by sex for both subsets. 
It can 
be noted that the difference of percent between males for both subsets 
is significantly greater than the difference of percent between females. 
This can be understood as an increase of job opportunities for women. As can be seen in all the above tables, most of the 
percentages of DFW within each do not to hundred The
employees category sum one percent. explanation for this is that the question concerning whether or not 
the 
at Love Field not answered
employee previously worked was by all employ
ees 
. 
In conclusion, the analysis of data presented in this chapter allows the study of such topics as: 
1. The 
impact of Dallas-Fort Worth Regional Airport in the distri
butions of DFW employee residential location. 
DFW According to Previous Work Place. 
Figure 4,12. Employees by Income Level 
TABLE 4.10. DFW EMPLOYEES BY AGE ACCORDING TO PREVIOUS WORK PLACE 
FORMER LOVE FIELD WORKERS NON-LOVE FIELD WORKERS 
PERCENT 
SAMPLE SAMPLE 
AGE TOTAL DFW EMPT.OYEES TOTAL DFW EMPLOYEES 
FREQUENCY FREQUENCY 

PERCENT OF OF DFW PERCENT OF PERCENT OF DFW 
EMPLOYEES WITHIN CATEGORY CATEGORY
EMPLOYEES WITHIN 
Under 21 29 0.89 11.55 214 6.60 85.26 
-
21 34 863 26.63 53.97 715 22.06 44.72 
-
35 44 528 16.29 68.30 234 7.22 30.27 
-
45 54 311 9.60 71.99 117 3.60 27.08 
-
55 64 98 3.02 71.01 39 1.20 28.26 
Over 65 4 0.12 57.14 3 0.09 42.86 
No Response 17 0.52 40.46 16 0.49 39.02 
TOTAL 1,850 57.08 57.08 1,338 41.28 41.28 
TABLE 
4.11. DFW EMPLOYEES BY SEX ACCORDING TO PREVIOUS WORK PLACE 
FORMER  LOVE FIELD  WORKERS  NON-LOVF  FIELD WORKERS  
PERCENT OF  PERCENT  OF  DFW  PERCENT  OF  PERCENT  OF  DFW  
SAMPLE  SAMPLE  
SEX  TOTAL  DFW  EMPLOYEES  TOTAL  DFW  EMPLOYEES  
FREQUENCY  EMPLOYEES  WITHIN CATEGORY  FREQUENCY  EMPLOYEES  WITHIN  CATEGORY  
Male  1,237  38.17  60.28  790  24.38  38.50  
Female  570  17.59  51.58  518  15.98  46.88  
No Response  43  1.33  51.19  30  0.92  35.71  
TOTAL  1,850  57.08  57.08  1,338  41.28  41.28  

2. 	The distribution of DFW employees by type of vehicle, reflecting necessity of considering the DFW employee vehicles as
the a significant part of the total volume of vehicles at the air
port access highways. 
3. 	The common and different characteristics of DFW employees who formerly worked at Love Field Airport in comparison with those employees 
who 	did not. 
CHAPTER V. ARRIVAL AND DEPARTURE DISTRIBUTIONS OF DFW EMPLOYEES 
The DFW taken a is of the
Airport, as whole, one largest employers in the Dallas/Fort Worth area and as such is a major traffic generator from the standpoint of employee vehicles alone. In addition, the arri
vals and departures of employees adds to the traffic volumes generated normal considered
by airline activity. This must be both in modeling airport access volumes and in the subsequent design of airport access facilities. Therefore the distribution of employees' arrival and departure times at the airport relative to their work shift times is of 
critical interest in this research. The purpose of this chapter is to describe the conceptual basis for the distribution of DFW employees’ arrival and departure times
modeling at the airport relative to the work shift starting and ending times. The term "time difference" will be applied to the difference between the starting work shift time and the time that DFW employees arrive, or 
the difference between the time that DFW leave the
employees airport and their ending work shift time. The term "time-difference data distribu
tion" will be applied to the statistical distribution of the above time differences. is divided into sections. The first section deals
two
This chapter 
with the selection of the periods of day for which the time-difference data distributions are going to be specified. The second section deals tests of selected theoretical distributions against
with goodness-of-fit 
the observed time difference data distributions. 

Determination of the Periods of Day for Analysis 
Figures 5.1 and 5.2 show histograms of the percentages of DFW employees versus starting work ending work shift times,
shift times and 
respectively, during a normal work day. In the case of percent of DFW employees versus starting work shift time, it can be Fig.
noted from 5.1 
that there limits
are five distinguishable periods whose were tentatively 10 for the
selected as: 0 through 4 for the first period, 4 through second period, 10 through 13 for the third period, 13 through 20 for the 
47 
48 
Employees 

DFW 
of 
Times 
Starting 

Shift 
Work 

of 
Distribution 

5.1. 
Figure 

49 
Employees. 

DFW 
of 
Times Ending Shift Work 
of 
Distribution 

b.2. 
Figure 

20 	24 for the With the of
fourth period, and through fifth period. help "Statistical Package for the Social Sciences" (Ref. 10) a range of alternative limits around the above tentative ones were tested, and fixed when for each period the percent of DFW employees distributed by five-
change significantly. objective was
minute interval tended to 	The to 
distinguish time periods during which the arrival or departure patterns of employees remained approximately constant. A followed for the
similar procedure was ending work shift periods. The definitive limits of the periods based on this process are presented in Table 5.1. 
this
In way, knowing the periods during the day, the next step was to determine which theoretical distribution best the time dif
explains ference data distribution for each of the above 
periods. 
Determination of the Theoretical Distributions 
With regard to selecting a test to use for goodness-of-fit analysis, Refs. 1,2, 5, and 13 describe the Kolmogorov-Smirnov Test (K-S Test) as a quite sensitive test for continuous distributions with certain advan
the
tages over Chi-square test. Bradley (Ref. 1) for example, presents 
a discussion of the K-S test vis-a-vis the Chi-square. From that dis
cussion the following conclusions are taken: 
1. 	modest that
The K-S test requires only the relatively assumption 
sampling is random and that the sampled population is continuous 
Whereas 	other conditions that
Chi-square assumed among things, can be fulfilled only when sample size approaches infinity. The
2. 	Chi-square test is only an approximate test, at all sample and is hard to where-
sizes, the degree of approximation assess, as the K-S test is exact at small sample sizes and its degree 
of approximation is more readily assessable. 
3. The K-S test uses 	every observation repre
ungrouped data, i.e., 
sents a 
point at which "goodness-of-fit is examined; Chi-square loses this information (if the hypothesized distribution is continuous) by requiring that data be grouped into cells. 
Therefore, 	the K-S test was chosen the test to
as 	be applied in this study 
TABLE 5.1. LIMITS OF PERIODS OF STARTING AND ENDING WORK SHIFTS 
TIME  
PERIOD  
STARTING  WORKSHIFT  ENDING  WORKSHIFT  

0 06 Second through 9 6 through First through 5 through 
5 10 Third 9 through 13 10 through 14 Fourth 13 through 21 14 through 19 
Fifth 21 24 19
through through 24 
The K-S test is based on the simple measurement of the maximum vertical difference between two cumulative distribution functions; in this case, probability
the cumulative time difference data distribution and selected theoretical cumulative probability distributions. This difference, once determined, compared
is then with the values of K-S statistics for the appropriate sample size, and level of significance. A significance level of 0.05 is assumed in this study. Figure 5.3 shows an example of a typical distribution of the time difference data. From a simple visual inspection of the time difference data distribution for each period, the following observations are made: 
1. The theoretical distributions most likely to fit the time 
difference distributions are: 
a) Normal distribution 
b) Lognormal 

distribution 
distribution 
d) Gamma distribution 
e) Erlang 

c) Exponential 	(negative) 
distribution asked what time they arrived at. left
2. 	When DFW employees were or the airport there was a tendency to express their answer to the nearest five minutes. Therefore, the intervals selected, were five minutes in length centered around even five-minute epochs, 
i.e., the actual boundaries were defined according to the formula 5 where is
(N + 0.5), N a positive integer (the first interval has the bound These interval
integer lower of 0.0). boundaries also ensure that individual integer responses 
fall 
within an interval and not on a border between two intervals. 
Figure 5.4 is a flow chart which was developed to show the sequence of the steps followed in finding the theoretical distribution that best fits the time difference data distribution. First of all, the data were divided randomly into two parts. This division was made because of the K-S that the the theoretical distribution
test requirement parameters of should not be obtained from the same sample that is tested (Ref. 5). From one subset of the data the mean and variance are computed
sample 
and used to estimate the parameters of each theoretical distribution. 

53 
Data. 
Time-Difference 

Of 
Distribution 

Typical 

5.3. 
FIGURE 

| 
(0.05) 
SIGNIFICANCE 
r 
1 
OF 
LEVEL 
. 
1
SPECIFY TABLE VALUE VERTICAL STATISTICS GREATER 
K-S VALUE HYPOTHESIS
riSTICS 
'YES 
rr r 
15 
I1
STATISTICS 
STA1 VERSUS REJECT
MINIMUM-MAXIMUM STATISTICS Distribution
X-S K-S 
USE EQUAL? NOT 
DO
OBTAIN K-S 
IS OR
SIZE COMPARE DIFFERENCES 
VALUE 
r
SAMPLE 
GET »” 
Difference 
pp 
PROBABILITY CUMULATIVE WITH DIFFERENCES -Time
DISTRIBUTION DATA 
HYPOTHESIS
f 
rf
CUMULATIVE THEORETICAL DISTRIBUTION TIME-DIFFERENCE VERTICAL
■ Fitting
DIFFERENCE 1 EACH MAXIMUM REJECT 
CONSTRICT TIME COMPARE PROBABILITY CUMULATIVE DISTRIBUTION for
GET
DIFFERENCE SAN 
FITS | 
fewW 
f
MAXIMUM DATA DIVIDE 
INPUT \ 
1
DISTRIBUTION MINIMUM RANDOMLY Procedure THEORETICAL CUMULATIVE 
4 
OF 

THEORETICAL DISTRIBUTION --
VARIANCE5.4. 
AND ETICAL 
DISTRIBUTION 

MEAN 1 
THEOR
HYPOTHESIS: f PARAMETERS ’ Figure
TIME-DIFFERENCE 
COMPUTE
NULL 
COMPUTE DISTRIBUTION CONSTRUCT PROBABILITY 
From the other of the the observed
part sample cumulative probability time difference data distribution is constructed. Then, each cumulative theoretical distribution is compared with the observed cumulative time 
difference data distribution, and the maximum vertical differences are 
obtained. The minimum of the maximum vertical differences is compared 
with the K-S test statistics (Ref. 6). If the computed value is less 
than the 0.05 K-S statistic, it can be said that there is no reason to 
reject hypothesis corresponding
the that the theoretical distribution fits the time difference data distribution at the 
specified significance 
level of 0.05. and show the of
Tables 5.2 5.3 periods day that particular theoretical distributions fit the time difference data distributions along with the functions and estimated Note that
corresponding density parameters. the theoretical distribution that fits most often is the Gamma, while the remaining satisfactory theoretical are particular cases
distributions of 
Gamma distribution. This does not discard the possibility that other 
distributions might also fit the time difference data distribution, but 
only presented nearly
that Gamma the minimum-maximum difference in all cases. Figures 5.5 and 5.6 show a sample graph which compare the actual observed distribution and the theoretical distribution for DFW 
employees originating leaving airport, respectively.
at and the 
OF 
0.1 0.05 0.05 0.05 0.05
LEVEL 
SIGNIFICANCE
SHIFTS. 
WORK 
K-S VALUE .189 .049 .105 .069 .105 
THEIR 
0.52 1.15 0.065 1.13 0.053 1.44 0.067 1.67 0.042
PARA-METERS 
== ====== =
STARTING 
ak akakak a 
I
EMPLOYEES )/(k-1) 
DFW ak
DISTRIBUTIONS FUNCTION exp" 
-ax
OF 
(k_1 aexp
PERIODS DENSITY } 
(a N 
k x Same Same Same = 
= 
OF THEORETICAL f(x) 
CHARACTERISTICS f(x) _, Exponential
NAME Gamma Gamma Gamma Gamma 
Negative
5.2. SAMPLE SIZE 41 716 166 379 144 TABLE 
59 1321 24 
OF (HRS)
PERIOD DAY through through through through through 
059 13 21 
OF 
LEVEL 0.05 0.10 0.10 0.05 0.05
SIGNIFICANCE 
SHIFTS. 
K-S VALUES .131 .106 .138 .052 .074
WORK THEIR 0.078 2.0 0.081 1.79 0.036 1.46 0.057 0.065 1.64 
==== == = ==
PARA-METERS 
akak ak ak
a
ENDING 
1
EMPLOYEES 	)/()» )»
ak
DFW 
FUNCTION ak 
-otx 
OF 
DISTRIBUTIONS _ 1^exp" 	aexp
k 
Same 	Same 
xx^ 	=
PERIODS DENSITY "k n (a 
. 
k 1 	f)x)
OF 	=
THEORETICAL 	f(x)=(a
f(x) 
NAME
CHARACTERISTICS 	Up) Negative ExponentialErlang (Round Gamma Gamma Gamma 
107 131 78 686 341
5.3. SAMPLE SIZE 
6 10 14 1921
TABLE 
OF (HRS.)
PERIOD through through through through through
DAY 
06 10 1419 
Figure 5.5. Sample Distribution for DFW Employees Arriving at The DFW Airport. 
Figure 5.6. Sample Distribution for DFW Employees Leaving the DFW Airport. 
CHAPTER VI. MODEL DEVELOPMENT 
ofis to
The purpose this chapter to present the methodology used model which be used the
develop a can to estimate employee vehicular volumes arriving at or leaving the airport. It is divided into three 
sections. The first section deals with the fundamental and theoretical 
concepts used in developing the model. The second section explains the 
flow chart followed in the the
computer program, required input data, 
and the nature of the The third section deals with the cali
output. bration and testing of the model. 
Fundamental and Theoretical Concepts 
it found
In Chapter V, was for particular periods of time during 
the day, the way by which DFW employees arrive at or leave the airport 
can be a distinct theoretical distribu
satisfactorily approximated by 
tion (one distribution for arrival and one distribution for leaving). 
That is to say, for all the DFW employees whose work shift times in a 
certain period, the way that they arrive at or leave the airport will 
be distributed approximately the same. 
Figures 6.1 and 6.2 show examples of how DFW employees arrive 
at 
and leave the airport, respectively. Note that the area under the curves 
for a given time slice (t, t + At) represents the probability that DFW 
employees, whose work shift times are and arrive at or leave the 
airport, respectively, during that time slice. The expected number of 
employees arriving at or leaving the airport for a time slice (t, t+ At) 
is given by the product of the above probabilities and the total number 
of employees corresponding to the work shift times example,
or For 
in Fig. 6.1 denote by P(t, t + At) the probability that an employee corn
ing airport to begin his work shift at airport
to the arrives at the 
in (t, t + At). Then, the in
expected number of arriving 
(t,t+ )is: 
E {N(t, t + At)} = N {P(t, t + At)} (1) 
where E{N(t, t + At)} is the expected number of employees with work shift time 
t 1 arriving in the (t, t + At) time slice, and N is the total 
59 
Figure 6.1. Employee Vehicles Arrival Patterns (Single Work Shift Time). 
6.2. 	Patterns (Single Work Shift Time). 
Figure Employee Vehicles Departure 
of 
.
volume (number) employees with work shift time t 
not 	sufficient to calculate precisely the
Although, data are variance of N(t, t 4-At), it is possible to get a rough estimate of this 
the
variance under following assumptions: 
of
1. 
Each the N DFW employees starting his work shift at has 

the same probability P(t, t + At) of falling into interval (t, t 4-At). 

2. 	
Each employee essentially constitutes a Bernoulli trial with the same probability, i»e., the above probability applies inde


all	N of
pendently and identically to employees a particular work shift. These 
assumptions mean that the number of DFW employees arriving at of each slice
or leaving (independently other) the airport at time 
has binomial

(t, 	t 4-At) a distribution with mean N (P(t, t + At)} and 
+
varianceN {P(t, t At)} {1 P(t, t 4-At)}. 
So far, the following conclusions be made:
can 
1. 	For DFW employees arriving at the DFW airport and whose work shift starting of employees arriving at
time is the number 
time interval (t, t + At) 'has a binomial distribution with mean 
N {P(t, t + At)} 
and 	variance 
-
N {P(t, t + At)} {1 P(t, t + At)} 
Where N is the total of DFW employees whose work shift starting time is and P(t, t + At) is the probability that employees arrive in time slice (t, t + At). 
2. 	For DFW employees leaving DFW airport, similar assumptions are made. The number of DFW employees with work shift ending time 
is leaving during the time interval of (t, t + At) has a binomial distribution with mean 
M {Q(t, t + At)} 
and 	variance 
-
M {Q(t, t + At)} {1 Q(t, t + At)} 
where M is the total of DFW employees whose work shift ending time 
-
is and Q(t, t + At) is the probability that employees leave in time slice (t, t + At). Figure 6.3 and 6.4 show multiple work shift time curves (start
ing ending). Considering work shift starting time, for example,
and note 
that many curves overlap in time slice (t, t-At). Therefore, the total volume the in
expected of employees arriving at airport a particular time slice (t, t + At) is the sum of the individual volumes of 
in that time each
employees slice from starting work shift. Assuming 
that the number of time slice for each start-
employees arriving in the ing shift is stochastically independent employees
work of the number of 
in that time slice for other also be 
any starting work shift, it can concluded that the variance of total volume of employees with different 
work shift starting time arriving in time slice (t, t + At) is the sum of each individual variance for that time slice since the variance of the 
sum of independent random variables is simply the sum of the individual variances. Similar considerations and conclusions apply to work shift 
volumes the the ends
ending times, and of employees leaving airport at 
of work shifts. Denoting by the number of employees arriving for the start of 
work shift i, i = 1,2, n. Then the total expected number of 
... 
in t+is
employees arriving at the airport (t, At) 
n 
E{N(t,t+At)}=lN. {P(t,t+At)} (2)
total° 3 i1 i=l where t + At) is the area under the employee arrival distribution for work shift i in time slice (t, t + At) as shown in Fig. 6.3, By the above assumption, the variance of + At) is 
n 
Var {N ,(t, t + At)}= YN. {P.(t, t + At)} {l -P (t, t + At)} (3)
total ~ii l. 
I=l 
the of the total number of
Similarly, mean and variance employees leaving airport a particular time slice from various ending work
the in shifts 
are 
Figure 6.3. Employee Vehicles Arrival Patterns (Overlapping of Curves) 
Figure 6.4. Employee Vehicles Departure Patterns (Overlapping of Curves). 
m 
E	=
{M(t1+At)* IM t+At)} (4)
total’ 	i
1 
-
j=l 
and 
m 
Var M 
(t	t+At)}=lM.{Q.(t,t+At)}(l Q.(t,t+At)}(5)
total » 	
J	2 J
j=l 
Where M is the number of employees leaving at the end of work shift j 
-
and +
Qj(t> t At) is the probability that work shift j employees leave the airport in the time interval (t, t 4-At), i.e., the area under the employees departure curve for work shift j in time slice (t, t + At) is shown in Fig. 6.4. The above 
equations (2), (3), (4), and (5), are all that one needs to compute the mean and variance of the total number of employees arrivand

ing at leaving the airport in any arbitrary time slice (t, t 4-At). Procedures for the actual computation of employee volumes are described below. 
Computation of Employees Vehicular Volumes 
Figure 6.5 shows the flow chart followed to determine the volumes 
of DFW employees vehicles entering or leaving the airport. The procedure basically is: 
1. 	Selection is made of the of time to be
first period analyzed and the theoretical distributions to be considered for that period. 
2. The probability distribution is obtained for each interval of 
time 
specified (see input data). 
3. 	
From all the work shift time only the starting work shift times or the ending work shift times are selected, and from them only those that fall into the period selected. 

4. 	
The vehicular volume and the variance at each time interval are 


calculated and those vehicular volumes and variances with 
common time interval are summed. 
65 
Computation. 

Volume 
Vehicle 
Employee 

of 
Chart 
Flow 

6.5. 
Figure 

A 	is considered and 1
5. 
new period 	steps through 4 are repeated. 

6. 
Once all the periods are considered and the successive computa


tions made, the standard deviation is converted from the total 
variance for each time interval. 
7. For a given time interval, one standard deviation is added and 
subtracted to the expected value. In this way, one would be able 
to calculate the percent of the time that observations fall in the interval "Expected value plus and minus one standard deviation." 
A computer program has been devised following the above flow chart list of
and a the program is presented in Appendix B. An explanation of how to use it follows. 
Input Data 
data
To explain the input to the computer program, three types of input cards will be discussed (see Fig. 6.6). The first type,called the data control card,specifies the number of periods in columns 7-9, the number of work shift times punched in columns 10-12, and the length of time interval columns
punched in 13-14. The second card is called work shift data each of which
input card, contains informaton about a particular work shift time. Column 7 is used to specify whether the work shift is starting (by punching the number the number the
"l") or ending (by punching "2"), columns 8-11 time at which the work shift starts or ends; columns 8-9, the hour; and in 10-11, the minutes. A third value, the number of employees using 
their own vehicles, is to be punched in columns 12-16. Note that this is the number of employees using their own vehicle, not the total employees starting or ending their work shift. The reasons for this are 
as follows: 
1. 	
The fact was encountered that the percentage of employees using other than their own vehicles was very small (See Chapter lV). 

2. 	
The their is


percentage of DFW employees using own vehicles significantly high (See Chapter IV). 
67 
Structure. 

Data 
Input 

6.6. 
Figure 

estimate of the number of and leaving the airport in a given interval of time is obtained as the corresponding number of employees who use their own vehicles . Based on 
the above this is felt to be reasonable 
Thus, an employee vehicles entering 
reasons, a assumption. The third type of input data cards are called period data cards. Each period data card contains: (1) in columns 7-10, the time at which the period begins (hour in columns 7-8 and minutes in columns 9-10); 
(2) 
in columns 11-14 the time at which the period ends (hour in columns 11-12 and minutes in columns 13-14); (3) in columns 15-24 the mean and 

(4) 
in columns 25-34 the standard deviation of the time before and after 


the shift that employees arrive or leave in that period, respectively. 
Four decimals and the decimal point are allowed; (6) in column 35 a 
number is to be punched to indicate the type of distribution to be 
considered. This number code indicates the distri
following respective butions. 1 Normal distribution 
2 Log normal distribution 
3 Negative exponential 
4 Gamma 
5 Erlang down
rounded distribution 6 Erlang rounded up distribution In column 36, it is then indicated whether this distribution, with the mean, the variance, and the period specified, is to be used for employees if
arriving at or leaving, the airport. If it is arriving, punch "1"; 
it is leaving, punch "2", and if it is for both arriving and leaving, 
punch "3". Figure 6.6 shows the entire input data structure. 
Output 
The computer output is shown in Figs. 6.7 through 6.9. In Fig. 
6.7 the first part of the output is illustrated. The information about 
work shifts, is shown as a printout of the original "Work Shift Data Card". In Fig. 6.S a printout of the period data card is shown. These printouts allow the user to check the data for keypunch errors and data 
IF TS
•0*K 3H 
E**»iofee* 
ITAHTIM5 fWDI*C
TIKI 
1 82 
SI* 

5 
It 
ill 

31 
!JI 

13 
sss 

t 
SS* 33 
S«* t 
i« 6 
SS* »• 
555 
3# 
386 
1i 36 

»• 
•2
MS 
t
til 
»]• 150 
Hi 
It 
690
T• 
20
71* 
26
m 
13 
71* 3*1 
7SS 

7?» 
6 
53
7*5 
«• 
1022 
SIS 

30 
*?» 20 
• 98S3« 
It 
s • 180 
*1* 

STS 
20 
STS 16 
I* 

• 59 
ISIS 10 
1*1* 125 
1»«S 

It 
89
II • 
It
Ills 
IIS* 33 
ll«S 

10 
u « til 
IS S 

13 
ISt* 13 
ISIS 

16 
60
IS1* 
79
IS * 
!•
111* 
66 
ms 
19 
111* 

11SS 
•6 
6 
I* • 

11*5 
262 
l«IS 23 
1*2* 

26 
1*1* 

191 
Figure 6.7. Work Shift Information. 
P  E  R  I  0  0  S  
BORDERS  
LONER  UPPER  MEAN  VARIANCE  DISTRIBUTION  WAY  
0  0  5  0  22,0732  a 23,9695  GAMMA  ARRIVING  
5  0  9  0  17.a120  269.J5P3  gamma  ARRIVING  
9  0  13  0  27,3976  521,6399  GAMMA  ARRIVING  
13  0  21  0  26,7309  a28,6152  GAMMA  ARRIVING  
21  0  2a  0  23,5«17  309.9913  neg.  exponential  ARRIVING  

Figure 6.8. Periods Information. 
employee vehicle VOLUME 
An0I V N0 leAVIN0
I 
minimum MEIN MAXIMUM
MINIMUM MEAN MAXIMUM
Tine 
VOLUME VOLUME VOLUME 	VOLUMEINTERVAL VOLUME VOLUME 
0 0 0	0
08
015 
09
0 0 0	0
030 
09
0 00
Po5 0 
0 0 00 010 0 0 99 080
115 
00 0 	0 9 00
130 
0 00 0 
2a 8 

0
1*5 
0 00 	0 0 
00 0 	0
C0
215 
0 00 9 2*5 0 
90
230 
0 0 00 0 30 0 0 
0 00 	0 
990090
315 
00
•	0
330 
*t 2	009
3<n 
50 0 0 4iS 710 0 
*e 13 
00 433 1217 21 9 
5 
00 0
4046 00 
50 79 

33
4*5 
6700 0
70 
92101 •000 530 l40 0 
515 82 
00 3*5 207 221 
128 152 
000 6 370 
193 
4090 0 90 389 
09 630 166 179 192 
100 111 0
615 	121 
9 6*5215 245 0 0 00 
230 0 
70 

0050a7 1 0 0
430 
009 730 329 3*7 
188
715 162 175 
000
311 
909
340 356
745 321 
00 00 
S0598 	647 0
622 
0 830 280 314 0 
160 173 166815 
297 00 
805 5« 62 

700 00 « 0 94 115 
0 00
105 
00
21 0
«15 13 17 
0 00
610 2227 33 
0
31 4400
9*5 36 
00 0 
04451 58 	0 10 
61 R047 54
1015 
740 00
58 66
1030 
00 0 
4248 0
1045 35 
48550 a 0 01
11 
32 0 0 
26 37 90 
1115 
42480 e 0 
351130 
54620 0 12e.60 6977 0 0 0 1215 35 47
1145 
0
42480 0 
00
40 0754 0
123* 
0 90 IS067 0 ? 
02 4956
12*5 
7S64 0 
1315 64 

72810 00 
0
79680 0
1130 71 
0
108 12900
1345 119 
0 144 0 00
14 132 156 
97107117 9 0 0
1415 
0
120 80
1438 109 131 
0 150 125 
99
1445 182 112 123 
0
114 1360 0 
6068769 00
1515 
8291e 00
731530 
0 163 77 
687786 0 0
15*5 
0
86950 9 
4047540 00
1615 
44510 00
1638 38 
2328 0 90 170 23 33 
1645 	34 
28 000 
22260 88
171715 
0 1745283339 0 
30350 0
1730 25 
00 180 27 
3236 0 00 1815 
1115180 a0 1830 15 19 
00
24 0 
0 190 18 
141622 0 0
1845 
14 22000 1915 18 28
23 900 
0 00
33 39 46
1930 
0
59670 0
1945 51 
0 5966000
20 51 
69 	000
2015 	12 
9 000 2045 17 22 
2030 12 16 
260 00 
380 09
0 26 32
21 
490 00
43
2115 36 
0
61 7700
2130 69 
21 000 220 243035 0 
2145 13 17 
00 
0 00 	0 00
221S 
0 00 0 00 2245 0 
2230 
00 0 00 230 0 0 
0 9 0	0 
0 00 2330 6»600 
0 00
2315 
00 2345 0 0 
0 0 90 249 0 00 0 
90 
Figure 6.9. Calculated DFW Employee Volumes 
accuracy. Figure 6.9, the last part of the output, contains the results 
of the analysis. For each time interval a range of volumes is shown whose boundaries are the minimum volume (mean minus one standard devia
tion), the mean, and the maximum volume (mean plus one standard devia
.
tion) These values are printed out for DFW employees arriving at and leaving the airport. 
Calibration of the Model 
In order to estimate the total number of employees who begin and end each the
work shift for input to the model, same proportion of employees on each work shift as obtained in the survey sample was applied to 
the 
total DFW 

employees. 
The model was then used to obtain estimates of the number 

of employ
ees entering and leaving the airport in each 15-minute interval through
out the day* Equations (1) through (5) were used for this puroose. The 
results were then compared counts in
with the 15-minute manual made the May 1975 survey (Ref. 4), the of DFW employees arriving at or
and volume leaving the airport who drive their own vehicles. This latter volume calculation, called the "survey-response volume," was made according to the of the
responses to questions Employee Travel Survey (see ChapterIII) on the time that people arrive at or leave for work and their mode of travel 
to 
the airport. Tables 6.1 and 6.2 show the survey-response volumes, the model estimates, and the traffic count volumes arriving at and leaving the airport, respectively. Comparing model estimates of the volumes with the count volume in Tables 6.1 and 6.2, it can be noted that the two values differ substantially for given explanation differences
time intervals. The for these stems a misunderstanding of the on-airport travel patterns
from of airport employees on part of the project staff prior conducting the
the to 
survey in May 1975. The staff was under the impression that employees 
used only the service road system (see Chapter III) for their on-airport 
access to work. It was not until the survey was underway that it was 
discovered that employees also use the main ’’spine highway" as well. The 
172 1079390816982697165644043383948394119434246 18162826313848 43
160 37
COUNT VOLUME
AIRPORT). 
76918695545134332635393818242222284667661216 263849772135
117 131 123 136
MAXIMUM 
THE 
VOLUME 
9 
AT MEAN 107 120 112 125688277 8647 44 2828 2230 3332 151918 1823 395959 1222 324369 1730 
MODEL 
97 109 102 114607368774038232317252827111514141833515169172636611324
MINIMUM
(ARRIVING 
RESPONSE
SOURCES 
08 37
4 884 88
66 70 49702521 5433 291333 25 254545 107 54
111 202 103 148 120 124
SURVEY VOLUME 
DIFFERENT TIME INTERVAL 1401-1415 1416-1430 1431-1445 1446-1500 1501-1515 1516-1530 1531-1545 1546-1600 1601-1615 1616-1630 1631-1645 1646-1700 1701-1715 1716-1730 1731-1745 1746-1800 1801-1815 1816-1830 1831-1845 1846-1900 1901-1915 1916-1930 1931-1945 1946-2000 2001-2015 2016-2030 2031-2045 2046-2100 2101-2115 2116-2130 2131-2145 2146-2200 
FROM 
7962 575848 93728663 8992 80828678 9697 99
COUNT VOLUME 68 116 169 143 108 124 119 148 110 104 123 128 125 146 
VOLUME 
62 5481
70 2133 445861 744855 3748 77 48 56 84 88 129 156
121 192 245 471 188 347 358 647 186 314 115
MAXIMUM 
VOLUME
EMPLOYEE 
62 17 273851 5466 42483242 546942 474975 7279
MEAN 111 179 230 450 179 329 340 622 173 297 105 119 144 
MODEL
DFW 
160 5494132231444758354126354760354042676471 108 132
100 166 215 430 162 311 321 598 280
MINIMUM
OF 
RESPONSE
COMPARISON 
8 8 132
54 99 66 4141 5853 25826666 4566 379154 17391 206
107 272 207 466 210 433 280 751 173 338
SURVEY VOLUME 
6.1. 
TIME INTERVAL 601-615 616-630 631-645 646-700 701-715 716-730 731-745 746-800 801-815 816-830 831-845 846-900 901-915 916-930 931-945 946-1000 1001-1015 1016-1030 1031-1045 1046-1100 1101-1115 1116-1130 1131-1445 1146-1200 1201-1215 1216-1230 1231-1245 1246-1300 1301-1315 1316-1330 1331-1345 1346-1400
TABLE 
113 178 157 209 162 153 233 84 87 122 145 78 200 12278 78 58 61 60 60 127 11256 61 46 46 49 81 10047 64 47
COUNT VOLUME
AIRPORT) 
85 85 349 176 295 155 358 172 346 174 493 228 501 230 178 85 11457 78 35 111 52 33 30 39 37 49 38 60 63 69 62
MAXIMUM 
THE 
VOLUME 
7676 76 
50 28 4358
MEAN 331 163 279 143 340 159 328 162 471 213 479 216 165 104 70 30 101 45 25 34 32 32 56 61 55 
MODEL
(LEAVING 
6868 58 6793436224 91 2028 3727 54
313 150 262 131 322 147 311 149 450 199 201 153 38 23 26 50 48 48 
4
MINIMUM 
SOURCES 
RESPONSE 
78 54 
285 260 318 186 313 206 280 210 516 260 454 280 223 41 15333 62 41 66 82 49 33 54 21 45 33 87 49 87 37
VOLUME 
SURVEY
DIFFERENT 
INTERVAL 1401-1415 1416-1430 1431-1445 1446-1500 1501-1515 1516-1530 1531-1545 1546-1600 1601-1615 1616-1630 1631-1645 1646-1700 1701-1715 1716-1730 1731-1745 1746-1800 1801-1815 1816-1830 1831-1845 1846-1900 1901-1915 1916-1930 1931-1945 1946-2000 2001-2015 2016-2030 2031-2045 2046-2100 2101-2115 2116-2130 2131-2145 2146-2200
FROM TIME 
VOLUME 
51484525 7994395541423739434140313839377359756147503613528791
118 144
COUNT VOLUME 
818460368393 87 754221 18193628146787443101210713285238
105 111
EMPLOYEE MAXIMUM 
VOLUME
DFW 
45653328985
MEAN 
7275533174849678 10167361714163123 11 10234532 
OF MODEL 
723321005652
646746256675867091593013111225 18 183927
MINIMUM
COMPARISON 7 
RESPONSE 
4 8 448 00 4 0048
123331278 9974 1814137 12 5825 12 12 162116 6221
124
VOLUME 
6.2. SURVEY 
TABLE 
TIME INTERVAL 0601-0615 0616-0630 0631-0645 0646-0700 0701-0715 0716-0730 0731-0745 0746-0300 0801-0815 0816-0830 0831-0845 0846-0900 0901-0915 0916-0930 0931-0945 0946-1000 1001-1015 1016-1030 1031-1045 1046-1100 1101-1115 1116-1130 1131-1145 1146-1200 1201-1215 1216-1230 1231-1245 1246-1300 1301-1315 1316-1330 1331-1345 1346-1400 
74 
precise proportion of employees who use the spine highway is not known 
are in the conduct

at present (there plans project to a mini-survey to 
estimate as a function of time of it is

this proportion day). However, 
expected to be significant and a major explanatory factor for the 
difference between the model estimate and the service road traffic 

counts. 
One other piece of vital information that should help explain the above differences The model estimates
is employee vehicle-occupancy data. in Tables 6.1 and 6.2 are based on an assumed employee occupancy of one. 
obtained from later
The actual figure (which will be a mini-survey as a part of this project, but not reported in this thesis) is almost certainly greater than one. Figures 6.10 and 6.11 show graphically the comparison of the model estimates and the count volume for )FW employees arriving
r
and the
at leaving airport, respectively. 
Once the above two factors, i.e., the proportion of employee vehicles that use the spine road instead of the service road and the 
average employee vehicle occupancy, are obtained and applied to the model 
that there will be closer
estimates, it is expected agreement between 
the model estimates and the service road 
counts. 
the with the
Comparing, now, survey response volume model volumes, 
it can be noted that in almost all the cases the volume
survey response is near if not inside the interval of the model volume. This close 
agreement is certainly expected since the model parameters are based on the This verifies that the model itself
survey response. comparison only was The real validation of the model involves the
accurately defined. above comparison of model estimates with actual traffic counts. Similar conclusions are applied for the case of DFW employees leaving airport.
the 
Figure 6.10. Calculated and Counted Employee Volumes (Arriving at the Airport) 
Figure 6.11. Calculated and Counted Employee Volumes (Leaving the Airport) 
CHAPTER VII. CONCLUSIONS AND 
RECOMMENDATIONS 
The 	the
findings of previous chapters appear to justify the following conclusions and recommendations. 
1* 	The 
results of past airport employee travel surveys have not been 
widely published. 
2. 	With regard to the Employee Travel Survey 
a) 
The 
of 
the 
be and usable data
to 
were designed. 
b) Based the this 

design 
survey 
appears good 
on in the wording of each question in future should be wording problems experienced survey, 
surveys 
analyzed until nearly all potential misunderstanding is 
eliminated. 
On
c) questions allowing multiple responses, every response should be so that all
analyzed response possibilities are covered. In the for
survey question regarding occupation, example, all types of occupation existing at the DFW airport were not covered by the choices.
response 
d) The rate of response was 24.24 percent. 
3. 	To facilitate locating DFW employee residences in the study area, 
DFW employees were grouped by zones designated by the North Central Texas Council of Governments, and also at the city level. 
The city of Dallas exhibited the largest single portion of DFW employees, followed by Irving, Fort Worth and Arlington inside 
the 	intensive and Roanoke and
study area Lewisville, Denton, 
Plano outside the ISA. 

In of DFW 	whether
the comparison employees according to or not they used to work at Love Field, former Love Field workers live 
predominantly in the mid-cities area of the ISA and Dallas 
On the other hand non-former Field
County. 	Love workers live 
predominantly in the mid-cities area of the ISA and Tarrant 
County. 
4. 	The used
type of vehicle most commonly for work trips is the vehicle.
employee’s own 
the

5. 	In determining theoretical distributions that replicate the way that DFW employees arrive or leave the airport before or 
after the work 
shift starting or ending time, respectively, the 
76 
used the 	and
goodness-of-fit test was Kolmogorov-Smirnov Test, 
it 	was found that Gamma was the most satisfactory distribution 
round
for both cases. Negative exponential and Erlang up 
(particular cases of Gamma) were the next most satisfactory dis
tributions 
. 
In 	the to the volume of
6. development cf a methodology estimate 
features
employees the following are to be considered: 
a) the variance of the total expected volume in a time slice is the. of the variances of the
(t, t + t) sum 	expected 
value of each 	under the that
work shift curve, assumption the values for the various work shifts in the time slice 
(t, t + t) are stochastically independent, that
and the total volume in a time slice is independent of the volume in other time slices. 
The variance of the number in each time slice from each work shift is a rough variance determined according to the formula: 
b) 	of employees 
Var 	= 
-
(N(t,t+At) N (P(t,t+At)}{l P(t,t+At)}
± 
This calculation is based on the assumptions that each of the DFW 
employees with a common work shift time (starting or ending) has the same 
+ 	into interval and
probability P(t, t At) of falling 	(t, t 4-At), that 
each employee essentially constitutes a Bernoulli trial with that same 
probability. 
7. With to the 	constructed to determine
regard computer program 
vehicular volume, the following features are summarized: 
a) 	The program allows the user to divide the day into different time periods. is important
This because it has been shown that arriving or departing characteristics are different during the day. b) The program also allows the user to specify the length of a and
time slice for analysis, the type of distribution, the choice the distribution with different means
of using type of 
and 

variances and for employees arriving at and/or leaving the 
airport. 
other than employee travel volumes and entities other than the DFW airport. only required 
c) 	The program can be extended in use to quantities 
The condition is similar characteristics and behavior in time 
d) 	As part of the input data, the probability distribution of how the 
employees arrived at or leave the airport relative to their work shift times are required. This is a disadvan
of 	the need
tage because for estimation of these values. However, this is considered minor compared to the benefits to be derived from the model. 
e) 	Obtaining another required input data item, the vehicular volume, also can create problems. In the case of DFW employees, where it was found that they drive their vehicles
own 
of 	the
approximately 90% time, obtaining the vehicular volume 
is 	a simple task. For those cases where there is more than 
one type of vehicle used significantly, the determination of the vehicular volume can be found, but 
with greater difficulty. 
f) 	The program is limited to only 5 distributions. These distributions were selected 
as the most common ones for vehicular volumes of those DFW employees driving their vehicles. Significant vehicular volumes involving other of
types 
vehicles 
may have distributions other than the ones specified previously. 
8. In order to 	the model
verify that itself is accurately defined, 
the 
survey response volume was compared with the model volumes. 
Since 	are based on the a
the model parameters 	survey response, 
found.
close agreement was 
9. The only source for validating the model was the service road 
trafficcount made in May, 1975, but because of the lack of in
formation on the proportion of employees that use the service 
road and vehicle
employee occupancy, the complete validation 
could not be accomplished in this thesis 
10. Further the 	in
research concerning methodology is necessary 
order to investigate, for example: 
a) 	The advantages of utilizing the methodology in other cases where, for a common work shift time (starting or ending), different types of traffic are being noted. These different types of traffic might have different distributions or the same distribution with different mean and variance. As 
an 	example, there is the possibility of two significant pop
ulations of employees, one driving their own vehicles and 
the other carpooling, each one with a different distribution. 
b) 	It is recommended that a mini-survey be conducted to determine the 
the
proportion of employees who use spine highway 
79 
as distinguished from those who use the service roads and the vehicle statistics. Both of the above
employee occupancy data items should be obtained as a function of time of day. 
APPENDIX A 
81 
o occ 
1——
—
OOr-^, ( ©OcMi/icT'OoOcNOOOOOr-i <r — ooo 
—
In< i i 
O S-ddddddddddddddodddddddddddddddddddddddddooooooo©©
o 
H 
>* 
U 
w 
£ O' ifs.nrNfM^nONN^'-'H^nr^oDvDn 00^^ 
< 
in a 
u. 
<ojro
r-.»o>0^(sim-.T^vDr^c^0^r^inr^oOO'0-<(Nr-\<finvc.coONOtMro<riAvDi^co^r-i<<i^vr)^ r;3;^jO“ 
o 
KJ 
S22SS^8SSo5o85S!22Sl2Scoool2oeSS8ooooS888oSS®8N2SS8S
ZONE 
< 
o H OO OOO 
O oooooooooooooooooooooooooooooooooooooooooooo 
H 
BY 
Jh 
w2 
W 
gu 
o 

l:
EMPLOYEES 
< 
U
DFW 
ZONE 99 100 101 102 103 104 105 106 108 110 in 113 114 115 117 118 119 120 121 122 123 124 125 126 127 129 130 131 132 136 137 138 139 141 142 143 146 147 149 151 152 153 155 156 157 158 159 165 166 
A.l. 
.J
TABLE 8SS82SP;S2gKSgS£883S5SS2388SS8g8S22SgSSSSP:aSS2S2aS o 
HO tsiddddddddd—'ddddddddoooooooooooooocoocooooooocoooo 
H 
>O 

W 
O 

o 
*““* r-i *-< CS1
— 
< t: VC -4 »—< 
U
U2 
2 r^r^r^r^r^r^ooc»w»wwwwwww 
O •a-ininmminiAi^»nir»m\ONDvDvO\0'D'£>vO'0>o-f^r^r^ 
K 
nLriNfNjC'^LnNnMv£)nvDvD^^o,'^ffi ir'f^vDOf\Nvc>^0'na'vDrri^^(Oa' 
mOO<J-<J-OOsC5<N^OO0Cr~ir^r^r-Kr^f^^»-<O— <OOOOr-.OOOOcN/OOOcsir-<OCNiOO^OOOOo^-( c 
u-< 
H OOOOOOOO—ir-iOOOCOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
o 
HH 

>
o 
w2 
w 

H r-< *-< CsJ <T LO CNICNJ •—< 
10 c* 
u. 
r
—ifMflN?4vrCO!>Or-<N<fOO(N^'4vDCOC'CsjLriNCOO'-'Nnsfl^'^f r''jLOvC 
nnnnvj^^sr-ciTiininLri^iA^^OvX)^} 
o 
K1 
in^ann^NO'O'O'fv^cMC'vDvDa'N 
*—< rn
Lo vX^OOf— <000—iAO \D O O ir\ <TC0OOCNjr~-OO<r ro 
< OCOO*— 
o ooo^ooooooooo
£ 
OO OOOOOOOO 
>< 
O
(CONTINUED) H 
2 
•J u: 
r
2 Nr>)^n^cosDNmO'£><rc^'-<f,*ininNrifiN'£i<f evjm 
O' •"*< •"* — 
»—< —< <f rMM m CN—J CNI CM 
<B CN 
tO CE 
fcu 

1. 
A. 
f^otrtfMr^inveotsjirivDr^ooO'O—<t^m<tiA\orvCOO--<<Mninp-cooNCMro<fu-ivor'eoosq^<Mri<ju-ir^co<^o 
~ --

ts3 nnr,r-,r-)rlfnPir)r)r-,r^)r-ic->c-i(”'ir')r->mr-ir~)r~>r->(->('‘>c->r-!<'->r-)nr-ic'-)C1<->r-iC>c-)r->r->r->r')(->c-)r-ic-ir'><'">(«-ir-)
TABLE W 
9>«iruoO'4>D»ri^nnNtMnvonnNnoninO'N<fOini5Nicisnicn^ff>j-i<rj5|-<NNOin>i».in
— 
oiNinrvNruntOO-JiCN 
o lOOO
s OOO—lOOOOO—i—'OOOOOOOOOOOOrsl^O'-iOOOOOOOOOOOO'-'OOOOO—
H 
>* 
o 

W 
2 
U3 
iO 
w 
Bi 
U< 

to 
W 
2 

SSSSS?;SS33555§555KS?;R§SSSSSSSKC^RSP:SR§SSSSSS5SSg
O MCNJCMCMCMrMr\jC'JCMCMrM 
K CMCM(NJCMCMCMr>jrMCMCMCMfNiCMCMC\iCMCMCMC'JCMCMrMCMrMCM<NCMrNjCMrMCMrMCMCMCMCMC ,cm cm 
OF 
0.06 0.03 0.03 0.03 0.06 0.03 0.06 0.03 0.03 0.03 0.03 0.03 0.15 0.09 0.15 0.06 0.15 0.06 0.03 0.19 0.09 0.03 0.18 0.03 0.06 0.06 0.06 0.09 0.03
TOTAL
7. 
111 1 
21l1 1 11111535252 63 6 2223
22
SAMPLE FREQUENCY 
621 623 630 631 634 636 640 651 660 669 677 686 700 701 702 703 704 705 706 709 710 711 712 715 717 718 719 722 733
ZONE 
OF 
TOTAL 0.03 0.06 0.06 0.06 0.06 0.06 0.22 0.03 0.06 0.09 0.03 0.06 0.03 0.06 0.03 0.03 0.03 0.03 0.03 0.03 0.03 0.06 0.09 0.03 0.03 0.03 0.03 0.03 0.03 
(CONTINUED) 7. 
1 1 111 11111 
1 22227 2312 21 1 1 123 1 
2
SAMPLE FREQUENCY 
1. 
A. 523 528 529 532 533 535 536 537 539 542 544 54 54 550 551 552 555 557 572 578 580 583 596 598 602 608 613 619 620
TABLE ZONE 6 7 
12 12 L2 1212 
OF 
0.03 0.03 0.09 0.09 0.06 0.03 0.03 0.09 0.09 0.96 4.29 0.86 0.09 0.12 0.03 0.19 0.03 0.40 0.09 0.12 0.04 0.34 0.03 0.40
0. 0. 0. 0.0.
TOTAL
7. 
11 1613421 
1134 2 33 4 3444
34 
31 28 13 1113
139
SAMPLE FREQUENCY 
'3 
8 
4
ZONE 467 468 469 471 472 474 478 480 482 484 496 497 498 499 500 501 503 507 508 511 512 513 515 516 519 520 521 522 
•J nO nO coO' vOcO rnin CNa-. O'm On
CN ro COinroco O'CNO'O'O' CO CN co *"■*ooO O CNOO o -O»
CN-o 4CN oCOOooO CN oc
n-COo oO'o o 
o ododooOooOOOOO oo©oOO ooo o oood oo
o oo OO o O “ 
oooo oooooo
H 
>
O 
MX 
D 
CNo CN o> ’■" *“*
X O' rnCN 00 roo CO 00 roNOr-vONOCOr-4 -* — 
<r <7CN CO co CN CO -1CO
<r CNCN
< w 23 CO30 
cc
ZONE w 
b* 
BY 
X n* On co n-COonO co<r in SO cror^.00 CO
cc CNro m inNO _ CN CO vOcc CO O' O' OnO' oO 
o S!o
o OOo CN CNCN CN CN CNCN CN CN CN CN CNCN CN CN CNCNCNCNCN CNCNCNCNCNCN 
EMPLOYEES w 
coconOvO vOnOin CN
CN CNvOcoCNmO'SO r-*'sDCNroNOSOCOVOCOvOcoC^>rN NO 
fc 
r-< -<

CN O Oao O CN OooOooCooOoO ooa oc C r-<
CN cooo CN 
OO ooo OO o o Oo oCo
o o o OC oo OO ooOO o©O oocoooocooooOocoo oc
H H
DFW <r 
>< 
o 

w
WORKER 
c* o 
in CN <5-CN00 in ** *■*
£o <r ro CN<r inn* CN CN <N-4CN CNO' NO*-* — CN— CN CO<r CN COvO200 
C* 
U4

FIELD < u 
w 
X innOocO co inn-COO'O CN sCO'O'r—i v£> r—iCN NO CN inNO in

<r CN 00O' COO' CO O-O' 
CN cn CN CN CN coco rororo CO<r ininin minLninin NONO NOn-n-
LOVE 
< 
cococnO' n-c>O'conOco cn<r mcooCO iOO'.On00 vOvOCOO'-NO NO O' 
r-oco OO OcCmOO <$■CNmo*“mCNOCOCOoOoOO oOcvO coinc OO o 

FORMER 
OOo4 
CN 
o 
o ooOo O O
cOO OO OooOooOoOooooacoOoOoooooooOooooOOoOO H 
2. 
A. 
o 
•J 
X 

O'
X oc coCNcoro in ro roOn CNNOO'O' ©oCNCN COCN CN CNOsDCN 00 in
co CN00
TABLE t: >-X 
co CN 
CO a: 
b» 

u 
2: m CNco innO cccn o CN lonCr-^00Ono inSOCC nOn-o CO CO o
fN ro<r CNm CC CN inNO O' CO C ininin mininin sDvONO\J0v£>sDvONOsD O'O'O' O'O'O' CO o
r-r^. n.00oc00 CCococ OS O'O'oo 
ON CO COOJ COONONCO
ONnO -O O' ON COON CO CO COroro COONro ON COO' COCO COOJ CO COO' co CONO OJ 
0 OO O CO O O_* O0OO O0OO00OOo > OP NO O' 
Pro NO O'cn nO ON cn ro cn v-O 
OO0 roO_Oro O0OO0 0 OOOOOOOO 0 
OOOOOOOOOOO0OO0O0OO00 OOo 0OOO0OOOO0OOOOO0 OO0O003 OO 
ro 
H 
h; 
0 
71 z 
r* 

d 
d 
XO ro 

iO O CO roro tOro 
O Coro COro NO CDrop 03 ro p ro 
roNO ro 
in 73 

> 
CO 
ro NO ON O 'O O NOO'00 COnO0 0 P0o 
Z 00NOO co 'O NOO CO0 COcnO' ro ro
pp cnON ro 
ro N)
cn 00CDNOO ro COCo NO NO OOOO roro rorocoCOCOCONOCO P
00NO NO NO
Op coP PP 
cn cn cn cncnOnon cn onO'O'O' ONO' ON ON ONp
cn 'O' cn Cncn cn cn cn cn cn Cn Cncn 
ro O' CJNON COONNOCO COO'ONCONOCO COO'COCo v^J CO ~J ONO'ONCOOJcoPCO r-0 O0OO O CO P_*roOOO0OO0O CS00 0O00 
1r roO' ON roroOO'ro NO00NO COO'ro 0OOo 
co 0O_ 4>
> O'OOO p ro CD ro0O'
“Tl 
n OOO O0O O0OO00O0O000OOo
3 OOOO0 OroOO0O 00O0O0OOOOOO0OO 
*< 
n
(CONTINUED) H _ C=1 Z r* 71 da cnO' O0 ro P roro fOcn
j: O O roro O' ro ONcn O'00 ropro 
m p 'Oro P
> ro 
73
cn 
2. 
A. 
71 
Co01 00 cnO' 00O 

cn ON
O 0NO NO roP 
00NO0 1—ro cn CD 
CO 'O-o
O roOO CONOCOp
O' O'O' -O 0000000000000<0'O0NOOO0
'O 0
O 
CO PP P PP■OP
COOJ CO CO CO CO Co CO COP ■0PPPPPp PCO COsO CO COCO COco COco Coco CO CO
TABLE 
P NO vO co0 ro 0COON CO oO (-• no on NO COON NO0 COro p
cn CO ■-O 00NO COcncn COcn CO ro CO CO 
ro 
ro cnCJO' COpO'00PO'ro0O0OOO ro CO0cjP0 > O OO0OOro NO 00pO Op P ro
O0 O' 
0 oO
3 OO0 O0 rOOO00OO0 OOOOOOOOO0O0OOO0OO0O0 00 OOO0 0 OO0
“* 
H 
>-< 
a 
z 
71 

►CO r-» >-• 
>— »-* *-•OP •— PO00 >-* ro OJ ON NO> 00r— 0 N3 CO -O v-O 
-0 NOon co NOO' Co COCO 0 O N-*—000 CO p CO 
COro ro pro ro ro ro
71 -O
> 
C/1 73 
O co 0ro p cn 'vJ 
co ON
cn NOO NO cnNO ro p CO CnCn 
-0 -o-0 CO CO CO CO NO<0 P P PoiCn CnC/l-n
0 CO CO CO CO00CD p
cncn C"»P ONO' ON ONON
cn OJ ro rororo roro rororo
ro COCO CO NO CO COco CO CO COco CO COco CO 
CM roro ro rororo rorororo 
(CONTINUED) 

2. 
A. 
TABLE 

OF 
7. 
SAMPLE 
OF 
7. 
SAMPLE 
OF 
7. 
SAMPLE 
TOTAT. 
FREQUENCY 

ZONE 
TOTAL 

FREQUENCY 

ZONE 
TOTAL 

FREQUENCY 

ZONE 
0.03 0.06 0.03 0.06 0.06 0.03 
1
12122 
715 717 718 719 722 733 
0.03 0.06 0.03 0.16 0.03 0.03 
12 51l
1 
704 705 706 709 710 712 
0.03 0.03 0.16 0.06 0.12 0.06 
11 
52A2 
669 677 700 701 702 703 
O''OoP' OJOJ 03orroOJOrOrvOO'ro0JroOJOJO'O'VOO'OJOJOrOJro03O'oo50O'vO00NJOJ
OJonO NJ O' VO PO vON) rO OJ 
X-	H-»—
00h—OO'O >-JonO 	oOO r—*rooooo roorooooooOoooo ro03ro00 OfOfOOO 
> OOr X.' ro 
OOoOO O OOooOooOOOooooOOooo OOOOOoo oOOOOOoOOo00o o o
T01 
t*i 
r* 
00vO 03OJvO
NJ >— 	OJ (ONJOJro ■o X> ■43 •o -O>-* ro 
O'P* OJ OJ -oOJP-Or rO03'OOJroro On■o
FREQUENCY 
ro 
CO NJ 
ro p-
ZONE 
BY 	r— co O OJonOroro Or•o00vOo ro P*onO'
ra roOJOnO'SINjOJP>onvQrooj03O' COO roOJP‘03O' VO 4> OJ 2 P*P*p'O'OrO'O'Or -'O'O ooooCo00030300CO0303O roOJ OJOJ4> P*4> oU3 03OnOn
o 	OJ 
ro njroro 	roro roro ro roro roro OJOJ OJOJOJOJOJOJ OJOJoj OJ OJoj
ro roronjro ro ro ro roro OJ 	ojOJO ro 
oo 
OJ O' OJ OjojO' OOJ vOOJVOVOOJ O'ro OJO'O'roO'OJoO'ro CO00roOrvOONOJojro
C O' OJoo vOor OJ ro OJ OJ OJ 
O oo
o oo ooooooo oooOOOOo»—OOo OOO ooo ro OO'oojO
> oo
EMPLOYEES N 
OH 
o 
o OooooooooooOOoOOOOOoOOOOOoOOOOoooOOOo OOOOOOOoo 
a 
Nf >-? 
DFW 
q 
ra 2
zn
r* 
o 
ro >-* 	r—OJ rorop* roroP* OJro4>vOvOO'roOrro
WORKER «< 
> C*1 roojro roP' OJ 00O' 
C/3 30 

20 
FIELD 
00 
ro h-* 	-J o O rooj4>On oroonO' 00 OJonO'-o ro OJ 03OJP-«vl03
c*i roO'r-O''OCO'Qo ro4>03O' vQ 00 vO 
OOOoO O H-4 ro rororo OJ
OJ P*on0/3onO'O'*o -O'OCOCOvO03VO'43ovOvO 0JojojOJ ro ro ro ro ro ro roro roro rorororo ro ro 
z 
roro 	ro 
NON-LOVE INO 
OJorVOooOJO' OJOJOjojOJonO'OJojOjOJVOOOJOJOJO'O'OJOJ03O'OJOJOJVOsoO'OJOrOJojOJOJOJojojOJO' 
O Oo 	K-
t—oooooooOoo ooro OoooooOo ooOooooo"4 Oooooooooo ro
> 
ooOoo o
ooO OooooooooooOoOOoo O OOoOooOoo ooo
P oOO OOOOOO 
JV? H 
3. 
A. 
*■< 
b2 
C*3 

oovn OJOJ 	fO on v0ro OJOJro 03 
ro OJ
TABLE o 
> $3 NJ 
01 

?3 
"*1 
r-o O' OJOn 03 
OO ro 	"O COoO >O'O'O' 00 

O 
On5355 5657 6061O'6566 72737479CO 82co038990vOvOvO969766 102o106 1151205 123125126127129130OJ 
-
ZONE 
S'0ON*0 Co CO NO 	u>nO CO COLOO' CO CO
r* COCDCO ON CO COmDON CO CD CO CoCOCOCOCO CO 
> 

OOO IN)0roOOOO0016,O0OOOOOo©oooOO©oooooo ©o 
g OOO0 0OOOOO00OO0OOOOOoooOoO©ooOoOoOOoo H 
*< 
cn z 
r* cn 

g 
3 
O 

CO 
cn -CO
> 
►— ro
cn rocoS' ON'O00NO cn0 00 ro00 o> On S' !S>0022 
O
Z 000300NONONONO 	oroCo0-300o
501 503 507 508 ** 513 NO 532 533 £ S' 596 o 709
522 s^ 
J 
N 
CO 	vOCO 
4r S' ONON££ NO00£NONOONCOCOCONNOCOrocoCNroO'ON COCCOCOCOcoCO CO'Orococo 
ro oooo 3o Ooo
►— ro00O00 O0000ooO 	o
O .16
'■n	> cn 
O Oo ooOooooOoOo3 OOO00OOO0©O00OO00000ooOo o 
>< 
cn 	Z
(CONTINUED) ,o H 
2 
n 
Crt Cn co 
> 
*n 
3. 
A. 
•o t— 	COo
h 	o> S'o~CO Cn S'cnON ro
CT 0 foCOS' ON00NO ro ro S' b0
b 	i 12) s^ 
S'S'S'S'S'S' S' S' S'S' S' S'S'S' S' S' S'0 S'S' S' S' S' S' S' S'S' S' S' S'S' S' S' S' 
CO
TABLE 
ro Cn.71 .28 ON COO' NOnOO' COro Cn CO03 CO ONco03noO'S'o COONco
onronoCJCNroO CO COON b 0 O O O CO O ooOS' ro o N3ocoCOoo
O CO15 	o 
> OcnO . Cn 
g 
OOOO0OOO OOO OOOOO OooOooooooooooOoo oo 
ro 
H 

cn 	z 
Cl
r* 
3 
2S'L0O COCO NOCO 00cnCO roS'S'00voroNO noO'N) O'oN)
3 ro •>0 ro 
fO
0 17 68 
03 	N»
> 	cn ro 
co 	73 
-n 

S' ro 
COMO S' 	§
ooo
CO nO
cncn On
ZONE 361 362 363 365 367 368 369 372 373 375 376 377 378 379 380 381 382 383 385 387 388 389 390 392 393 •o 395 396 408 
S' x>S'S'
CO CO
COCOCO 	CO 
APPENDIX B 
PROGRAM FMPPIST ( INPUT»OUTPUT»TAPE 1 = INPUT ) 
COMMON 

PF(24,9O),XMEAN,VAR,AC9),B(9),APRC24,6O),XLEA(2« 6O),GARC24
f 
*,60),GLE(?4,60),IWAYC99), I HOUR(99),MIN(99),EMP|,(99) ,XVAR(24,60) ,VL *E(2<1,60),GXVC24,60),GVL(24,60),JW(3) READ 1, NPFR,NDATA,INI 
PRINT l/N'PF. R»NDATA»INT 
1 FORMAT(6X,2J3,I2) 
PI s 4,O*ATAN(I.O) 
PRINT 19 

19 	FORMAT(IHI,/«9X,*W 0 R K SHIFT S*,//, 
* 
*,//, 
* 4 feX » *T I ME STARTING FNDING*,//) 
DO 77 J=I,NDATA 
READ 7, IWAYCJ), IHOURCJ),MINCJ),EMPLCJ) 
7 FORMAT CfcX,l 1,212,F5,0) 
IFCIWAY(J) .EG, 1) GO TO 101 
IFCIWAYCJ) ,EQ. 2) GO TO 102 
101 CONTINUE 
PRINT 15,1 HOUR(J),MINCJ),EMPL(J) 
15 FORMAT (46X,212,2X,F10,0) 
GO TO77 
102 CONTINUE 
PRINT 16,IHOUR(J),MIN(,J),EMPL(J) 
16 F0RMAT(46X,212,12X,E10,0) 
77 CONTINUE 

PRINT 29 
29 FORMATCI HI,/53X,*P £ R I 0 D S*,//, 

* 
31X,*B0RDERS *,//, 
* 29X,*L0WER UPPER MEAN VARIANCE DISTRIBUTION 
* 
WAY*,//) 
DO 36 M=1»1440 

GAR(M) = 0,0 
GLE(M) = 0.0 

=
GXV(M) 0,0 
6VL(M) = 0,0 
38 CONTINUE 

DO 11 Nsi,NPER 
READ 5, IHOURI,MINI,IHOUR2,MIN2,XMEAN,VAR,IDIST,IWI 
5 FORMATC6X,ai2,2FIO,R,2II) 
JfUl) 10H ARRIVING

= 
JW(2) = i3H LEAVING 
JK(3) 10HARR.-LEA,

= 
IFCIOIST ,EU, 1) GO TO 5010 
IFCIOIST .EG. ?) GO TO 5020 
IF (101 ST ,£Q. 3) GO TO 5030 
IFCIDTST .EG . 4) GO TO 10*10 

IFCIOIST .EG. 5) GO TO 1050 IFCIOIST .EG. 6) GO TO 6010 5010 CONTINUE PRINT 35,1 HOURI,MINI,IHOUR2,MIN2,XMEAN,VAR,JW(IWI) 35 F0RMAT(3PX,212,3X,212,2(2X,F10,4),7X,*N0RMAL*,5X,A10) CALL NORMAL(N) 
GO TO 1060 5020 
CONTINUE 
PRINT i»5,IHOUR1 MINIIHOUR2,MIN2,XMEAN,VAR,Jw(IWI)

,
, 
45 	F0RMAT(30X,212,3X,212,2(2X,F10.4),5X,*L0GN0RMAL*,4X,A10) CALL LOGNRML(N) GO 
TO 1060 5030 CONI INUE PRINT 65,1 HOURI,MINI,IHOUR2,MIN2,XMEAN,VAR,JW(I1)
W 
65 
F0RMAT(3QX,212,3X,21?,2(2X,F1f1.4),2X,*M£G, EXPONENT lAL*A 1«J) CALL NEGEXPCN) 
GO TO 1060 
1040 	CONTINUE 
PRINT 75,1 HOURIrMINI,IHOUR?,MIN2,XMEAN,V AR,JW(IW1) 
75 F0RMAT(30X,212,3X,212,2(2X,F10.4),7X,*GAMMA*,6X,A10) 
CALL GAMMA(N) 

GO TO 1060 
1050 CONTINUE 
PRINT 05, I HOUR 1,M1N1,1 HOUR2,MIN2,XMEAM,V AR,JW(IWI) 
05 F0RMAT(30X,212,3X,212,2(2X,F10.4),4X,*ERLANG DOWN*,3X,AIO) 
CALL ER|„ ANG(N,0.0 ) 
GO TO 1060 

6010 CONTINUE 
PRINT 95,IH0UR1,MINI,IH0UR2,MIM2,XMEAN,VAR,JW(IWl) 
95 F0RMAT(30X,212,3X,212,2(2X,F10.4),6X,*ERLANG UP*,4X,AIO) 
CALL ERLANG(N,I.O) 

1060 CONTINUE C DO 66 1=1,90 C PRINT 13,PE(N,L) c 13 format(2x,Eio.7) 
c 66 
CONTINUE 
TIMEI IHOURI * 100 + MINI

= 
=
TIME 2 IHOUR2MOO + MIN 	2 
DO 22 J=I,NDATA 
IF(IW1 ,F 0,3) GO TO 110 
JFCIWI ,NE. IWAY(J)) GO TO 22 

110 	CONTINUE 
TIME = IHOUR(J)*100 + MIN(J) 
IFCTIME TIMED GO TO 22

.EE. 
IFCTIME ,GT, TIME2) GO TO 22 
IFUWAY(J) .EO. 1) GO TO 1090 
IFCI WAY CJ) .EO. 2) GO TO 2010 

1090 	CONTINUE 
DO 33 K=l» 90 

IHT 	= IHOUR(J) 
MW = 
MIN(J) 
7777 

CONTINUE 
IF(K LE. MW) GO TO 9020 

, 
IF(IHT .EO. 0.0) IHT=2a.O 
IHTI IHT 1.0

=
IHT 	= IHTI 
= MW 	+60.0
NJ 
MW = 
MJ 
GO TO 

7777 
9020 CONTINUE 
MJr MW K +1.0

~ 
IHI = IHT +1.0 
ARR(IHI,MI) = EMPLCJ)*PE(N,K) 
GAR(IHI,MI) = GAR(IHI,MI) + ARRCTHI,MI) 
0 = 1.0 -P E(N,K) 
XVAR(IHI,MI) = Q* ARR CIHI,MI) 
GXV(IHI,MI) GXV(IHI,MI) + XVAR(IHI,MI)

= 
C PRINT 70,TH1,M1,ARR(IH1,M1),GAR(IH1,M1),XVAP(JH1,M1),GXV(IH1,MD 70 F0RMAT(9X,212,2(9X,F10.0),2F10.5) 33 
CONTINUE 
GO TO22 

2010 CON fINUE 
DO 44 K =1,90 
IHT I HOUR(J)

= 
MW = 	MIN(J) 
MJ=	K+ MW 
2222 	CONTINUE 
IF(MJ ,LT, 60.01 GO TO 9030 

IHTI =INT + 1 
IHT = IHTI 

IF(IHT .EQ. 2fl) IH T =0,0 
MI=MJ 60

-
=
MJ 	MI 
GO 	TO 2222 
9030 	CONTINUE 
=
IHI IHT+ 1 
Ml =	MJ+1 
XLEA(IHI,MI) EMPL(J)*PE(N,K)
-
=
GLE(IHI,MI) GI.E (IH I,M 1 ) + XLE A C IHI #MI ) 
= 
-
0 1.0 PE(N,K) 
VLE(IHI,MI) = O*XLEA(IHI,MI) 
GVL(IHI,MI) = GVL(IHI,M 1) + VLE(IHI,MI) C PRINT 1Ifl,IHJ ,M 1,XLEAUHI ,M 1>#GLE(IHI,MI),V|_E(IHI,MI),GVL(IHI,MI) 
114 F0RMAT(2X,212,4F10.0) 
44 CONTINUE 
2? CONTINUE 

11 	CONTINUE 
PRINT 39 

39 	FORMAT(IHI, VEHICLE VOLUME*,//, 
* 
42X,*A. R RI VI NG*,l6X,*L E A V I NG*,//, 

* 
?7X,*TIME *,2(3X,*MINIMUM*,SX,*MEAN*»4X,*MAXIMUM*),//, 

* 
25X, *INTERVAL*, 6(i)X,*VOLUME*),//) 


J1= 2.0 
J 3 = 60.0 

=
INTI INT 
DO 28 1=1,24 
GO TO97 
98 CONTINUE 

=
INTI INT 
J1= 	J2+ 1 
J 3 = 60.0 
97 CONTINUE 
DO 99 J=JI,J3,INTI 
IF(INT .NE. INTI) GO TO 3333 

=
CA Rl 0.0 
=
CAR 0.0 
CLEI 	= 0.0 
CLE = 0.0 
=
CV AR 0.0 
CVAP 1= 0.0 
CVLE 	= 0.0 
CVLEI 0.0
= 
3333 	CONTINUE 
J2 =J +INTI 1

-
NINT = J2 
IF(NI NT .GT. 60) NINT=6O 
DO 68 K=J,NINT 
CA R 1 = CAR + GAR(I,K) 

=
CAR CAR 	1 
CVARI = CVAR + G XV(I,K) 
CVAR = CVARI 
CLEI = CLE + GLE(I» K) 

=
CLE CLF. 1 
CVLEI = CVLE + GVL(T,K) 
=
CVLE CVLEI 
66 	CONTINUE 
IFCI ,NF« 2fi) 00 TO 17 
IF CNINT .LI. 60) GO TO 17 

=
CARI CAR + GAR(I ,1 ) 
CAR CAR!
-
CVARI CVAR + GXV(I, 1) 
CVAR 
= CVARI 
CLE) = CLF + GLE(1,1 ) 
CLE CLEI 
CVLEI CVLE + GVL(1»1)

= 
CVLE = CVLEI 
IH2 24

= 
M2s 00 
GO 

TO 	113 
17 	CONTINUE 
IF(J2 ,GT. 60) GO TO 99 
IH2 = 

-
I 1.0 
-
M2= NINT 1 
113 CONTINUE 
-
CAR 2 = CAR SORT(CVAR) 
CAR 3 = CAR + SORT(CVAR) 
CIE2 CLE SQRT(CVLE)

=
CLE 3 = CLE + SQRT(CVLE) PRINT 16,IH2,M2,CAR2,CAR,CAR3,CLE2,CLE,CLE3 16 F0RMAT(27X,212,6F10.0) 99 CONTINUE 
IF(J2 ,LE. 60) GO TO 96 
J3= J?-60 INTI =J 3 J1= 1 
28 	CONTINUE 
CALL EXIT 
end 
SUBROUTINE NORMAL(N) 
COMMON PE(24,90),XHEAN,VAR,A(9),R(9),ARR(26,60),XLEA(26,60),GAR(24 *,60),GLF(26,60),1WAY(99),1 HOUR(99),MIN(99),EMPL(99),XVAR(26,6O),VL *E(26,60),GXV(26,60),GVl(26,60), ,1W(3) 
1 FORMAT 4lX*>27 1 ;PEr>S SET TO ZER 0 * E 10,3) 
PI = 4,0* ATAN(I.K) 
CONST = 1.0/SQRT(2,O*PI*VAR) 

=
AREA 0.0 
=
ALAST 0.0 
=
FLAST O.P 
T= 
0.0 
XEXP = -o.s*<T-XMEAN)**2/VAR 

IF( AHS(XEXP) ,GT, 271,0 ) GO TO 2010 
FEAST = CONST*EXP(XEXP) 
DO 1020 1=1,90 
DO 1010 J=l ,10 
T= T 

+O.l 
XEXP = -O.5*(T-XMEAN)**2/VAR 
IF ( ARS(XEXP) .GT, 271,0 ) GO TO 2010 
F CONST *ExP(XEXP)

= 
AREA s AREA + O,OS*(FLAST+F) 
FLAST F

= 
CONTINUE 
PE(N,I) AREA ALAST 

1010 
=
=
ALAST AREA 
1020 	CONTINUE RETURN 
2010 CONTINUE PRINT 
I.XEXP 
DO 81 1=1,90 
P£(N,I) = 0.0 

81 	CONTINUE 
RETURN 

END 
SUBROUTINE LOGNRML(N) 
COMMON PE(24,9O),XMtAN,VAR,A(9),R(9),ARP(24,6O),XLEA(H4,6O),GAR(24 *,60),G1F(24,60!,1RAY(99),1 HOUR(99),MIX(99),tMPL(99),XvAR(2 4,60),VL *E(24,60),GXV(24,60),GVl (24,60) 
1 FORMAT!/, 41X*>271fPFr>S SET TO 7FRP*MO,S) 
A(2) =AI.OG(XMEAN)-O.b*ALOG( (VAR/(XMEAM**2))+I,O) 
B(2)=AIOG((VAR/(XMEAN**2)) + l,0) 
PI = 4,fI*ATANU,O) 
CONST=I,O/SQRT(? O*PI*B(2)>

t
ARE A= 
0,0 
AL AST=0 0 

f 
FUST = 0.0 
=
T 0,0 
DO 	1020 1=1,90 
DO 	1010 Jsl,lo 
T=T+o„l 
XEXP = -0,5*((AIOG(T)-A(2))**2)/R(2) 
2010

IF ( ABS(XEXP) ,GT 271,0 ) GO TO 
F CONST*EXP(XEXP)/T 

, 
= 
AREA 	= 
AREA t O.OS*(FIAST4F) 
FUST F

= 
1010 	CONTINUE PE CN,I ) a AREA -AIAST ALAST AREA
= 
1020 	CONTINUE RETURN 
2010 
CONTINUE 
PRINT I,XEXP 
DO 	61 1=1,90 
=
PE(N, I) 0,0 
81 	CONTINUE 
RETURN 
END 

SUBROUTINE NEGEXP(N) 
COMMON PE(24,90)f XMEAN,VAR,A(9)*B(9),ARR(24,6ft),XLEA(24,6(i),GAR(24 *,60),01E(29,60),IWAY(99) THOURC99),MIN(99)FMPL(99 ),XVAR(24,60),VL
,, 
*E(24,60),GXV(24,6U),GVL(24,60)  
1  FORMAT (/,41X*>271 >PEr>S  SET  TO  7ERO*FIO,3*  IN  NEGATIVE  *  
*  *EXPONENTIAL.  DISTRIBUTION*)  
A (3)sXME AN  
Tso,o  
ALAST=O.O  
DO  1010  1=1,90  
T=r+i .0  
XEXP  =  »T/A ( 3)  
IF (  ABS(XFXP)  t  GT,  271,0  )  GO  TO  2PIO  
AREA  =  1„0  -EXP(XEXP)  
PE(N,I)=AREA~ALAST  
alast=area  
1010  CONTINUE  
RETURN  
2010  CONTINUE  
PRINT  1 /XEXP  
DO  B 1 1=1,90  
PE(N,I)  =  0,0  
61  CONTINUE  
RETURN  
END  

subroutine gamma(N) 
COMMON PE(24,9O),XMEAN,VAR,A(9),H(9),ARR(24,6/1),X1.EA(24,6'1),GARC24 A»6O)#GLF(24#6O),IWAY(99),I HOUR(99),MIN(99),fcMPLC9 9),XV AR(24,6?),VL *E(24,60) ,GXV(2<I, 60) ,GVL (24,60) 
1 FORMAT(/,4IX*(GAMMA)»K GT 150 EXPECTED FREQUENCIES*
-
* * SET TO ZERO IN GAMMA DISTRI BUT 10N*G13,2) 2 FORMAT(/,3SX,"ATTEMPT TO RAISE EXP GT 271(1F,*F 1 0,3 *) GAMMA
IN * 
* 
A(5) = XMEAN/VAR 
B(5) XKEANAXMEAN/VAR

B 
IF( B(5) GT« 150,0 ) GO TO 2010 
, 
=
AREA 0,0 
-
ALAST 0,0 
FIAST=O,O 
T = 0,0 
CONS T s A(5)/GAMMAF (B(5)) 
DO 1020 1=1,90 

DO 1010 J=l, 1 0 
TsT+o,l 
XEXP = «A(5)*T 

IF( ABS(XEXP) ,GT, 271,0) GO TO 2020 
FsCONST*((A(S)*T)**(B(S)-I,O))*EXP(XEXP) 
AREA=AREA+B,OS*(FLAST+F)
FLASTsF 

CONTINUE
1010 
P£(N,I)=AREA-ALAST 
ALASTbAREA 1020 CONTINUE RETURN 2010 CONTINUE 
PRINT 1, B(5) 
DO 61 10t,90 
PE(N,I) 0,0

n 
81 CONTINUE return 2020 CONTINUE PRINT 2, XEXP 
DO 51 15 1,90 
PE CN,I) * 0,0 
31 CONTINUE 
RETURN 
End 

SUBROUTINE E R|_ ANG (N, XROUND) 
COMMON PE(2u,9O),XMfcAN,VAR,A(9),B(9),ARR(24,6 3),XLEA(2*1,60),GAR(24 *,60),G1E(2«,6&),IWAY(99),IHOUR(99), MIN(99),tMPL(99),XVAR(?u,60),VL *E(24,60 ) ,GXV(24,60),GVL(24,601 
-
1 FORMAT!/,35X*(ERl.ANG),« GT 150(*F10.3a) EXPECTED FRF.QUENC Ifc S a * * SET TO ZERO IN ERLANG DISTRIBUTION*) 2 FORMAT!/,9OX,*ATTEMPTED TO SET K EQUAL TO ZERO IN ERLANG* 
** 
DISTRIBUTION*) 3 FORMAT!/,aiX*>27l>PFrfS SET TO ZERO*FIO,3* IN ERLANG DISTRIBUTION*) NO IST a XROUND + 6,fc 
A(NDIST) = XMEAN/VAR 
IB=XMEAN**2/VAR+XROUND 
B(NDIST) IB

s 
IF! B(NDIST) ,GT, 150.0) GO TO 2010 
IF(IB,EQ, 0 ) GO TO2020 
AREA a 

0,0 
ALA3T=O,O 
FLASTsO,O 

T= 
0.0 
IBMIsIB-i 
IFACTsI 
DO 1010 lai.lßMl 
IFACTsIFACT*I 

1010 	CONTINUE CONST=A(NDIST)/IFACT 
DO 1030 1=1,90 
DO 1020 J=l,lo 
TaT+0 # 1 

XEXP a -ACNDIST)*T IF! ABS(XEXP) ,GT, 271,0 ) GO TO 2030 FeCONST*! (A(NDIST)*T)**IBMI)*EXP(XEXP) AREA=AREA+O,OS*(FLAST+F) FLASTaF 1020 CONTINUE PECN, ALASIaAREA ie3o CONTINUE 
RETURN 
2010 	CONTINUE PRINT 1, B(NDIST) DO 61 I=l*9o P£(N,I) 0,0
a 
61 CONTINUE RETURN 2020 CONTINUE 
PRINT 2 
DO 61 1=1,90 
PE(N,I) = 0.3 

61 CONTINUE RETURN 2030 CONTINUE PRINT 3, XEXP 
DO 51 1=1,90 
PE(N,I) = 0,0 
51 CONTINUE 
RETURN 

END 
FUNCTION GAMMAP(X) 
C W GAUTSCHI,ALGORT THM22I,CACM, H,WERNER,R.COLLINGE,MOCIS,I9b-)97,
f
DATA AOO/ 0,90999 99999 9999/ 
DATA AOl / 0.92278 93351 0233/ 
DATA A02/ 0,91189 03301 6678/ 
DATA AO3/ 0,08157 69261 29155/ 
DATA AU9/ 0,07929 89159 19999/ 

DATA AO5/~0,00026 61865 99953 06/ 
DATA AO6/ 0,01119 97193 35778/ 
DATA AO7/-0.00283 69625 20372 8/ 

DATA AOB/ 0,00206 10918 50225 5/ 
DATA AO9/-P,00083 75696 85135 17/ 
DATA AlO/ 0,00037 53650 52263 07/ 

DATA All/-0,00012 19173 98706 32/ 
DATA Al2/ 0,00002 79832 88993 83/ 
DATA Al3/-0,00000 30301 90810 28/ 

C ZsX 
IF(X GT,150,0) GO TO 91 
GAMMAF=I,O 
1F(Z,GT,3,0) GO 

e 
TO 20 
GO TO30 
IFCZ,GT O,O) GO 10 

1F(Z,GE,2.0) 
TO 
# 
Z=AMOD(X,I,) 
IF(Z,CU,O,O) 

GO TO90 
ZsX 
10 	gammaf=gammaf/z 
Z= Z+i .0 
Jr(Z«GE,2,O) 30,10 
20 ZsZ-1,0 
GAMMAF=GAMMAF*Z 
IF(Z GT.3,O) GO TO 20 

t 
30 
TsZ-2. P=l(((C(Al3*T+Al2)*T+All)*T+Alo)*T+Ao9)*T+AoB)*T+Ao7)*T+Ao6 Ps (((((P *T+AOS)*T+AO9)*T+AO3)*T+AO2)*T+AOI)*T+AOO 80 GAMMAF=GAMMAF*P RETURN 90 PRIM 95,X STOP 3 
91 	PRINT 96,X 
CALL SYSTEMP(O,O,O,O,O,O,I,6LGAMMAF) 

STOP 9 96 FORMAT(*OX = *E20.13* GT 150*) 
95 	FORMATC*OBAD ARGUMENT FOR GAMMAF(Z)*E2O,I3) 
END 

REFERENCES 
1. James
Bradley, V., Distribution—Free Statistical Tests, Prentice-Hall, Inc., Englewood Cliffs, 1968, pp. 296-303. 
2. 	Statistics: With View Toward
Breiman, Leo, a Applications, Houghton Mifflin Company, Boston, 1973, pp. 298-316. 
3. 
Chance, Merrit 0., Airport Access and Ground Traffic Study Review, 
Graduate Report, Institute of Transportation and Traffic Engineering, University of California, Berkeley, 1968, pp. 18-31. 
4. 	Dunlay, William J., et al., Survey of Ground Transportation 
Patterns of the Dallas/Fort Worth Research

Regional Airport, Report 15, Transportation Studies,
Council for Advanced The University of Texas at Austin, August 1975, pp. 13-23 
5. Gerlough, Daniel L. and Barnes, Frank C., Poisson and Other 
Distributions in Traffic, Eno Foundation for Transportation, Saugatuck, 1971, pp. 35-49. 
6. and 	and Other Distributions
Gerlough, Daniel, Barnes, Frank, Poisson 
in Foundation

Traffic, Eno for Transportation, Saugatuck, 1971, 120-121.
pp. 
7. 
MAPSCO, Dallas 1975, Handy Map, Inc., Dallas, Texas 

8. 
MAPSCO, Fort Worth 1975, Handy Map, Inc., Dallas, Texas. 

9. 	
Naylor, Thomas H., et al., Computer Simulation Techniques, John 
77-701.



Wiley & Sons, Inc., New York, 1966, pp. 
10. 	Statistical Package for the Social
Nie, 	Norman, et al., Sciences, McGraw-Hill Book Company, 
1975. 
11. North Central Texas Council of Governments, 	data furnished from 
their in-house computer operations. 
A 	of Local
12. Robinson, John P., and Nordlie, Peter G., Survey Origins and Destinations of Users of Washington National Airport, U.S. HSR-RR-61/S—MS-b, February 1961,
of 
37-43. 

Department Commerce, 
pp. 
Statistics for the Behavioral Sciences,
13. 	Siegel, Sidney, Nonparametric McGraw-Hill Book Company, 1956, pp. 47-52. 
100 
The vita has been removed from the digitized version of this document.