TRANSPORTATION
A SUBSTUDY OF THE REGIONAL OFFICIAL PLAN
CA 30K Hw GCC 7 rf eT
scoreny, UMN
HAMILTON PUBLI \
I} ny
es WW,
ti
REPORT NUMBER 3
DEVELOPMENT @F A TRAVEL FORECASTING MODEL THE PLANNING AND DEVELOPMENT DEPARTMENT OF
THE REGIONAL MUNICIPALITY OF HAMILTON-WENTW ORTH OCT.’76
Digitized by the Internet Archive in 2023 with funding from Hamilton Public Library
https://archive.org/details/transportationsuOOunse_ 2
THE REGIONAL MUNICIPALITY OF
HAMILTON WENTWORTH
October 25th, 1976
MEMORANDUM NO. 423
ioye The Chairman and Members Regional Planning and Development Committee
Supqject: hegilonal Official Plan Study = PENSE Forecasting
Background
One Of the major tasks of the Regional Official Plan Study is to determine the location and type (road/transit) of transportation facilities required for each alternative
development plan. lo accomplish this: task, future travel demands associated with each alternative development plan have to be forecasted, The attached report describes the
computerized travel forecasting model developed by the Resional Plan Division for this purpose.
In addition to Lts usefulness in the Regional Official Flan Study, the model provides the Region with a valuable Scontipuing planning tool for determining the ampact ox future major development proposals on the Regional Trans portation system,
Since the model enables forecasts of future travel to be made quickly and accurately for any area of the Region, such travel forecasts can be made available for use by the area municipalities on requesc.
Recommendation
hg Copies of the report to be provided to the area municipalities for information.
2s The Regional Plan Division make travel forecasts available for use by the area municipalities on request,
Respectfully submitted,
ya Luk
Db. A. Lychak, M.C.1i.?P. Commissioner of Planning and Development,
GMMcC:KP
Planning and Development Department. 100 Main St., East, Hamilton, Ont. L8N 1G8
Chapter
TABLE OF CONTENTS
Title
Table of Contents bast Gl Tables List of Figures
Summary Introduction
Development of Base Year (1971) Data
3.1 The Study Area 3.1.1 Zone System 3.1.2 Road Network 3.1.3 Screenline Selection 3.2 Internal Travel 3.2.1 Home Based Work Trips 3.2,2 Other Home Based Trips 3.2.3 Non Home Based Trips 3 External Travel .4 Land Use Data 5 Ground Counts
Model Calibration 4.1 Calibration Process 4.2 Results
Model Validation 5.1 Validation Process 5.2 Results
Appendices
A. Simulated Travel Time Contours - from Downtown Stoney Creek & Dundas
B. Screenline Documentation
C. HBO Trip Attraction Equation
D. External Travel Adjustments
E Land Use Data Bank (1971)
Lt
LIST OF TABLES
TELE
Home Based Work Trip Rates
External Travel Proportions
Other Home Based Trip Rates
External Travel by Corridor
Screenline (Ground) Counts
Calibrated Travel Time Exponents
Simulated versus Observed Corridor Volumes
Page
HRS
16
i,
18
22
24
29
Lid
Llot OF FIGURES
rae tS es
Model Calibration & Validation
Internal Traffic Zones
External Traffic Zones
Simulated Travel Time Contours (from Downtown Hamilton)
Simulated versus Observed TLD (HBW)
Simulated versus Observed TLD (HBO)
10
Lek
cs
Zo
26
1.0 SUMMARY
A Eo teerstaeea travel forecasting model was required in order to assist in the development of transportation - related impacts of future growth options for the Regional Municipality of Hamilton-Wentworth. This report describes the complete development of such a simulation model including,
(1) documentation of the input data requirements,
(2) calibration - refinement of the model parameters, and
(3) validation - comparison of the model's simulated results with observed data.
Figure 1-1 illustrates the sequence of activities which are involved in the model calibration-validation procedure. A series of iterative adjustments is made to the parameters of the model. The magnitude of each adjustment is gauged by comparing simulated results against actual observed travel data. This process is continued until an acceptable agreement is reached (in a base or present year) at which time the model is considered to be repre- sentative of the actual travel characteristics of the population in the study area and is ready for use as a forecasting tool.
A comprehensive traffic prediction package, PMPMOD (a version modified by M.T.C. staff to allow select link analyses and input of external trip tables) was obtained from the Ministry of Transport- ation and Communications. In order to provide relevant input for planning future transportation facility requirements it was decided
to model all person trip travel (excluding commercial vehicles)
Seeuring an the P.M. peak hour (4:30-5:30 P.M.).
FIGURE 1-1 MODEL CALIBRATION & VALIDATION
TRAVEL CHARACTERISTICS SURVEY
TRIP PRODUC. & ATTRAC. RATES
BASE YEAR (197]) ROAD NETWORK LAND USE
ZONAL PRODUC. & ATTRACTIONS
ZONE TO ZONE
TRAVEL TIMES
RIP DISTRIBUTION MODEL
OBSERVED Tele Dees
ADJUST TRAVEL SIMULATED TIME EXPONENTS Delieke Ss
ne CALIBRATION
TRIP ASSIGNMENT
MODEL
CHECK ASSUMPTIONS & REITERATE FROM POINT OF ADJUSTMENT
SIM. SCREENLINE CROSSINGS
OBSERVED GROUND COUNTS
NO VALIDATION
The study area, which consists of the Regional Municipality of Hamilton-Wentworth and the City of Burlington, was divided into 148 traffic zones. These became the basic unit for COlMlLECELCH
pr the required land use data, that is, employment (by type)
’
area and population.
Travel within the study area (internal) was simulated using three trip types; home based work (HBW), other home based (HBO), and non-home based (NHB). The two classifications of home based trips, which together accounted for approximately 87% of total P.M. peak hour internal travel, were generated on a zonal basis and distributed using a gravity model concept. NHB trips were simu- lated using a corridor factoring technique based on observed corridor travel characteristics.
Estimates of 1971 P.M. peak hour external travel ie. (trips whose origin and/or destination are outside of the study area) were obtained from the Ministry of Transportation and Communications in the form of an origin-destination (O-D) table. Several modi- fications were made to this data on a corridor movement basis by comparing assigned volumes to base year ground counts. The only corridor requiring major adjustment (assigned volumes reduced)
was the Brantford-Hwy 403-Toronto Corridor.
meeravel characteristics survey* of area residents, conducted in the Fall of 1974,provided much of the background information
required for development and eventual calibration of
a alt i ne a me eee me ree ee eee Meme atone * Travel Characteristics Survey, Regional Municipality of Hamilton- iravel Cnaracteristics ois,
Wentworth: Transportation, Report Number 2, October 1975.
the travel forecasting model. In particular, the following base year data were extracted from survey results. (for each trip purpose) :
(1) trip generation/attraction rates,
(2) “directional split: of travel,
(3) temporal variations in demands,
(4) proportional importance of trip purpose by corridor,
(5) mean and. distribution of trip lengths and,
(6) pattern of trip movements on an O-D basis.
Calibration of the model was accomplished by varying the travel time exponents in the friction function of the trip distribution phase until an acceptable agreement was reached between the shape and mean of the simulated versus observed trip length frequency distributions.
The final calibrated exponents and resultant simulated mean trip lengths (MTL) were 0.125/0.205 and 14.1/11.6 (min.) for the trip purposes of HBW/HBO respectively. Differences between the simulated versus observed MTL's were well within the +3% standard and the shapes of the trip length distributions closely matched each other when compared visually (see Figures 4-1 and 4-2).
Validation of the model's simulation results was achieved primarily through comparison of total person trip travel assigned to various corridors. In total, simulated estimates of P.M. peak hour person trip movements were compared to observed ground count data across twenty-five(25) different corridors or screenlines
which represented all major intra- and inter- Regional travel
movements,
Overall, the simulated results compared favourably with the corresponding ground count data (see Table 5-1). Most of the corridor differences were within +15% and more importantly, the magnitude of deviations were generally less than 1,000 peak direction person trips.
In general, simulated results tended to be higher than the ground counts by approximately 10%. This was considered to be an appropriate adjustment for more meaningful comparison purposes due to the following:
(1) ground (observed) counts were not available for minor roads crossing the screenlines
(2) increase in the activity rate ie. (effective work trip generation rate) between the base year (1971) and the year in which the trip rates were calculated (1974-5).
(3) absence of simulation estimates for commercial vehicle movements.
On the basis of validation checks such as the above it was concluded that base year (1971) travel patterns had been adequately simulated and that the model was thus capable of producing reasonable
forecasts of future P.M. peak hour corridor travel movements.
2.0 INTRODUCTION
AS part of the transportation - related input to Hamilton- Wentworth's Regional Official Plan it was necessary to estimate the future transportation System facility requirements and resultant impacts of alternative growth strategies for the Region. To facilitate the above and provide input for other transporta- tion system planning activities it was decided to implement a
computer-based transportation (travel) forecasting model.
A computer package, PMPMOD (a version modified by M.T.c. stafé to allow select link analyses and input of external trip tables) was obtained from the Ministry of Transportation and Communications and utilized to simulate total person trip travel occurring in the P.M.
peak hour (4:30-5:30 P.M.).
Before travel can be simulated with any degree of accuracy it is necessary to adjust the parameters of the model to reflect the trip making characteristics of the people within the study area. This process of fine-tuning of the model, commonly referred to as model CALIBRATION and VALIDATION, is carried out in a base or present year so that simulated results (at each iteration) can be compared to an already documented situation.
Once the model has been calibrated and validated in the base year, that is the model adequately simulates base year travel patterns, future land use scenarios can be input and thus future travel demand forecasts can be made.
One of the most important sources of information in the development of the simulation model was a travel characteristics survey* conducted by the Regional Plan Division in the fall of LOT Be
ee Ln eee
* Travel Characteristics Survey, Regional Municipality of Hamilton- Wentworth, Planning and Development Department, OC te oe oe mene erons
This survey was a one percent, random sample of all households within the study area (Regional Municipality of Hamilton- Wentworth and the City of Burlington) and provided base year information on trip making characteristics such as trip genera- tion» rates, distribution of trip lengths and directional spi1c o© “travel.
Figure 1-1 is a flow diagram of the complete model calibra- tion - validation process and related inputs which will be further
documented in the body of this report.
3.0 DEVELOPMENT OF BASE YEAR (1971) DATA
This chapter documents the sources and derivation of the major input requirements for the development (calibration and validation) of a travel forecasting simulation model.
3.1 The Study Area
The internal study area, within which a detailed travel simulation was carried out, incorporates the entire Hamilton- Wentworth Region and the City of Burlington. The City of Burlington was included due to the significant inter-action between it and the City of Hamilton.
Trips to/from places outside this area are relatively few in number and were handled separately under the classification of external travel.
3.1.1 Zone System
The zone system that was employed is illustrated in Figures 3-1 and 3-2. There are 174 traffic zones in all; 148 internal zones ie. (within the internal study area) and 26 larger external zones covering the rest of the Province of Ontario.
Of the internal zones, 135 are located in the Hamilton- Wentworth Region and are geographically compatible with both the T.A.R.M.S. system and the one used to code the travel character- istics survey. The remaining 13 internal zones cover the City of Burlington and are a direct aggregation of the T.A.R.M.S./ Survey zones.
For ease of comparison and presentation a superzone system
was formed (within the Region) consisting of 30 districts.
——S oR Yop _ i ——_—$_{=$SSOSQSx= —_= —_ SS ———————EE EE Eee ee
— 7 oe ameer == SESE =e HIZOMINIM- NOLIWYH 3U ASIVdIUINGN FYNUIUse are = 2
ray ane
a Mor
iA mOD
4
eo : J ». | _\ezo go
GLO\_U 52 ae 4
pe Ey SE =
S3NOZ DdIiddVUL TVNUFLNI [-¢€ ainbiy
| 9143 ayD7 : I -SSNOZ DIiddVYL IWNUSLXS Z7-€ aunbiy oa \ > 6SI Ne pee a as ee aN SN an ee ee, i \ | Sef 9) cn o9!1\— jal oer a ¢9l ; eo ai ey eee 691 291 an ae ae N77 es ; Y, ~ ms. 7 , A ys ec eon a 4 | iy L \ ell \ 6bI * nT os “ s aoe Nes oes. ‘a SS uoJnY @yD \ pli \ \ \
These superzones are direct aggregations of the traffic zones and correspond to the Region's Planning Districts.
3.1.2 Road Network
A road network consisting of approximately 649 nodes was utilized to simulate the transportation network. in the P.M. peak hour. All highways, arterials and a number of collector roads are represented in the network. Speeds were coded which reflect actual driving conditions in the P.M. peak hour. Much of the road link information was extracted from the report, Transportation Inventory, Regional Municipality of Hamilton-Wentworth Planning and Development Department,
Octeber 1975. The consistency of the network was initially checked by comp-
aring simulated network travel times with actual or known driving conditions in the P.M. peak hour. Travel time contour maps, which show the travel time (in minutes) required to reach all points in the study area from selected key origins, were used to display simulated conditions. Figure 3-3 is one such example showing the contours resulting when downtown Hamilton is used as the focal point. Other contour maps from downtown Stoney Creek and Dundas can be found in the Appendix. 3.1.3 Screenline Selection screenlines or corridors were selected as a basis for summarizing and comparing travel demands. Twenty-five (25) screenlines in all were utilized, eight encircling the Region and seventeen intersecting various movements within the Region
(see Apvendix).
(s#uauasduI aynulWw ¢ ul) SINOJUOD OUI} JIADIL
pues “ureces MOL mw ANIJWNLYVdIO LNINdOIIAIG ONY ONINNW Id
uoBuying jo Api oy, puy HL8OMINIM- NOLIWYH 40 ALITYdIDINAW TWNOIOIN FHL
(YNOH NV3d “Wd ) NOLIWVH NMOLNMOG WOdd SYNOLNOD AWIL IAAVAL GALVINWIS
on
a ee
€ a1nbig-
{
ee EL
Screenlines located in the City of Hamilton closely approximate those used in the Hamilton Transportation Strategy Study of 1973 so that corridor information could be collected and compared on a consistent basis.
Where possible screenlines were located so as to follow major physical and/or jurisdictional boundaries and intersect distinct corridor travel movements. A complete description of each screenline along with a listing of all major and minor roads crossing it can be found in the Appendix.
3.2 Internal Travel
Internal trips are defined as those which both start and end within the study area. Three trip purposes were used to simulate internal travel; namely home based work (HBW), other home based (HBO) and non-home based (NHB). As mentioned previously, a detailed discussion of the derivation of internal travel characteristics eg. (trip rates) can be found in the report, Travel Characteristics Survey - Reg. Mun. of Hamilton-Wentworth, October 1975.
3.2.1 Home Based Work Trips
In the P.M. peak hour home based work trips account for only about 51% of total internal travel. This survey result is Significant due to the relatively low observed proportion of "essential' trips ie. (work related) in the P.M. peak hour, a
characteristic which is significantly different from the overall
average used £or the T.A.R M.S) area.
ne
- 15 -
Analysis of the directional split of travel indicated that 94% of all home based work trips (in the P.M. peak hour) were destined for the residential end.
The final HBW trip rates which were used to simulate P.M. peak hour travel are shown below in Table 3-1. The trip
production and attraction rates are disaggregated by population
and employment density categories respectively.
TABLE 3-1 P.M. PEAK HOUR HOME BASED WORK TRIP RATES Trips Produced = Rate * Population
Population Density (persons/acre) Rate
0-10.0 .104 10.1-20.0 mela h 20. 1-30;,0 Cikey 30.1-40.0 Reals > 40.0 ~L65 Trips Attracted = Rate (1) * Wholesale & Manfuacturing Emp. + Rate (2) * Retail Employment + Rate (3) * Service Emp. + Rate (4) * Other Emp. Employment Density Wholesale/Man. Retail Service Other (employees/acre) Rate (1) Rate(2) Rate(3) Rate (4) 0-4.0 330 2309 <oU9 WE) 4.1-10.0 sone 308 308 -oue LO, =50.0 20" ae dy & 314 sou 22040 343 ; ,2L9 405 ~405
There was one area of exception to the above trip rates and that was for the City of Burlington. Analysis of existing
travel data indicated that the proportion of external trips
(as compared to total travel) generated by the City of Burlington
was significantly higher than the average for the whole study
area (due to Toronto bound commutertrips). Since the trip rates are representative of internal travel only this difference made it necessary to reduce the internally generated travel (trip rates) for the City of Burlington accordingly.
Using the results of Table 3-2 (below) a correction factor of 0.82 was calculated and applied to the HBW trip production
rates for all zones located in the City of Burlington.
TABLE 3-2
PROPORTIONS OF EXTERNAL TRAVEL
Total Study Area* (inelL. *Cityeol Buri.)
City of Burlington=*
%$ external
work trips 23% OF total source: * Travel Characteristics Surveyors
** 1971 Census - Place of Work - Place of
Residence and Journey to Work Survey- Burlington, 1972 3.2.2 Other Home Based Trips Home based trips for purposes other than work (HBO) account for approximately 36% of all internal travel in the P.M. peak hour. The purposes included (in relative order of importance) are shopping and personal business, social - recreation, and school. Table 3*3 lists the P.M. peak hotir HBO trip, cates 19 ne trip production rates are disaggregated by population density
Similar to those for home based work.
Scie:
The travel characteristics survey was not originally designed to collect the level of detail of information required to generate home based other trip attraction rates. An attraction equation was synthesized, however, by utilizing information obtained from the survey on the proportional attractivity of various land use types to home based other trips. By relating the intensity of each land use type to the corresponding amount of employment or population an equation was developed in terms of variables which are easier to forecast Appendix for derivation of HBO trip attraction equation).
TABLE 3-3 Trips Produced = Rate * Population
Population Density (persons/acre) Rate
O-10.0 20923 tO e200 .O08l2Z AOL = Uit0 ; s0596 30.1-40.0 ~0525
> 40.0 s0asl
.0103 * Manufacturing & Wholesale Emp. .285 * Service Emp. .285 * Retail Emp. 1.719 * Other Emp. .0224 * Population
Trips Attracted
++t+il
3.2.3 Non Home Based Trips Non home based (NHB) trips, which neither start nor end at a persons place of residence, account for only about 13%
of total internal travel in the P.M. peak hour. Insufficient data was available from the travel character=-
istics survey to produce trip generation rates for NHB trips.
As a result a factoring technique was utilized to generate trips
of this type. The complete survey sample was assigned to the
{see
composite road network and summarized by travel corridor (screenline) under two categories; home based trips and non- home based trips. The observed proportions of total trips to home based trips (across each corridor) were then utilized as factors to scale up simulated home based trips (HBW plus HBO) to represent total internal travel.
A complete list of the NHB scaling factors can be found ine lable o=i-
3.3 External Travel
External trips are defined as those having one or more of their trip ends outside the study area. In order to incorporate trips of this type an O-D (origin - destination) table was obtained from the Ministry of Transportation and Communications which relected P.M. peak hour person trip travel external to the study area. Subsequent assignment of this trip table to the network resulted in the corridor volumes shown in Table 3-4*.
These external trips can then be directly added (on a corridor basis) to the internally generated trips to produce estimates of total person trip travel demand.
TABLE 3-4
EXTERNAL CORRIDOR TRAVEL (in P.M. peak hour, pk. direction person trips)
Screenline (Corridor) 1971 External Trips 1 2,080 2 520 ig 520 4 27:0 5 700
; Several modifications (on a corridor basis) Were made to the external trips as shown here. These are documented in the Appendix.
Screenline (Corridor) 1971 External Trips 6 400 7 700 8 1,960 9 40
10 850 aa 460 2 210 13 270 14 190 5 930 16 90 17 200 18 1,890 19 5 ations 20 tO O 21 500 PME) 360 53 480 24 2,080 25 850
3.4 Land Use
Specifically, the land use requirements for operation of the model are: population and employment by type (wholesale and manufacturing, retail, service and other) for each traffic zone. The area of each zone is also required for calculating density-related measures, eg. (see Table 3-3).
The source of the 1971 data base was a publication of the Ministry of Transportation and Communications, Socio-Economic Data, 1971 (TARMS). Minor modifications were made to the above on the basis of census data for the area. A complete LiESseing of the data base can be found in the Appendix.
oS. Ground. Counts
Representative base year (1971) ground counts in the P.M. peak hour were required for all major road facilities (at peint
of crossing screenline) for comparison and eventual refinement
of model simulation results. The major difficulties encountered
in collecting this data were centred around,
(1) the multiplicity of data sources required to obtain all the necessary counts and,
(2) the lack of specific information on peak hour
travel as opposed to daily (A.A.D.T.), 24 hour volumes.
For facilities within the City of Hamilton the City Traffic Department proved to be an excellent source of ground count data. They provided traffic volume overlay maps for both the P.M. peak hour and the 24 hour period and vehicle classification counts by time of day.
Outside the City of Hamilton little information was available on peak hour traffic volumes. Instead,24 hour A.A.D.T. (average annual daily traffic) volumes had to be used and multiplied by a series of factors to produce estimates in terms of P.M. peak Hour, peak direction person trips. AsA.D.0. volutes for county roads were obtained from the Regional Engineering Department's records while similar data for Provincial Highways came from the Ministry of Transportation and Communications.
The transformation of 24 Hour A.A.D.T.. counts into equlv.-— lent volumes expressed in P.M. peak hour, peak direction person trips involved the estimation of the following parameters; vehicle occupancy factors, peak hour factors, directional split factors and truck factors. In the City of Hamilton most of this information was readily available. For roads outside the
C44
City the factors were developed from examination of City of
Hamilton data in the same cooridor, specially-requested Provincial counts on major facilities and design hour volume estimates.
Since the Hamilton Transportation Strategy Study (1973) had already documented 1971 P.M. peak hour, peak direction person trip movements in certain common corridors within the City of Hamilton it was decided to utilize this data where available. However, peak hour volumes were re-computed from A.A.D.T. data (for these corridors) as a check on the consistency of estimation of the associated transformation (peak hour) factors .
Wapleng—5 Lists the A.A.D.T. counts for each screéentine and the related factors that were applied to produce corridor volumes in terms of P.M. peak hour, peak direction person trips. A comparison with peak hour data from the H.T.S.S. is also given
where available.
sjzunos puno1iy SS.LH
uoT}OeATqC Yyeod ul sdtay, uoszeg InOH Need “Wd
IO49e A Ont,
ao0j70e7 ATT Tds Teuotz0eITG
(euTTUseADS Aq)
(“ay pz/ayu *Xd)
TORR FT InNOH yeed
SZNNOO CGCNNOYD
IONOR TA Aouednooo STOTYUSA
CLIT 8e VLO'PP GLE‘9Z T06‘66 EZL‘ 82
Opp’ TOT 067‘ 8hT
OT8‘L6 O6P‘LT QLE‘TS 7LO‘E6 006‘TT Oré6 ‘PT O62‘ET OPs‘LT CEL‘ TZ 0z0’9
0S9‘S9 OT6‘6T OTE ‘6
0€6’2T OT2‘8T OLE “eT OSP‘8
EPG’€E
lLANMAMNOR AD
G=€ SIEVG
4.0 MODEL CALIBRATION
£51. Calibration. Process
Model calibration, in the context of this report, refers to the process of iterative adjustments that are made to the friction functions in the gravity model of the trip distribution phase. The nature and magnitude of the adjustments required are gauged by comparing simulated results with observed data at each stage. Once acceptable agreement has been reached the model is considered to be calibrated,in that the final friction functions thus obtained are considered to be indicative of the actual travel characteristics of the population in the study area.
Generally speaking, as the travel time between two zones increases the likelihood or amount of travel between those zones decreases. This relationship is expressed in the PMPMOD computer package in the form,
- (6 * ttij) F (dij) «e
where F (dij) = represents the friction function 9 = a Calibrated constant. oom ag = the minimum path travel time
between zones i and j.
It is the exponential constant, theta (9), which is adjusted at each iteration of the calibration process. Theta is a reflection of the sensitivity of the study area population to changes in travel time, where larger values reflect greater sensitivity to travel distance thus resulting in proportionally fewer Long trips.
Comparisons between the trip length freguency distributivons “simulated by the model and those calculated from observed data
provided the basis for adjustment of the Eriction function
exponents (@). The following standards were used to define
an acceptable level of agreement between the two (thus signifying
a calibrated relationship): (1) the shape and position of the trip length distribution curves should be relatively close when compared visually and,
(2) the difference between simulated and observed mean trip lengths should be within + 38.
4,2 Results The final (calibrated) friction function exponents which
were selected are shown below in Table 4-1. One exponent was
required for each of the trip purposes distributed by the gravity
model. Comparison of the observed versus simulated mean trip lengths (MTL) indicates that their difference is well within
the + 3% standard.
TABLE 4-1 CALIBRATED EXPONENTS
Simulated MTL
Friction Function Exponent
Observed MTL
Trip Purpose
HBW 0.225
HBO 0.205 Figures 4-1 and 4-2 are a comparison of the observed
versus simulated trip length freguency distributions for the
trip purposes of HBW and HBO respectively. AS can be seen from
the graphs the agreement in the shape and position of the two
Ss Of curves is’ quite cood.,
("NIW) AWIL THAVaL
09 OS OV O€
02 OT 0
ci G OT ST "NIW T°PL = HDONAT dIuL NVEW daLVINWIS “NIW Z°PT = HLONAT diud NVYAW ATAUNS 0Z GZ (MaH) SNOLLNAIULSIG AONANODHA HLONAT diva Oe
GaAuaSdO SNSYHA GALVTAWIS T-? gdenola
THAWUL JO LNadoOddd
(*NIW) AWLL THAWILL 09 0S Or O€ 02 in 0
eae |
“NIW 9° TT = HLONAT dIdL NVAW GaLVINWIS ~ aa.
“NIW G°TT = HLONYT dIdd NVAYW AdANNS
(OdH) SNOLLOGINLSIG AONTNOTUA HLONAT dius GHAUASHO SOSUBA GALWINWIS @-p BAUNDIA
THAVEL JO LNaoddd
uy
N
5.0 MODEL VALIDATION
5.1 Validation Process
The model validation phase is a final series of checks that are made on the simulated, base year results by comparing the model outputs with observed * data from other sources. Once the base year conditions are modeled satisfactorily then it only remains to account for any changes in the input variables eg. (population and employment growth) in order to use the model for forecasting purposes.
The potential sources of data available for validation (cross-checking) purposes included the following:
(1) Census. - Place of Work - Place of Residence Matrix, 1971
(2) Travel Characteristics Survey - Trip Summaries, 1974
(op6 Venivcie (Ground) Counts,. 1971
No meaningful comparisons could be made between the Census - Place of Work data and simulated results because of the uncertainties involved in converting to a common unit of measure. Expressing both in terms of P.M. peak hour person trips, for example, would involve the estimation of trip rates, peak hour factors and directional split of travel on an origin - destination (interchange) basis.
P.M. peak hour trip data from the Travel Characteristics Seer was summarized in an aggregated O-D table format where each zone represented one area municipality. Unfortunately, the ee eyes k Yr Sis) eS eee
* by observed is meant non-simulated
small survey sample size and the need to extract only PM:
peak hour data left most cells in the table With an insignt ficant number of entries. Also, there was a time difference
of approximately 3 years between the two sets of data. Therefore the survey results could only be utilized to provide a gross check on the trip distribution phase for the major, home based inter-municipal movements.
The base year ground counts (as documented in section 3.5) provided the easiest and most direct comparison for simulated versus observed results. Agreement at this level of detail provides a positive check on all aspects of simulation process from trip generation through to trip assignment.
5.2 Results
Table 5-1 compares the total simulated demand with the observed ground count (by corridor) for each of the twenty-five screenlines. The simulated demands are further disaggregated into internal and external type trips due to their different methods of derivation.
Examination of the observed versus simulated volumes indicates that the agreement overall is good. Most of the differences are within + 15% and the magnitudes of deviation are generally less than 1,000 peak direction person trips (aggregated on a corridor basis).
In general the total simulated corridor volumes are
higher than the corresponding ground counts by approximately 103%.
6 8 E 9 S 4 3 c T
puewsd pueuwsd puewsgd
puewedg ZOROeT oseg sdtig pearesqo TeRIOL peyeTnuts Teo, TeurNA XY
TeureAuL TeIOL awWOH UON peseg owoy aUuUT[USETOS
SLINSHY NOILVAIIVA ; T-S ATEVL
This was considered to be an appropriate adjustment for more meaningful comparison purposes due to the following:
(1) ground (observed) counts were not available for minor roads crossing the screenlines
(2) incréase in the activity rate ie. (eiiective work trip rate) between the base year (1971) and the year in which the trip rates were calculated (1974).
(3) absence of simulation estimates for commercial vehicle movements.
A limited comparison between simulated results and factored-up survey data showed reasonable agreement in the amount of home based travel distributed between and within the various area municipalities. Where a large number of survey observations was available, such as travel within the City of Hamilton, the simulated number of trips very closely approximated those of the survey ie. (less than 5% difference).
On the basis of validation checks such as described above it was concluded that the model had adequately simulated base
year conditions and was ready for use as a forecasting tool.
APPENDIX A
Simulated Travel Time Contour Maps From
Downtown Stoney Creek & Dundas
( sjuawesdul aynuiw ¢ ul)
oe at ee r " wee
‘ (ee
HJYOMINIM - NOLUWVH JO ALITYdIDINQW TWNOID38 3HL
SINOJUOD BWI} JAADAL
(YNOH AW4d ‘W d) WAIYD AANOLS NMOLNMOQ WOUd SYNOLNOD JWIL WAVaL G3aLV INWIS
uojBuiyying yo Ayn eq, puy HIBOMINIM - NOLIWYH JO ALITYdIDINNW TYNOIOIY IHL
2 -.|
= 1
agent Vane |
(syuawaisul aynuiw ¢ ul ) SANOJUOD BWI} JBADIL
(YNOH vad ‘Wd ) SVGNNG NMOLNMOG
APPENDIX B
Screenline Documentation
ejerg Of 4!
SNOHV3DO01 ANITINASYDS
= 37 =
SCREEN LINES
SL-1
This screenline extends along the eastern boundary of the study area (below the escarpment), between Lake Ontario and the Niagara Bccar pment...) .1C. ineersects the Hamilton = St. Catharines corridor.
Major roads crossing are: Q.E.W. Highway #8
Minor roads crossing are: Base Line Road North & South Service Roads
SL-2
This screenline extends along the eastern boundary of the Hamilton-wWentworth Region (above the escarpment), between the escarpment and the South-East corner of the Regional Boundary.
Major roads crossing are: Highway #20 Regional Road 25A Regional Road 11 Regional Road 50 Regional Road 22
Minor roads crossing are: Smith Road
Highland Road
9th Road East
8th Road East
7th Road East
Golf Club Road
2nd Concession Road 7th Concession Road 8th Concession Road 9th Concession Road
SL-3
This screenline follows the southern edge of the study area from the South-East corner of the Hamilton-Wentworth Region, west, to the Brantford-Onondaga Township line.
Major roads crossing are: Highway #56 Highway # Regional Road 13 Regional Road 33 .(Tyneside Road’ County Road 22 (Trinity & Carluke Road)
Road allowance between lots, 30 & 3l.
28 & 29 24 & 25 20 & 2h Trinity Church Road
Minor road crossings are:
Lee ee
Miles Road Ferris Road Glancaster Road Nodsworth Road - Shaver Road Carlduke Road Trinity Road Butter Road
Spa
This screenline extends along the southern boundary of the Hamilton-wentworth Region from the Brantford-Onondaga Township Line (just east of Highway #2), up to, but not including High- way #8. It intercepts the Hamilton-Brantford corridor.
Major road crossings are: Highway #2 Highway #99 Highway #5 County Road 17 County Road 14
Minor roads crossing are: Alberton Road Book Road Ferguson Road Parsonage Road Ronald Road Misener Road Powerline Road Harrisburg Road 2nd Concession Road Patric Road 5th Concession Road 6th Concession Road 7th Concession Road Maclean Road
SL-5
This screenline extends along the northern boundary of Flamborough Township from Highway #8 (inclusive), east, to the regional bound- ary between Halton & Wellington. It contains the K/wW-Hamilton
and Guelph-Hamilton corridors.
Major roads crossing. are: Highway #8 Highway #97 & 52 Highway #52 Highway #6
Minor roads crossing are: Studiman By- Road Clyde Side Road Sheffield By-Road 8th Concession Road 10th Concession Road Foreman Side Road
- 39 -
Valens Side Road Lennon Side Road lith Concession Road 14th Concession Road Centre Road
SL-6
This screenline extends along the eastern boundary of Flamborough Township from the corner nearest Highway #401, south-east to the Niagara escarpment.
Major roads crossing are: Highway #5 County Road 18 County Road 36
Minor roads crossing are: Mountsberg Road lith Concession Road 10th Concession Road 9th Concession Road 8th Concession Road 7th Concession Road 6th Concession Road 5th Concession Road 4th Concession Road
SL-7
This screenline extends along the southern boundary of Flamborough (the escarpment) between King Road (inclusive) and Highway +6 (inclusive). It cuts across the Guelph-Hamilton/Burlington eorridor.
Major roads crossing are: Highway #6
. Minor roads crossing are: Waterdown Road Snake Road
SL-8
This screenline runs parallel to Highway #6 (on the west side) -from the Niagara escarpment down to Hamilton Harbour. It intersects the Toronto/Burlington-Hamilton corridor.
Major roads crossing are: Highway #403 Highway #2 york Road
Minor roads crossing are: Old Guelph Road Heath Street
SL-9
This screenline extends along the escarpment i.e. (the boundary between Flamborough and Ancaster), from Highway #6 (exclusive) to Highway #99 (exclusive).
Major roads crossing are: Highway #8 Regional Road 5 (Sydenham)
Minor roads crossing are: Valley Road Wier Side Road
Binkley Road
SL-10
This screenline runs parallel to Highway #52 (on the east side) from Highway #99 (inclusive) south to Highway #2 (inclusive), then extends east, parallel to Highway #2/53 (on the southern side), to the City of Hamilton Limits.
Major roads crossing are: Highway #99 County Road 23 : Highway #2 County Road 16 County Road 48
Minor roads crossing are: Mineral Springs Road
Powerline Road Shaver Road
Smith Road She 11
This screenline runs along the southern city limits immediately south of Highway 53 (Rymal Road).
Major roads crossing are: Highway #6
Minor roads crossing are: Aldercrest Avenue Seneca Avenue Springside Drive Miles Road’ Nebo Road Glover Road Trinity Chureh Road Glancaster Road
SL-12
This screenline extends along Highway 53 (on the southern side) from Trinity Church Rd. (exclusive)up to, but not including First Road East (of Highway 20) in Stoney Creek, then directly north to the escarpment.
a 47 =
Major roads crossing are: Highway #20 Highway #56 Regional Road 11 Regional Road 40
Minor roads crossing are: Highland Road
Green Mountain Road
SL-13
This screenline runs along the western Hamilton City Limits on the mountain between Highway #53 (inclusive) and the escarpment.
Major roads crossing are: Highway #53 Mohawk Road Proposed Mountain Freeway
Minor roads crossing are: Golf Club Road
SL-14
This screenline extends along the eastern Hamilton City Limits on the mountain from Highway #53 (inclusive) to the escarpment.
Major roads crossing are: Highway #53 (Rymal Road) Proposed Mountain Freeway (possibly)
Minor roads crossing are: Highland Road Mid soLre oe
SL-15
This screenline extends along the escarpment from Highway #403 (exclusive) to just east of the Upper Wentworth Street. It includes the major wester mountain accesses.
Major roads crossing are: Beckett Drive (Queen Street access) James Street Mountain Road
Claremont Access volley Cuc
Sr-16 This screenline extends along the escarpment from Wentworth Street to just west of Highway #20.
Major roads crossing are: Sherman Access Kenilworth Access
Mount Albion Road
SL-17
This screenline extends along the escarpment between Highway #20 (inclusive) and the eastern edge of the Hamilton-
Wentworth Region.
Major roads crossing are: Highway No. 20 New Mountain Road Regional Road 50
Minor roads crossing are: Dewitt Road McNeilly Road
SL-18
This screenline is located to the east of Highway #403 between Hamilton Harbour and the escarpment. It intersects the western approaches to Hamilton's CBD below the mountain.
Major roads crossing are: Aberdeen Avenue Main Street King Street YOrKNeiLvdae
Minor roads crossing are: Fred Street
SL-19
This screenline runs parallel to Wentworth Street (on the west side) from the Harbour to the escarpment. It intersects cross- town, east-west traffic.
Major roads crossing are: Charlton Avenue Main Street East King Street East Cannon Street Barton Street Burlington Street Proposed Perimeter Road
Minor roads crossing are: Stinson Street King William Street Nightingale Street Wilson Street Century Street Birge Street Shaw Street Burton Street Ferrie Street
SL-20
This screenline intercepts the eastern approaches to Hamilton's CBD. It extends along the Red Hill Creek fromthe harbour
up to the escarpment.
Major roads crossing are:
Minor roads crossing are:
SL-21
43 -
King Street East Queenston Road Barton Street Burlington Street
Hixon Street Melvin Avenue Brampton Street
This screenline is located along the CN mainline between the Harbour and Wentworth Street (exclusive).
Major roads crossing are:
Minor roads crossing are:
SL-22
James Street Wellington Street
Bay StLreec McNab Street John. Street Maly oErect Ferguson Avenue Victoria Avenue
This screenline is located along the CN mainline between Wentworth
Street (inclusive) and Red Hill Creek.
It separates the Bayfront
Industrial area from most of the major east-west arterial links.
Major roads crossing are:
Minor roads crossing are:
SL-23
Wentworth Street Birch Avenue Sherman Avenue Gage Avenue Ottawa Street Kenilworth Avenue Parkdale Avenue woodward Avenue
POLccriggqe Strece Avondale Avenue
This screenline is located along the CN mainline between the Red Hill Creek and the eastern edge of the Hamilton-City limits.
= ge
Major roads crossing are: Highway #20 (Centennial Parkway)
Minor roads crossing are: Nash Road Kenora Avenue . Lake Avenue Grays Road
SL-24
This screenline is located along the Beach Strip Canal which separates Hamilton from Burlington.
Major roads crossing are: QEW (Niagara)
Minor roads crossing are: Beach Blvd.
SL-25
This screenline is located along the western portion of the escarpment. It extends from east of Highway #403 to west of
Sulphur Springs Road.
Major roads crossing are: Highway #4403 Highway #2
Minor roads crossing are: Old Dundas Road (County Road No. 32) Sulphur Springs Road
APPENDIX C
Home Based Other (HBO) Trip Attraction Equation
Derivation of HBO Trip Attraction Equation
1. Assumptions
(a)
(b)
that the amount of land devoted to residential, manufacturing plus industrial, retail plus service and other uses is proportional to the corresponding amount of population and/or employees of a similar classification.
that future changes in the proportional mix of
employment (by type) and population reflect similar changes in the relative attractivity of the corresponding type of land use.
2. Vinput~ Data (1971)
(a)
(b)
(c)
(d)
(e)
(f£)
total study area manufacturing plus wholesale employment, Em = 81,696
total study area retail plus service employment, Es = 90,242
total study area other employment, Eo = 2,696
total study area population, POP = 488,294
total P.M. pk. hour HBO trips produced, PROD = 42,140
proportion of total HBO trips attracted (ok.nr.) aero function of land use*: - manufacturing plus wholesale,
Am = 0.02
- retail plus service,
As = 0.61
= Other,
Ao = 0.11
- residential
Apop = 0.26
—_—_—_———— EEE
* source:
Travel Characteristics Survey, 1974 ELS ee ee A Ee Wea PA
Equation Form
ATT (3)
-b
Rate(1) * Em(j) + Rate(2) * Es(j) Rate(3) * Eo(j) + Rate(4) * POP(4)
where ATT(j) = the number of P.M. peak hour HBO trips attracted to Zone 7,
Em(j}) = the number of manufacturing plus wholesale employees located in zone j (and so on).
Derivation of Coefficients
(a) calculate scaling factors (S) to convert the unit contribution of each variable to a common basis (population was selected as basic unit):
Spop a
Sm = POP/Em = Ses! Ss = POP/Es = 54k So = POP/Eo =e ro
(b) calculate relative attraction indices (R) for the four predictor Variables as the product of the unit Beaute factor (S) and the relative attractivity of the variable (A) as outlined in section 2(f).
Rm = Sm * Am = Deeb be
Rs = Ss * As = 3490
Ro = So * Ao = 19,9 Rpop = Spop * Apop = 0.26
(c) calculate scaling (balancing) factor which balances total HBO attractions (as estimated from rates in (b)) to productions.
SF = PROD/ (Rm*Em+Rs*Es+Ro*Eo+Rpop* POP)
0.0863
(a) determine final attraction co-efficients as product ose the relative attraction indices (R) and the scaling factor (st).
RATE(1) = Rm*SF = 0.0103 RATE(2) = Rs*SF = 0.285 RATE (3) = Ro*SF = 1.719 RATE (4) = Rpop*SF = 0.0224
APPENDIX D
External Travel Adjustments
Derivation of Screenline Adjustment Factors for External Travel
Adjustments to the magnitudes of external travel were made on
a screenline or corridor basis if sufficient data were available.
Depending on the nature of the travel in the corridor the adjust-
ment procedure would follow one of two procedures:
(1)
(2)
External Corridors (only external travel crossing):
Correction factors were computed as the ratio of ground counts (observed) to simulated data.
Internal Corridors (both internal & external travel crossing)
Since no explicit data was available on the observed amount of external travel in such corridors (during the P.M. peak hour) it was only possible to adjust the external assignments to the extent that adjustments were made to external corridors (as described in 1).
The corrective action was to subtract/add a percentage of the amount by which external assignments had been red- uced/increased in a related external corridor. The per- centage reflected the proportion of external travel that was common to both corridors.
(See table on next page)
TOPTAIOD (Teuzszx9) pazeTser sy} wory TeAer} Teuzreyzxe paysn{peun vsyz =D !HLON
0S8 eee = es 0Z°0 ELC < 8728'S CVC ; 98S°T yeod 4
= OSL‘ ae 6TL°9 _ sete € P8E‘E yeod vz
G20 | ose‘T | ie pos‘s || sez’p eT OT €L0‘9 6S8'Z
8ST‘?
Yeod-N ae :
uoT}OeATQ |(euTTUaaZOS tOptirio,
peysnlpy peysnlpeun
Teuta
(7) FOV ARANS einpedso0rg jUuSeuZSsnlpy
SLNUWLSOACGY THAVWYL IVNYALXa
APPENDIX E
Land Use Data Bank (Eo 7a)
Source: Socio-Economic Data, 1971 (TARMS) Census, 1971 (Statistics Canada)
Population Density Analysis in
Residential Neighbourhoods, 1972 (City of Hamilton)
LAND USE DATA BANK (1971)
Traffic Population Manufacturing Retail Sexvace Other Area Zone Construction (acres) Wholesale
1 470 12 aN 8 50 9801
2 665 24 55 119 7 3970
5 13370 0 21 74 73 12756
4 ase Si 47 97 168 138 9929
5 1,310 23 52 112 95 8993
6 620: fe) 2 5 31 8855
q 1,620 8 46 129 10d 22448
8 lite 26 16 33 58 7801
) 760 61 16 66 44 9968
10 1,265 39 10 45 8333 at 240 27 5 66 18 2492 12 2,560 88 49 204 26 2975 ass 1,800 i 12 101 99 6540 14 rane) fo) 99 rial 32 1862 15 880 a 12 20 26 2522 16 2,045 ey 138 73 404 isp 545 fe) 26 36 6 552 > 1,080 170 ae 231 20 LS a7 1 350 fe) 0 fg 10 1340 20 3,605 39 197 286 4,160 1 320 170 fe) Soe
22 1,420 374 > 498 128 es 2,970 D322 at egi 45 227 24 53450 (ER aNES 266 96 20 valet 25 355 Rone fo) 46 10 2069 26 2,795 4 37 176 5 1635 eT 3,415 5 sp eat 10 2344 28 Dat D 7 106 450 35 1281 29 2,035 oe 29 125 10 950 30 1,500 ) 24 3) 105 12342 a2 945 0 24 10 eke. 7456
- 55 - Traffic Population Manufacturing Retail Service Other Area
Zone Wholesale (acres) Construction 32 285 22 i. 59 30 4728 ce 385 . 5 5 cial 30 2364 ot T40 4. 9 4 65 6462 35 765 fo) 13 6 9 1448 36 1,795 an 56 454 68 5644 ay 750 0 25 31 ees 3989 38 660 2 23 a7 A846 39 450 O 2 2 28 3546 40 1,005 0 a1 17 1438 41 685 4 O 19 28 1812 42 980 39 4 21 28 5516 43 985 21 154 105 38 5142 44 510 O O 4 19 8540 45 425 6 4 6 32 — O79 46 925 2 21 37 33 ATAF 47 L235 70 ae 7 100. 8895 48 ols 41 19 40 30 3014 49 690 23 als) 8 7 1219 50 360 zu 4 8 8 1292 51 _ 800 4 8 O 10 1183 oe 2,855 6 Wee 23 ahs) 364 53 6,915 193 187 225 15 493 54 3,490 4 23 62 15 493 aye) 1435 150 18 82 25 502 56 640 615 7 33 25 1645 aL 2,300 62 48 340 25 2029 58 1,140: © 14 13 18 15 O72 op, 1.639 108 T 299 35 Tons 60 2,150 722 40 203 25 739 61 600 2,190 145 339: 5 621 62 O ) O 24 —_—, 381 63 2,280 66 39 285 16 229
64. 5,075 633 50
gic ill os zone
65 66 67 68 69 70 7a ee te 74 75 76
93
2. 96
Population
230 1g 60 3,055 9,650 4,620 787100 See) 9,665 - 9,300 95535 Soles 2,035 2,965 De esas Tue TS 25 7 GO a goe? : Bye) 8,360
os 6 iia eee
Manufacturing Retail Service Ovner Construction Wholesale
2,904 110 ee 0 2 5050. 212 = 187 0 18,981 66 266 0 4,220 29 50 0 5524 29 68 O 4,969 fe) 319 fe) 2h 93 174 4 ie | Bye 64 fe) 26 fe) 30 0 peo 120 687 fe) 523 116 401 0 T4097 ib 87 0 30 33 45 0 ee 128 141 0 578 372 O7 O 515 201 566 fe) 310 tal 503 0 107 34063 156 0 ty230 4&2 865 O oo 341 897 0 2.5 304 621 eee. 0 484 666 3,298 O taeeo 890 2,158 oun 1,225 890 2,158 0 1,225 S90) 27ee 0 2,040 2178. 32380 0 600 227 828 0 121 234. 3,705 0 600 227 828 O 157 O a O 3 alae 98 0 3,090 309 eau 0
Traffic Population
Zone
10,005 7,040 55325 5470 3,940 2,845 59535 3,940 3,810 6,995
95 2,240 150
oF
5 5835 7,160 11,840 9,065
12,950
- 9,040 8,200 8,660 3,660 3,840 1,670 4,375
73590 —
4,445 2,975 560 280 285
Sas te
Mamufacturing Retail Service
Construction Wholesale
aii 602 oe) 158 awe 36
293 S25
Other
aoe On OT 0.0 0079 00 OOo SOO 650 6 oO Ww Oo 6 oOo ooo
- 58 -
Traffic Population Manufacturing Retail Service Other Area Zone ‘Construction (acres) Wholesale 129 1,060 3 7 108 on 642 130 740 nly 58 42 9 678 131 600 55 14 56 O 389 132 1,010 5 aly 1. ©) 365 Dagee" > 1,065 9 49 Te 0 650 134 335 36 fe) aS fe) 5438 Lo 200 2 fe) fe) O 410 136 igo 6 12 938 O 296 i347 33292 (aks, 385 478 Om L7Gs 138 8,990 £12 . 265 422 O 995 139 21,826 2.137 ee 2,979 447 2416 140 10,596 606 1,665 718 0 1654 141 8,624 808 193 bub! O 174 142 14,132 405 253 240 6) 1861 143 1,676 497 41 239 fe) 3644 144 12,085 Pay 283 1.000 O 5250 145 Pio (2 486 Ng) 78 0 4137 146 1,662 183 21 63 a2 12866 147 1,376 50 15 83 sbi 17744
148 = 916 O 4 21 ef 14630