HomeMy WebLinkAboutGP_UPDATE_TRAFFIC_STUDY_METHODOLOGY*NEW FILE*
GP -UPDATE -TRAFFIC -STUDY
_METHODOLGY
0
MEWPORT BEACH TRAFFIC MODEL
(MBTM) DRAFT METHODOLOGY
OVERVIEW REPORT
John Kain, AICP
Carleton Waters, P.E.
Bill Lawson, AICP
Scott Sato, P.E.
n 41 Corporate Park, Suite 210 Irvine, CA 92606
p: 949.660.1994 • f: 949.660.1911
e: admin@urbanxroads.com • www.urbanxroads.com
NEWPORT BEACH TRAFFIC MODEL (NBTM)
DRAFT METHODOLOGY OVERVIEW REPORT
Prepared For:
Mr. Rich Edmonston
CITY OF NEWPORT BEACH
P.O. Box 1768
Newport Beach, CA 92658-8915
Prepared By:
URBAN CROSSROADS, INC.
41 Corporate Park, Suite 210
Irvine, CA 92606
John Kain, AICP
Carleton Waters, P.E.
Marlie Whiteman
February7, 2002
JK:CW:MW:rd
00460-02
TABLE OF CONTENTS
SECTION
PAGE
1.0 INTRODUCTION.............................................................................................. 1-1
1.1
Project Purpose
r 1.2
Model Consistency Issues
1.3
Goals and Objectives
1.4
Methodology Overview
1.5
Validation and Consistency Criteria
2.0 MODEL STRUCTURE DESCRIPTION............................................................ 2-1
2.1
Traffic Analysis Zone (TAZ) Structure
2.2
Land Use to Socioeconomic Data (SED) Conversion Factors
2.3
Trip Generation and Mode Choice
2.4
Trip Distribution
2.5
Time of Day Factoring
2.6
Roadway Network Representation and Traffic Assignment
2.7
Post Assignment Data Refinement Procedures
NCHRP-255 TURN MOVEMENT ESTIMATION ALGORITHM DISCUSSION ............. A
NBTM TAZ TRAFFIC ANALYSIS ZONE (fAZ)
CORRESPONDENCELISTING..................................................................................... B
LIST OF EXHIBITS
EXHIBIT PAGE
1-A DRAFT NBTM OVERALL COVERAGE AREA ....................................... 1-5
1-B DRAFT NEWPORT.. BEACH TRAFFIC MODEL (NBTM)
PRIMARY STUDY AREA ............. :...........................................................
1-7
1-C DRAFT NEWPORT BEACH TRAFFIC MODEL (NBTM)
OVERALL MODELING METHODOLOGY ............................................... 1-8
1-D NEWPORT BEACH TRAFFIC MODEL SCREENLINES & INTERSECTION
ANALYSISLOCATIONS........................................................................ 1-13
2-A DRAFT TRAFFIC ANALYSIS ZONE (TAZ) SYSTEM ............................. 2-2
2-B DRAFT TIER 3 TRAFFIC ANALYSIS ZONE (TAZ) SYSTEM ................. 2-3
2-C DRAFT NEWPORT BEACH TRAFFIC MODEL (NBTM) TRIP
GENERATIONPROCESS....................................................................... 2-5
2-D DRAFT NEWPORT BEACH TRAFFIC MODEL (NBTM) TRIP
DISTRIBUTION PROCESS.................................................................. 2-13
2-E DRAFT NBTM TRAFFIC ASSIGNMENT ALGORITHM
VOLUME/CAPACITY RATIO TO TRAVEL SPEED RELATIONSHIPS 2-21
2-F DRAFT GENERALIZED NBTM POST MODEL REFINEMENT
PROCESS............................................................................................. 2-23
2-G DRAFT DAILY ROADWAY SEGMENT TRAFFIC VOLUME
REFINEMENTPROCESS................................................................. 2-24
LIST OF TABLES
TABLE PAGE
2-1 DRAFT LAND USE TO SOCIOECONOMIC DATA
CONVERSIONFACTORS....................................................................... 2-7
2-2
DRAFT NBTM SOCIOECONOMIC DATA (SED) BASED
TRIPRATES............................................................................................
2-9
2-3
TYPICAL DRAFT NBTM RESIDENTIAL TRIP GENERATION
2-11
EXAMPLES...........................................................................................
2-4
DRAFT NBTM TIME OF DAY FACTORS ..............................................
2-14
2-5
DRAFT NBTM LINK ATTRIBUTES........................................................
2-16
2-6
DRAFT NBTM ROADWAY LINK FACILITY TYPE CODES ..................
2-17
2-7
DRAFT NBTM USE CODES..................................................................
2-17
2-8
DRAFT NBTM ROADWAY LINK SPEED ASSUMPTIONS ...................
2-19
2-9
DRAFT NBTM ROADWAY LINK PEAK AND OFF-PEAK CAPACITY
ASSUMPTIONS.....................................................................................
2-20
NEWPORT BEACH TRAFFIC MODEL (NBTM)
DRAFT METHODOLOGY OVERVIEW REPORT
February 7, 2002
1.0 INTRODUCTION
This report has been prepared to document the update of the Newport Beach Traffic
Model (NBTM) to provide consistency with the version of the subregional travel demand
model currently being used for long range planning purposes in Orange County. The most
current version of the ,Orange County Transportation Analysis Model is Version 3.1
(OCTAM 3.1). All data included in this report is preliminary and subject to further review in
the model development process.
This chapter of the report introduces the reader to the NBTM update project and presents
the goals and objectives of the NBTM work effort. It also provides a summary of the
overall NBTM modeling methodology and the criteria used to evaluate the consistency of
the NBTM model results with the parent OCTAM 3.1 results.
The NBTM travel demand forecasting tool will be developed for the City of Newport Beach
to address traffic and circulation issues in and around the City. The NBTM tool will be
developed in accordance with the requirements and recommendations of the Orange
County Subarea Modeling Guidelines Manual (August, 1998). The NBTM is intended to
be used for roadway planning and traffic impact analysis, such as:
• General Plan/Land Use analysis required by the City of Newport.Beach.
• Amendments to the Orange County Master Plan of Arterial Highways (MPAH)
• Orange County Congestion Management Program (CMP) analysis.
The NBTM is a vehicle trip based modeling tool, and it is intended for evaluating general
roadway system supply and demand, problems and issues.
1-1
1.1 Project Purpose
The NBTM travel demand forecasting tool is a subarea model which is intended to
provide a framework for ongoing analysis throughout the City of Newport Beach
and surrounding areas. The overall project effort will include developing the NBTM
travel demand forecasting tool in accordance with the draft Orange County
Consistency Guidelines and providing a variety of automated post model
refinement procedures/inputs that can be used to develop refined traffic projections
suitable for various traffic impact analysis needs.
1.2 Model Consistency Issues
Advancements in travel demand forecasting methodologies and computer
technology have resulted in a proliferation of microcomputer based modeling tools.
The development of many local area models with overlapping areas allow for
multiple forecasts from different sources being developed for the same roadways.
These forecasts are often different due to differences either in the underlying input
assumptions or modeling methodologies themselves.
The need to develop travel demand models which are based on consistent inputs
and which produce consistent results has been increasingly recognized at the
regional and local levels. Legislative requirements have been introduced at both
the regional and local levels to ensure that travel demand models are in fact based
on consistent inputs and produce consistent results, with any differences being
explainable in terms of enhanced accuracy for either the local or regional model. A
local model, for instance, may introduce additional local detail that can increase the
accuracy of forecasts on the local street network. At the same time, a regional
model may include more sophisticated procedures for forecasting travel demand at
any number of levels, providing for more accurate forecasts by the regional model
in some cases. An example might be more sophisticated models of transit system
characteristics -and therefore more accurate, transit patronage forecasts.
1-2
The general credibility of travel demand models is also enhanced when different
modeling tools can produce consistent forecasts, enhancing the ability to explain
why different forecasts are being provided for the same roadways. A consistent
approach enables regional, subregional, and local jurisdictions to have a common
basis for evaluating transportation plans, programs, and projects.
1.3 Goals and Obiectives
The overall goals of this project are to develop and apply a subarea model that (1)
satisfies the model consistency guidelines for Orange County, and (2) provides a
useful tool for City traffic analysis purposes. The consistency guidelines have been
developed in response to recently adopted federal guidelines that are more
stringent than previous federal requirements. The objectives for this work effort
support the overall goals. The objectives include:
1. Developing a subarea model that is consistent with the parent model
(OCTAM) and provides forecasts within a reasonable time frame.
2. Provide the framework necessary to prepare future traffic forecasts for 2007,
2025 and General Plan buildout conditions.
3. Providing a subarea modeling tool which results in more refined traffic
volume forecasts for the primary study area.
4. Automate/facilitate the development of refined future traffic volume
4 projections.
r
The objective of developing a subarea model that is consistent with OCTAM and
` provides forecasts within a reasonable timeframe is important to support
F
interjurisdictional study efforts with the various local jurisdictions bordering the City
of Newport Beach. The proposed subarea model approach and structure must also
execute (run) within a relatively short time frame, compared to the parent OCTAM
tool. An initial objective of less than 4 hours to execute is recommended by Urban
Crossroads, Inc. staff. An upper limit of 6 hours execution time is believed to be the
L
1-3
absolute maximum that the typical end user will accept before frustration and
disillusionment set in. Additional analyst's time is required to prepare necessary
Input files and review and rdocument run assumptions and results.
The objective of providing a framework for developing future traffic projections for
2007, 2025, and General Plan buildout timeframes will be satisfied by developing
highway networks and vehicle trip tables for each of the proposed time frames, in
addition to the calibration/validation year network(s) and trip table(s).
The objective of automating/facilitating the subarea modeling process is addressed
by providing a complete set of data inputs and extraction/refinement tools that
automate the post -assignment data refinement process.
The objective of providing more refined traffic forecasts is generally expected to
result from a more refined TAZ structure and roadway network within the primary
study area. The detailed TAZ structure and roadway network are expected to
improve the sensitivity of the subarea model to changes in local area
socioeconomic data (SED)/land use and/or the roadway system.
1.4 Methodology Overview
This subsection provides a broad overview of the NBTM structure. The detailed
parameters of the NBTM (e.g., trip rates, time of day trip table factors, etc.) will
be presented in subsequent chapters. The overall coverage area of the NBTM is
depicted on Exhibit 1-A. The NBTM coverage area includes the five county
urbanized area which is included in the parent OCTAM 3.1 tool. f
The basic model structure recommended in the subarea modeling guidelines is a
"focused" modeling approach. The concept of a focused model is to provide the
greatest level of detail within the primary analysis or study area, with the least detail
1-4
I
EXHIBIT 1 A
DRAFT NBTM OVERALL COVERAGE AREA
Ln
included in those parts of the model which are geographically distant from the
primary study area. This concept is further refined in the guidelines as a three tier
system.
Tier 1 is the least detailed component of the subarea model. The intent of the Tier
1 level of definition is to provide the minimum amount of detail necessary to
accommodate regional (OCTAM 3.1) traffic as it enters the Tier 2 coverage area.
The Tier 1 level of detail is not intended to support detailed analysis within the Tier
1 area.
The Tier 2 level of detail corresponds directly to the parent (OCTAM 3.1) model,
while Tier 3 incorporates more detail than the parent model. Exhibit 1-A also
presents the limits of each tier or level of detail. While the Tier 3 area incorporates
additional detail surrounding the City of Newport Beach, the City will be the primary
study area for this work effort.
The primary study area of the NBTM is shown on Exhibit 1-B. The primary study
area is generally bounded by Brookhurst Street/the Santa Ana River on the west,
Adams Avenue/Baker Street/Campus Drive/SR-73 on the north, Crystal Cove State
Park on the east, and the Pacific Ocean on the south. As described previously, Tier
2 area level of detail and vehicle traffic forecasting capability is equal to that of the
parent OCTAM 3.1 travel demand forecasting tool. The Tier 2 area is generally
bounded by the northwest Orange County line, 1-5 Freeway, Fairhaven Avenue,
Santiago Canyon Road, El Toro Road, Santa Margarita Parkway, Trabuco Creek,
and the Pacific Ocean.
The NBTM is highly dependent on the Orange County Transportation Analysis
Model, Version 3.1 (OCTAM 3.1). Exhibit 1-C provides an overview of the NBTM
modeling process. The general modeling steps or processes are:
1-6
M� +
EXHIBIT 1-B
DRAFT NEWPORT BEACH TRAFFIC MODEL (NBTM) PRIMARY STUDY AREA
R ®® V
NEWPORT BEACH TRAFFIC MODEL UPDATE Newport Beach CalHomia •00460:04 URBA�
EXHIBIT 1-C
DRAFT NEWPORT BEACH TRAFFIC MODEL (NBTM)
Sulmr" Trip
Generation
OVERALL MODELING
Regional Tdp Ganeretion,
Moda Cholu, and Trip
Distribution Models
Reglonal YaMde
Trips (Drive alone,
HOYs, T0ti5)
Tier 162 Zones
Are local taps
or netwo
substantlallydW renl
from reglonei trlps or
network?
No
LU S.E
pdA Growti>/
Expansion Factors
Fratar/Matdx Expand
METHODOLOGY
I'
Regional Highway
Network
Revaluate
Reglonai Tra%.*
Charecteri:tks
Subarea Network I
Yes
i
I
I
i
HORAN !
1-8
I
. Land use to socioeconomic data (SED) conversion.
. Trip generation and mode choice.
. Trip distribution.
. Time of day factoring.
. Traffic assignment.
. Post -assignment data refinement processing.
NBTM relies on regional model estimates of trip generation, trip distribution, and
mode choice. The model structure accommodates changes in land
use/socioeconomic and network characteristics in the following manner.
Trip Generation - Trip generation estimates are based on socioeconomic
data driven trip generation rates. The socioeconomic
data is usually derived from the City of Newport Beach
land use within the City of Newport Beach Sphere of
Influence. The calculated trip generation is then used
to adjust the regional trip generation results to match
the more detailed local (NBTM) trip generation
estimate.
Trip Distribution - Trip distribution estimates are based on trip distribution
patterns estimated by the regional travel demand
model and incorporated into the subarea model. The
number of trips attributed to the primary study area in
the regional model are adjusted to match the project
trip generation using an analytical approach commonly
referred to as the Fratar model. This process
automatically adjusts the trip distribution patterns as
necessary.
WO
Mode Choice - Mode choice is estimated by using regional model
mode share results, which are then incorporated
directly Into the subarea model.
Traffic Assignment - Traffic is assigned to the roadway system on the basis
of travel time and cost. Tolls are explicitly included in
the traffic assignment process using the procedures
obtained from the regional travel demand model.
Traffic is assigned separately for the AM, mid -day, PM,
and nighttime periods of the day, to allow for more
accurate representation of the effects of congestion on
the choice of travel routes by drivers.
f
Post Model Refinements -The goal of the future traffic volume forecast
refinement or post model refinement processing is to
utilize all available data to prepare the best possible
estimate of future traffic conditions. The NBTM
procedure incorporates 2001 traffic count data, 2001 i
modelvalidation data (traffic estimates), and future
(raw) model forecasts (estimates) as inputs.
1.5 Validation and Consistency Criteria
The focus of the current work effort is to develop an Newport Beach Traffic Model
(NBTM) that satisfies the Orange County Transportation Authority (OCTA)
consistency requirements and provides more detailed/accurate local area traffic
forecasts than can be obtained from the subregional modeling tool. The most
recent set of consistency guidelines were published by OCTA in August,1998.
1-10
This section describes the consistency and validation criteria used to evaluate the
performance of the NBTM tool in the context of the parent OCTAM regional travel
demand model and the comprehensive traffic count database collected and/or
compiled in support of this work effort.
Demonstrated model consistency between the NBTM and the OCTAM 3.1 model is
required before the NBTM can be -used to design regionally significant infrastructure
improvements (e.g., substantial arterial roadway widening). Similarly, the
Congestion Management Program (CMP) and, Growth Management Plan (GMP)
each requires analysis of regionally significant land use decisions, also based on
the OCTAM 3.1 model or a consistent subarea model. This report, therefore,
provides all of the information required to verify that the NBTM is consistent with the
OCTAM 3.1 travel demand forecasting tool.
` As described previously, the basic model structure recommended in the subarea
modeling guidelines is a "focused" modeling approach. The NBTM tool has been
developed using the recommended model structure. Because the Subarea
Modeling Guidelines preferred model structure has been applied, the effort to
demonstrate socioeconomic data, trip generation, trip distribution, and mode choice
consistency has been optimized.
The basis for testing of the NBTM performance with respect to traffic volume
forecasts (model validation) are criteria set forth in the document Highway Traffic
Data for Urbanized Area Project Plannino and Desian. This document is also
known as National Cooperative Highway Research Program Report 255
(NCHRP-255). This report documents the difficulties associated with evaluating
model performance. Variability in day-to-day traffic conditions is a well
established fact, and the criteria included in NCHRP-255 provide a reasonable
basis for evaluating model traffic forecasts.
Screenlines are a primary tool used to evaluate model performance. A screenjine
is an imaginary cordon drawn across a series of roadways that serve a common
traffic flow. The 11 screenlines used to evaluate the NBTM are presented on
Exhibit 1-D. NCHRP-255 presents graphs that represent the maximum desirable
deviation from actual traffic counts that can be expected from a modeling tool such
as the NBTM. Separate graphs are presented for screenlines and individual
roadways. Exhibit 1-E reproduces the NCRHP-255 graphs.
'it should be noted that NCHRP-255 explicitly states that "Although no specific rules
exist as to when an assignment should be considered acceptable, the vast majority
of links should have assigned traffic volumes that fall within the maximum desirable
deviation" as shown on Exhibit 1-E.
NCHRP-255 does not establish standards for peak hour intersection capacity 4
utilization (ICU) validation. The performance of the previous NBTAM model, as
reported in the model validation report, has been used as the validation target for
this work effort. The previous ICU validation resulted in an average difference of
0.02 for the AM and PM ICU values throughout the study area, with a standard
deviation of 0.11 for the sample size of 59 intersections.
1-12
Infer rF In v � 'e Wes•-R Y' M V .n N • � a 1 ) 1 H 1
NEWPORT BEACH TRAFFIC MODEL
EXHIBIT 1-D
8 INTERSECTION ANALYSIS LOCATIONS
LEGEND:
J O -SCREENLINEID
PACIFIC SCREENLINELOCAMON
OCEAN ---=FUTUREROADWAYS
• . = INTERSECTION ANALYSIS LOCATION
EXHIBIT 1•E
- 5S DAILY TRAFFIC VALIDATION CRITERIA
ERIA 2 ,
40
SO
40 \
Y
•
a 00 Screenline
efrom Table A4
e and Figure A-11 !•!
u
20
OD
10
OS
Lgand
A - Sereenlines Depleted
in Figure A-t and Table n-4
Desirable Deviation
•C
OA
00 25. 50 75 100 125 150 175 200
Total Sereenline 24-Hour Sass Year Traffic Count (1000'si
Figure A-9. mazimum desirable deviation In total screenline volumes.
ease Year Aaeigru/ant - Base Tear Count
fareent Devlatton• ease Year Coupe
4►
Link A
t 2e/� 40,000 now 40/000 ie/000 ao/Boo lO m •100/aoo
a 10/aa0 20/000
sate Year Cant:
fyee A•1. Haab$" dr* W Mf Mr 1* wt P
spuRM,MTiOB4AL COOPATM (#t WAY MMAPM PWWAM WM 30RMx"'-M
1'-14
4 T
2.0 MODEL STRUCTURE DESCRIPTION
This chapter of the' report describes the Newport Beach Traffic Model (NBTM) structure
and presents the NBTM parameters that remain constant for all modeling time frames.
The specific parameters have been adapted from the subregional modeling tool (the
Orange County Transportation Analysis Model, Version 3.1 or OCTAM 3.1), with
adjustments as necessary to reflect the specific travel demand characteristics of the
City of Newport Beach.
Each of the general model steps is discussed as a separate heading. In addition, the
model traffic analysis zone (TAZ) structure is discussed first, as this is an equally
important aspect of the subarea modeling process.
2.1 Traffic Analysis Zone (TAZ) Structure
The NBTM overall TAZ Structure is shown on Exhibit 2-A. The primary study
area (City of Newport Beach) TAZ structure is shown on Exhibit 2-13 and
incorporates _ TAZs for purposes of aggregating individual land uses to a level
of detail suitable for local area modeling. By contrast, the OCTAM 3.1 TAZ
system includes _TAZs for the same area. The additional TAZ structure detail
is intended to support accurate forecasting of traffic on all arterial roadways (as
well as study area freeways) within the study area.
i The NBTM TAZs aggregate to the OCTAM 3.1 TAZs within the primary modeling
i
area. This is a requirement of the consistency guidelines. Within the NBTM
secondary (Tier 2) analysis area, the NBTM TAZs correspond to the OCTAM 3.1
TAZs on a one-to-one basis. Appendix "A" (to be provided) contains a complete
` listing of the relationships between the NBTM TAZs, the OCTAM 3.1 TAZs, and
various other geographic areas such as counties, Community Analysis Areas
r (CAAs) and Regional Statistical Areas (RSAs).
2-1
EXHIBIT Z A
DRAFT TRAFFIC ANALYSIS ZONE (TAX) SYSTEM
N
N
NENPORT BEACH 7RAFMC MMEL N .Heath Canfomlo — OO460:tt23 —rev. 02 10
o- r n m n. � s •n n + � � . �
N
W
EXHIBIT2-B
DRAFT TIER 3 'TRAFFIC ANALYSIS ZONE (TAZ) SYSTEM
TO BE PROVIDED BY CITY OF NEWPORT BEACH
2.2 Land Use to Socioeconomic Data (SED) Conversion Factors '
The conversion of land use to SED is the first step in the NBTM modeling
process. Exhibit 2-C illustrates the overall land use to SED conversion and trip
generation process. The City of Newport Beach maintains land use data that is
used for many purposes, including providing input data to the NBTM traffic 1
forecasting process.
Recently adopted regional modeling consistency requirements necessitate use of
consistent input data that provides trip generation estimates that are also consistent
with the regional modeling tool. OCTA has provided the following variables as the
Input data for OCTAM V : t
• (Total) Population i
• Household Population I
• Employed Residents
• (Non -Institutionalized) Group Quarters Population j
• Occupied Single -Family Households
• Occupied Multiple -Family Households (including all households other than t
single family households) !
• (Total Occupied ) Dwelling Units 1
• Retail Employment 1
• Service Employment
• Other Employment (Non -Service and Non -Retail)
• Total Employment
• Median Household Income
• ElementarylHigh School Enrollment
• (Non -Resident or Commuter Student) University Enrollment
c
2-4
EXHIBIT 2-C
DRAFT NEWPORT BEACH TRAFFIC MODEL (NBTM)
TRIP GENERATION PROCESS
NEWPORT BEACH
LAND USE DATA
CONVERT LAND USE'
TO SOCIOECONOMIC
DATA (SED)
DAILY TRIP RATES GENERATE TR
BY PURPOSE AND FROM SED
PRODUCTIONS/
ATTRACTIONS
DAILY TRIPS BY
PURPOSE FROM SED
ADD DAIL
TRIPS BY
PRIMARY STUDY AREA
LOCAL DAILY VEHICLE
TRIPS BY PURPOSE AND
LEGEND:
INPUT/OUTPUT DECISION MODELING
DATA RULE PROCESS
LAND USE SUPPLEMENTAL
BASED SED SED
ADD SED
OVERALL
SED
SPECIAL GENERATOR
DAILY VEHICLE TRIPS
2-5
i
Many of these variables are self -descriptive. A brief explanation is provided for
those variables which are not self -descriptive.
Non Institutionalized Group Quarters Population: Non -institutionalized group
quarters population refers to military personnel living in barracks and students living
I
In dormitories. It also includes similar populations, such as seminaries, convents,
orphan homes, agricultural workers living in dormitories/barracks, homes for unwed
mothers, and institutional staff (at hospitals, prisons, etc.) who live on the premises
where they work.
Retail Employment; The definition is consistent with the definition presented in the
documentation for the OCTAM '3.0 (Table B.1, Appendix "B", OCTAM III Trip
Generation Program Documentation, June 30, 1997 version). Per this definition, all
employment falling into Standard Industrial Classification (SIC) Division G (major
groups 52 - 59) is considered retail employment.. These codes include retail shops,
eating and drinking establishments (SIC 58), etc.
Service Employment: For the purposes of this modeling effort and consistent with
OCTAM 3.1, all employment falling into Standard Industrial Classification (SIC)
Divisions H and I (major groups 60-89) is considered as service employment.
Examples of service employment include banking, insurance agents, real estate
offices, hotels and other lodging, personal services (dry cleaners, beauty salons,
etc.), auto repair shops, medical/dental offices, educational services (schools,
libraries, etc.), and social services. }
Conversion factors for each of the land use included in the City's land use have
been developed by Urban Crossroads, Inc. staff. Table 2-1 summarizes the i
NBTM land use to socioeconomic data conversion factors. The conversion
factors have been derived from other recently completed studies. Certain
variables, particularly non -institutionalized group quarters population and
2-6
TABLE 2-1
N
I
J
DRAFT LAND USE TO SOCIOECONOMIC DATA CONVERSION FACTORS
NSTM
LAND
USE
CODE
NBTM LAND USEDESCRIPTION
UNITS
OCCUPANCY
RATE
SINGLE
FAMILY
DWELLING
UNITS
MULTI.
FAMILY
DWELLING
UNITS
GROUP
QUARTERS
POPULATION
POPU-
LATION
RESIDENT
WORKERS
RETAIL
EMPLOY-
MENT
SERVICE
EMPLOY-
MENT
OTHER
EMPLOY-
MENT
TOTAL
EMPLOY-
MENT
ELEMEN-
TARYIHIGH
SCHOOL
STUDENTS
UNIVERSITY
STUDENTS
1
Res-l- w(SFD)
DU
0.97
1
0
0
3.05
1.70
0.00
0.00
0.00
0.00
0
0
2
Res-Medium(SFA)
DU
0.98
0
1
0
2.80
1.60
0.00
0.00
0.00
0.00
0
0
3
Apartment
DU
0.98
0
1
0
220
1.40
0.00
0.00
0.00
0.00
0
0
4
Elderly Residential
DU
1.00
0
1
0
1.40
025
0.00
0.00
0.00
0.00
0
0
5
Mobile Horne
DU
0.98
0
1
0
2.20
1.00
0.00
0.00
0.00
0.00
0
0
6
Motel
ROOM
0.90
0
0
0
0.00
0.00
0.10
1.00
0.20
1.30
0
0
7
Hotel
ROOM
0.90
0
0
b
0.00
0.00
0.10
1.00
0.20
1.30
0
0
8
Resort Hotel
ROOM
0.90
0
0
0
0.00
0.00
0.10
1.00
020
1.30
0
0
9
Regional Commercial
TSF
0.90
0
0
0
0.00
0.00
2.10
0.15
0.05
2.30
0
0
10
General Commercial
TSF
0.90
0
0
0
0.00
0.00
2.10
0.15
0.05
2.30
0
0
11
CommJRecnsdlon
ACRE
1.00
0
0
0
0.00
O.OD
1.20
4.00
0.00
5.20
0
0
12
Regional Commercial
TSF
0.90
0
0
0
0.00
0.00
2.10
0.15
0.05
2.30
0
0
13
Restaurant
TSF
1.00
0
0
0
0.00
0.00
3.20
0.00
0.00
320
0
0
14
Family Restaurant
TSF
1.00
0
0
0
0.00
0.00
3.20
0.00
0.00
320
0
0
15
Fast Foal Restaurant
TSF
1.00
, 0
0
0
0.00
0.00
4.00
0.00
0.00
4.00
0
0
16
Auto Dealer/Sales
TSF
1.OD
0
0
0
0.00
0.00
1.60
0.30
0.00
1.90
0
0
17
Yacht Club
TSF
1.OD
0
0
0
0.00
0.00
0.60
2.50
0.00
3.10
0
0
18
Heallh Cub
TSF
1.00
0
0
0
0.00
0.00
0.60
2.50
O.OD
.3.10
0
0
19
Tennis Club
SG
20
Madia
SG
21
Theater
SEAT
1.00
0
0
0
0.00
0.00
0.01
0.02
0.00
0.03
0
0
22
Newport Dunes
SG
23
General 011ice
TSF
0.90
0
0
0
0.00
O.OD
0.20
0.78
2.20
3.00
0
0
24
Medical 011ice
TSF
0.90
0
0
0
0.00
0.00
0.30
3.50
0.00
3.00
0
0
25
R a D
TSF
0.90
0
0
0
0.00
0.00
0.00
0.90
1.50
2.40
0
0
26
Industrial
TSF
0.90
0
0
0
0.00
0.00
O.OD
0.00
1.80
1.60
0
0
27
MiniStoragaWarehouse
TSF
0.90
0
0
0
0.00
0.00
0.00
0.10
1.70
1.60
0
0
28
PreSchoollDay Care
TSF
1.00
0
0
0
0.00
O.OD
0.00
7.00
0.00
7.00
0
0
29
ElementarylPdvate School
STU
1.00
0
0
0
0.00
0.00
0.00
0.00
0.11
0.11
1
0
30
JudorMigh School
STU
1.00
0
0
0
0.00
0.00
0.00
0.00
0.11.
0.11
1
0
31
CdturaIlLeaming Center
TSF
1.00
0
0
0
0.00
0.00
0.00
3.00-
0.00
3.00
0
0
32
Library
TSF
1.00
0
0
0
0.00
0.00
0.00
3.06
0.00
3.00
0
0
33
Post OlOce
TSF
1.00
0
0
0
O.OD
0.00
0.10
3.00
1.00
4.10
0
0
34
Hospital
SG
35
NuminglCorw. Home
BEDS
1.00
0
0
1
1.00
0.00
0.00
0.33
0.00
0.33
0
0
36
Church
TSF
1.00
0
0
0
0.00
0.00
0.00
0.8D
1.00
1.60
0
0
37
Youth CU/Sondce
TSF
1.00
0
0
0
0.00
0.00
0.00
7.00
0.00
7.00
0
0
38
Park
ACRE
1.00
0
0
0
0.00
0.00
0.00
0.60
0.00
0.60
0
0
39
Regional Park
ACRE
1.00
0
0
0
0.00
0.00
0.00
0.60
0.00
0.60
0
0
40
Golf Course
ACRE
1.00
0
0
0
0.00
0.00
0.10
0.60
0.00
0.70
0
0
41
Resod Golf Course
ACRE
1.00
0
0
0
0.00
0.00
0.10
0.60
0.00
0.70
0
0
42
Saiboa Area Comm1
TSF
0.90
0
0
0
0.00
0.00
2.10
0.15
0.05
2.30
0
0
43
Balboa Ama Restaurant
TSF
1.00
0
0
0
0.00
0.00
320
0.00
0.00
320
0
0
U:\UcJobs\004601ExceR)sedconvadsjT 2-1
I
university enrollment, will be maintained as special SED variables that are not
derived from the underlying data in the citywide land use database.
2.3 Trip Generation and Mode Choice
Subarea models are now required to match (nearly exactly) regional trip i
generation estimates derived from socioeconomic data (SED) at the regional
model traffic analysis zone (TAZ) level. It has long been recognized that there
are definite differences between land use and SED based trip generation
approaches. These differences have been addressed and very nearly resolved l
during the course of this work effort. The approach taken for NBTM is to convert
land use to SED and generate traffic that is fairly consistent with the regional trip ;
generation estimates (and controlled to the regional TAZ level totals for the
consistency scenarios). NBTM trip generation data is developed for the following
7 trip purposes:
• Home -Work
• Home -Shop r
• Home -Other
• Home-Elementary/High School
• Home -University
• Other -Other
• Other -Work
Examples of types of trips that fill into the "Other" category include social or
entertainment related trips and recreational trips. Table 2-2 summarizes the daily
vehicle trip generation rates by purpose and Input variable. These purposes are
aggregated to the 5 regional model trip purposes available following mode choice: I
• Home-Work/University
• Home-Elementary/High School .
l&:
0
VARIABLES
Single
Family
Residential
Multi Family
Residential
Population
Employed
Residents
Income
Retail
Employment
Service
Employment
Other Employment
EIJHS Enrollment
University!
Coil. Enroll.
DU
DU
POP
E-R
$MIL
EMP
EMP
EMP
Stu
StuION
TRIP RATE
0
0
0
1.27
0
0
0
0
0
0
0
0
0
0
0
1.83
1.07
1.01
0
0
1.05
0.60
0.24
0
13
0
0
0
0
0
P
0.89
0.46
0.11
0
11
0
0
0
0
0
0.44
0.43
0
0
2
5.2
1.08
0.24
0
0.2
0
0
0.04
0
0
0
0
0
0
0
0
0
0.147
0
0
0
0
0
0
0
ION TRIP RATES
H-W
0.1
0.1
0
0
0
1.24
1.24
1.26
0
0
O-W
0.25
0.25
0
0
0
3A4
0.6
0.54
0
0.2
H-0
0.4
0.39
0
0
1
3A6
0.9
0.1
0
0
HSho
0
0
0
0
0
5.54
0
0
0
0
0-0
0.41
0.45
0
0
2
4.84
1.1
0.2
0
0.2
H-U
0
0
0
0
0
0
0
0
0
0.91
H-Sch
0
0
0
0
0
0
01
01
0.88
0
DAILY
1 3.641
2.681
0.5371
1.271
291
25.551
6.991
1351
0.881
1.51
U:IUcJobs100460\Exce6[06460-01.xIstT 2-2
Home -Other
Other -Work
Other -Other
The number of trips generated by a typical dwelling unit (single-family,
condominium, or apartment) is a function of the dwelling unit ,population, the
number of resident employed workers, and income. Table 2-3 presents example
calculations for single family dwelling units, condominiums, and apartments and
illustrates the similarity to the land use based trip generation rates used in the
previous version of NBTAM. Although the overall number of trips generated by
NBTM do not exactly match ITE land use (driveway level) trip generation, the
overall differential Is relatively small (on the order of 10-15%). The differences
are generally greatest for hon-residential land use categories that are often part
of a larger shopping center and/or are frequented by a relatively high percentage
of pass -by trips (e.g., banks, gas stations, fast.food restaurants, etc.) It may be
appropriate to modify the traffic study guidelines to provide for a separate local
access analysis to ensure that site access driveways are adequate to serve the
projected traffic volumes. The (separate) local access analysis could then be
based on ITE trip rates.
Most mode choice (e.g., transit, etc.) issues are regional in nature, superseding
cities' boundaries. For this reason, the NBTM approach is to acknowledge the
role of mode choice through data obtained from the regional mode choice model.
This data may be used directly for minor adjustments to account for future
system refinements. It is necessary to return to the regional model for evaluation
of major transit system changes. Adjustments to the NBTM are then reflected ]n
terms of zonal vehicle trip generation adjustments.
2-10
TABLE 2.3
TYPICAL DRAFT NBTM RESIDENTIAL TRIP GENERATION
EXAMPLES
-....--.......-
- ---
Daily Trip
Daily
LE
Units
Quantity
Rate
Trips
I Units
DUs
1
3.54
3.540
on
POP
3.05
0.537
1.638
W Residents
E-R
1.7
1.21
2.159
(Median Annual)
$MIL
0.07
29
2.030
TYPICAL CONDOMINIUM UNIT DAILY TRIPS
Daily Trip
Daily
VARIABLE
Units
Quantity
Rate
Trips
Dwelling Units
DUs
1
2.68
2.680
Population
POP
2.8
0.537
1.504
Employed Residents
E-R
1.6
127
2.0321
450
Income (Median Annual)
$MIL
0.05
29
TOTAL
7
TYPICALAPARTMENTUNIT DAILY TRIPS
Daily Trip
Daily
VARIABLE
Units
Quantity
Rate
Trips
Dwelling Units
DUs
1
2.68
2.680
Population
POP
2.2
0.537
1.181
Employed Residents
E-R
1.5
1.27
1.905
Income (Median Annual)
$MIL
0.04
29
TOTAL
6160
U.XUcJobs=4601Fxoe t[00460-01xls]T 2.3
2-11
2A Trip Distribution
Exhibit 2-D illustrates the NBTM trip distribution process. Separate procedures
are employed for consistent scenarios and for scenarios where the local model
deviates from the subregional model inputs and assumptions. Trip distribution is
required to match the regional model to within 10% at the Community Analysis
Area (CAA) level for consistent scenarios. The NBTM structure is based directly
on the regional trip distribution. The regional trip tables are disaggregated
directly from the regional tool for those model scenarios required to demonstrate
consistency. Alternative model scenarios are then developed from the consistent
scenarios through a factoring process (the Fratar Model) that reflects changes in
trip generation by TAZ. This approach results in substantial dependence on the
regional model. For large changes in study area land use, It will be necessary to
return to the regional model (OCTAM 3.1) and reevaluate regional changes in trip
generation, trip distribution, and mode choice. .
2.5 Time of Day Factoring
The NBTM time of day factorsaresummarized on Table 2-4. These factors have
been derived from the regional model time of day factoring procedures.
2.6 Roadway Network Representation and Traffic Assignment
The NBTM network processing procedure replicates the OCTAM 3.1 coding
conventions within the study area. The highway network is represented by roadway
links. The units per mile factor used in the network representation for NBTM is
5,280 (e.g., the coordinate system is coded in terms of feet). This is consistent with
the NAD-83 (North American Datum, 1983) system that Is used throughout the
State of California for geographic information systems (GIS) applications. Use of
this coordinate system enhances the compatibility of the NBTM tool with other GIS
databases maintained by the City of Newport Beach and other public agencies.
2-12
EXHIBIT 2-1)
DRAFT NEWPORT BEACH TRAFFIC MODEL (NBTM)
TRIP DISTRIBUTION PROCESS
TABLES BY PURPOSE
AND VEHICLE OCCUPAN
AGGREGATE PRIMARY
STUDY AREA TRIPS TO
FRATAR DISTRICTS
DISTANT
PARTS OF
TRIP TABLES
FRATAR DISTRICT PRIMARY STUDY
INTERMEDIATE AREA LOCAL DAILY
REGIONAL TRIP TABLES VEHICLE TRIP BY
PURPOSE AND
FACTORS - LOCAL
TRIPS/ REGIONAL
CALCULATE
AGGREGATI
FACTORS
APPLY FRATAR POST-FRATAR
TRIP GENERATION/ INTERMEDIATE
TRIBUTIONPROCEDURE REGIONAL TRIP TABLES
LEGEND:
INPUT/OUTPUT DECISION MODELING
DATA RULE PROCESS
TIER 1/2
INTERMEDIATE
REGIONAL TRIP TABLES
TRIPS)
OCrAM TAZ OR FRATAR
DISTRICT DISAGGREGATION
FACTORS
TRIP TABLES
TO NBTM TAZa
BY PURPOSE
PRODUCTIC
ATTRACTION FC
nla - 00
2-1 'A
TABLE 2-4
DRAFT NBTM TIME OF DAY FACTORS
HOME-
HOME-
NOME-
WORK-
OTHER-'
TIME PERIOD
DIRECTION
WORK
OTHER
SCHOOL
OTHER
OTHER
PEAK
(7:00 AM -10:00 AM)
A P
0.0301
0.0686
0.0376
0.3245
0.1098
PM PEAK
(2:45 PM - 6:45 PM)
A P
0.3814
0,4146
0.2476
0.1023
OA090
PEAK TOTAL
1.0000
1.0000
1.0000
1.0000
1.0000
MID DAY
(10:00 AM - 2:45 PM)
P-A
0.2493
03030
0.2651
0.4404
03862
q-P
0.2043
0.2376
OA073
0.4354
0.4231
NIGHTTIME
(6:45 PM - 7:00 AM)
P-A
0.2835
0.1310
0,0235
0.0679
0.0910
A-p
0.2629
0.3284
0.3041
0.0563
0.0997
OFp•PEAK,TOTAL
1.0000
1.0000
1.0000
1.0000
1.0000
u:%UcJobs100460T-xcel4GO46MI,.XlsjT 2-4
I
I
I
2-14
1 k
If
Each link is identified in terms of a unique upstream and downstream node. The
upstream node is also referred to as the ANODE, while the downstream node is
called the BNODE.
Table 2-5 summarizes the attributes that must be coded explicitly for each roadway
link in the NBTM networks. The different types of facilities included in NBTM are
shown in Table 2-6, including descriptions of their physical and usage
characteristics. The link classification code is used to further describe the
characteristics of a roadway. Roadways with the same basis cross-section
(number of through lanes and median treatment) exhibit substantial differences in
free flow speed and capacity. Factors that can influence roadway speeds and
capacities include the number of mid -block access points, signalized intersections
per mile, posted speed limit, mid -block traffic control devices such as stop signs,
etc.
The use code is intended for further use in future versions of the OCTAM
subregional travel demand model. Table 2-7 defines the use codes that are
included in the NBTM network definition. These codes have been included to
facilitate future model updates corresponding to the updated subregional model
(future OCTAM versions). They are also used in conjunction with the facility type
and area type codes to determine the speeds and capacities for freeway ramps and
freeway to freeway connectors (these two distinct types of facilities share the same
facility type code).
They are also in conjunction with the facility type and link classification -codes to
j determine the speeds and capacities for freeway ramps and freeway to freeway
connectors (these two distinct types of facilities share the same facility type code).
The corridor capacity augment code is used to provide additional capacity for the
roadways retained within the Tier 1 model coverage area. A code of 1 results in a
200% capacity increase, while a code of 2 results in a 75% capacity increase. The
2-15
TABLE 2.5
DRAFT NBTM LINK ATTRIBUTES'
�ffl�ELDNAM�E
NBTM USAGE DESCRIPTION
ODE
IdenMes the "from" node of the link
BNODE
Identifies the "to" node of the link
Assignment Group Code
Identifies the facility characteristics (i.e., freeway,
ramp, divided, undivided, etc.)(See Table 2-6).
Link Distance
Defines length of the link in miles
Link Group 1
Two digitnumber = xy; x = Link classitloation
code and =use code S00 Table 2.7 .
Link Group 2
Two digit number = xy, x = corridor capacity
augment code and = number of lanes
Link Group 3
Used in conjunction with various post -processing
utilities oni------------
1 Use of TRANPI. fields is idential to OCTAM, except the tens column of
Link Group 2 field (corridor capacity augment code).
U:1UcJobs\004601EKOOPt00460-01.xlsjT 2.5
s
1.
I
1
I
2-16
TABLE 2.6
DRAFT NBTM ROADWAY LINK FACILITY TYPE CODES'
ACILITY
CODE
FACILITY CLASS
CHARACTERISTICS
0
Toll
Pay for use facilities
1
Freewa
Limited Access
2
6 + Lane Divided
6 + Primarily serves through traffic with
limited local access
3
2-5 Lane Divided
4 lane (or less) divided serves mostly
through traffic with some local access
allowed
4
3 + Lane Undivided
Serves throu h and local traffic
5
2 Lane Undivided
Serves most) local traffic
6
Smart Street
6-8 lane divided, with possible signal
coordination, intersection capacity
im rovements and/or rade se aratlon
7
High Occupancy Vehicle HOV
Limited Access - Use by Carpools Only
8
Fwy-to-Fwy Connector or Fwy
Interface between freeways and other
lRamp
1freeways or roadwa s
9
ICentroid Connector
Zonal Access
TABLE 2-7
DRAFT NBTM USE CODES'
USE CODE
I DESCRIPTION
2
Mixed Flow Freeway, Toll Road
3
HOV 2+
4
Surface Street(Centrold Connector
5
Ramp Meter
7
HOV 3+
8
amps
9
F - FwY Interchange
' All codes are identical to OCTAM codes.
2-17
number of lane code describes the number of travel lanes (one-way) available on
each roadway link. The number of lanes directly affects roadway capacity in the
NBTM tool.
The traffic assignment procedure is dependent upon the characteristics of the
roadway system that affect travel speeds and roadway capacities. The roadway
characteristics that are of interest include the facility type, the number of lanes,
and the link classification variable. Table 2-8 and Table 2-9 summarizes the
roadway link speed and capacity characteristics, respectively.
The NBTM consistent scenarios must match regional model daily traffic volumes
at the screenline level. The OCTAM 3.1 subregionai model incorporates 4 time
periods. The NBTM traffic assignment procedure therefore also utilizes 4 time
periods per regional model procedure, with conversion to AM and PM peak hour
volumes directly from the AM and PM peak periods, respectively. The
conversion factors have been derived from recently completed projects. The
conversion factor for the 3 hour AM peak period to the AM peak hour is 0.41,
while the conversion factor for the 4 hour PM peak period to the PM peak hour is
0.30.
The assignment procedure.uses the estimated traffic volumes, along with the
free flow speeds and roadway capacities, to determine operating speeds during
each time period. An iterative process is employed that seeks to balance the
traffic among all available travelpaths between any two origin -destination points.
.Exhibit 2-E summarizes the NBTM roadway link volume/capacity ratio to travel
speed relationships that are used during the assignment process to determine
the shortest path between each origin -destination pair. The NBTM assignment
2-18
N
kD
TABLE 2-8
DRAFT NBTM ROADWAY LINK SPEED ASSUMPTIONS
FACILITY CLASS
FACILITY
CODE
USE
CODE
LINK
CLASSIFICATION
CODE
1
2
3
4
5
6
7
8
9
Toll
0
2
65
65
65
65
65
70
75
65
55
Freeway
1
2
65
65
65
65
70
65
701
601
75
6 + Lane Divided
2
4
20
35
45
40
50
50
601
30
45
Prima 2-5 Lane Divided
3
4
20
35
45
40
50
50
601
30
45
Seconds 3 + Lane Undivided
4
4
20
30
40
35
50
45
35
_ 25
40
2 Lane Undivided
5
4
20
30
40
35
45
45
65
25
40
Smart Street
6
4
35
40
50
40
45
55
65
45
45
High Occu an Vehicle HO
7j
31
651
651
65
65
70
70
70
60
70
Fwy Ram'
8
8
30
30
30
30
30
30
40
30
30
F -to-F Connector
8
9
40
45
45
451
501
50
60
501
50
Centroid Connector
9
4
20
25
30
40
40
40
40
40
40
U:\UcJobs\00460\ExceR[00460-01.xls]T 2-8
N
N
0
TABLE 2-9
DRAFT NBTM ROADWAY LINK PEAK CAPACITY ASSUMPTIONS
FACILITY
USE
LINKCLASSIFICAT
ION CODE
i 2 3 4 5 6 8 9
CAPACITY
PER
LANE
PER HOUR
FACILITY CLASS
CODE
CODE
Toll
0
2
1.950
1,950
1,950
1,950
1,950
1,950
1,950
1,950 1,950
Freeway1
2
1,950
1,950
1,950
1,950
1,950
1,950
1,950
1,950 1,950
6+Lane Divided
2
4
1,000
1,000
1,000
1,000
1,200
1,000
1,000
1,000 1,000
Prima 2-5 Lane Divided
3
4
850
850
850
850
850
1,000
850
850 850
Seconda 3 +Lane Undivided
4
4
750
750
750
750
950
750
750
750 750
2 Lane Undivided
5
4
650
650
650
650
650
1,000
650
650 650
Smart Street
6
41.200
1,200
1,200
1,200
1,200
1,2D0
1,200
1,200 1,200
Hi h Occu an Vehide HOV
7
3
1,950
i,950
1,950
1,950
1,950
1,950
. 1,950
1,950 1,950
F Ram
9
8
600
600
600
600
600
60D
600
600 600
F -10-F Connector
8
9
1,600
1,600
1,60D
1,600
1,600
1,600
i,600
1,600 1-600
CenVoid Connector
9
4
C
0
0
0
0
0
0
0 0
DRAFT NBTM ROADWAY LINK OFF-PEAK CAPACITY ASSUMPTIONS
LINK
CLASSIFICATION CODE
FACILITY
USE
1 2 3 4 5
6 8 9
CAPACITY
PER
LANE PER HOUR
FACILITY CLASS
CODE
CODE
oil0
lFreewav
+Lane Divided
Lane Divided
3 + Lane Undivided
divided
et
WhOccuan Vehide HOV
Connector
onneGor
2
3
4
5
6
7
9
8
9
2
21
4
4
4
4
4
3
8
9
4
1,95101
1,950
1,000
850
750
650
1,200
1,950
600
1,600
0
1,950
1.9501
1,000
$50
750
650
1,200
1,950
600
1,600
0
1,950
1,9501
1,000
850
750
650
1,200
1,950
600
1,600
0
1-9501
1,950
1.000
850
750
650
1,200
1,950
600
1,600
0
1,9501
1.9501
1,200
850
950
65D
1,200
1,950
60D
1,600
0
1,950
1,950
1,000
1,000
750
1,000
1,20D
1,950
600
1,600
0
1,950
1,950
1.0001
850
750
650
1,200
1,950
600
1,600
0
7,950
2,000
1,000
850
750
650
1,200
1,950
600
1,600
0
1,950
2,000
1,000
850',
750
650
1,200
1,950
600
1,600
0
u xUcJot*Am60hEx xDD4w-ol) isrr 2-9
EXHIBIT 2-E
DRAFT NBTM TRAFFIC ASSIGNMENT ALGORITHM VOLUME/
CAPACITY RATIO TO TRAVEL SPEED RELATIONSHIPS
100%
90%
80%
70%
0
UW.1 60%
60%
O
-J 40%
uj 30%
LL
O 20%
10%
IIAI■1111111101■ir��s��
0 0.6 1.0 1.5 e.0 c.o
VOLUME/CAPACITY RATIO
0.0
LEGEND.
.........-E"ANDABLEARTERULL
- BUILT -OUT ARTERIAL
- -FREEWAY
)MID -
2-21
1'
procedure also incorporates static turn movement impedances; particularly for
1
left turn movements, which are generally calibrated during the model
development process (impedances may vary by future horizon year/AM or PM
peak period). Turn prohibitors are also utilized to ensure the appropriate use of l:
freeway ramps and enforcement of other turn prohibitions (e.g., right turn only
Intersections where left turns are prohibited/impossible due to center median, ( "
etc.). j
2.7 Post Assignment Data Refinement Procedures
Although not discussed explicitly in modeling consistency guideline documents,
regional agencies responsible for travel demand modeling and traffic forecasts
are becoming increasingly aware of problems with raw (unadjusted) model
turning movement estimates. The NBTM post assignment refinement
procedures are based on processes that have been gaining widespread
acceptance among regional agencies involved in travel demand forecasting
throughout southern California. I
The goal of the future traffic volume forecast refinement or post -processing is to
utilize all available data to prepare the best possible estimate of future traffic
conditions. The recommended procedure incorporates 2001 traffic count data,
2001 model validation data (traffic estimates), and 'future (raw) model forecasts 1
(estimates) as inputs. Development of refined future daily traffic volume
forecasts is discussed first. The refined daily traffic volumes are then input to a �.
separate peak hour data refinement and turn movement estimation procedure.
Exhibit 2-F illustrates this generalized hierarchical procedure.
The future daily volume refinement process is presented on Exhibit 2-G and is I
described by the following steps:
2-22
. ' EXHIBIT 2-F
DRAFT GENERALIZED NBTM POST MODEL
REFINEMENT PROCESS
ADT9
PEAK HOUR
LINK VOLUMES
INTERSECTION
TURN MOVEMENT
VOLUMES
EXISTING
DAILY ROADWAY SEGMENT
TRAFFIC VOLUME PROCESSING
PEAK HOUR DIRECTIONAL
ROADWAY SEGMENT
TRAFFIC VOLUME
\ POSTPPROCESSING
/ INITIAL \
PROPORTION BASED
ON EXISTING TURNS OR
NCHRP-255
ITERATIVE
PROCEDURE
FUTURETURNI
MOVEMENT
FORECASTS
ESTIMATED
EXHIBIT 2-G
DRAFT DAILY ROADWAY SEGMENT
AENT PROCESS
COUNTED
VOLUME
APPROACH)
GROWTH - FUTURE
GROWTH %
ALLOWED GROWTH'
COUNT 94: MINIMUM
ALLOWED GROWTH %
ITERMEDIATE GROWTH
INITIALFUTUREGROW
YES �
FINAL REFINED VOLUMI
- EXISTING COUNT
IS
3 COUNT
G NBTM
NBTM
VOLUME
/ IS -�,
IITIAL FUTURE GROWIF
:,MINIMUM ALLOWED"
GROWTH 4
YES NO
LEGEND:
INPUT/OUTPUT DECISION MODELING
DATA RULE PROCESS
IS
I
EXISTING
NBTM
VOLUME
I
N I
PMO
PH ACH)
I
CALCULATE GROWTH RATIO - !
RE NET1&MUTING N8
M
GROWTH RATIOE INITIAL FUTURE
VOLUME -EXISTING COUNT !
�It MODEL GROWTH RATIO
INITIAL FUTURE VOLUME
CALCULATEINCM f RE
GROWTH - INITIAL FUTURE
VOLUME - EXISTING COUNTED
VOLUME I
I?_MA FUTURE GROWTH !
INTERMEDIATE GROWTH -�
MINIMUM ALLOWED GROWTH
m
REFINED'VOLUME-
ISTING COUNT+
RMEDIATE GROWTH
I
.00460:011
!4 i
1. Determine whether the existing counted volume Is greater than the existing
modeled volume (Option A) or the existing counted volume is less than the
existing modeled volume (Option B).
2A. Use the growth increment (additive) refinement process. Growth is equal to
the future model volume minus the existing model volume. The refined
future forecast is equal to the existing traffic count plus the growth.
EXAMPLE:
Existing traffic count=17,000 vehicles per day (vpd)
Existing traffic model estimate=15,000 vpd
Future traffic model raw forecasts = 24,000 vpd
I Calculated growth = 24,000—15,000 = 9,000 vpd
' Refined future daily traffic volume forecast = 17,000 + 9,000 =
26,000 vpd
2B. Use the growth ratio (multiplicative) refinement process. The forecast
refinement ratio is equal to the existing counted traffic volume divided by
the existing modeled traffic volume. The forecast refinement ratio is then
multiplied by the future model forecast to calculate the refined future daily
T traffic volume forecast.
r
r EXAMPLE:
Existing traffic count = 25,000 vehicles per day (vpd)
Existing traffic model estimate = 28,000 vpd
Future traffic model raw forecasts = 34,000 vpd
Forecast refinement ratio = 25,000 / 28,000 = 0.893
Ratio refined future forecast = 34,000 * 0.893 = 30,400 vpd
Calculated (refined) growth = 30,400 — 25,000 = 5,400 vpd
2-25
There are two minimum growth checks included in the ADT refinement process.
The first check is based on a user input segment specific allowable minimum
growth percentage. During input, the software provides for a user specified
global default that can be modified for each individual segment. For instance,
this allows the use to prohibit negative growth throughout the study area, then
specify less stringent minimum growth requirements for segments where new
parallel capacity is being provided.
The second minimum growth check is only allowed at the global level. It is
Intended to "smooth" small negative growth (typically less than 10%) that may be
caused by the modeling process and is not reasonable (e.g., no new parallel
facilities) and no decreases in surrounding development travel activity.
The resulting daily traffic volumes are suitable for reporting, analysis, and input to
the future peak hour data refinement process.. At new link locations or existing
locations where existing, counts are very low relative to future projections, raw
model assignments are utilized.
The recommended peak hour procedure incorporates existing peak hour
intersection counts, existing model validation peak hour link directional data, and
future (raw) model peak hour link directional data. The procedure uses model peak
hour inbound and outbound traffic flows by intersection leg. Model directional flows
are generally considered more reliable than the individual modeled turning
movements.
Once the raw model peak period data has been converted to raw model peak hour
volumes for each intersection leg approach and departure, the link directional peak
hour refinement ,procedure can be performed. The first two steps of the process
are identical to the calculation of refined daily traffic volume forecasts. The result of f
these first two steps for the peak hour data refinement process is referred to as the
initial model derived peak hour growth for purposes of this discussion. f
2-26
Subsequent steps (3 and on) are:
3. The initial model derived peak hour growth is compared to the daily traffic
volume growth at this point in the process. Future peak hour volumes that
are less than the existing volume are only allowed if the future daily traffic
volume is less than the existing daily traffic volume. Such "negative growth"
may be reasonable when a new transportation facility is provided which
parallels the route being analyzed. In such instances, negative growth is
'I+ allowed, consistent with the change in daily volume (e.g., if the future daily
traffic volume is 15% less than the existing daily traffic volume, up to 15%
negative growth in the final refined peak hour directional volume is allowed
as well). Otherwise, the final refined future peak hour directional volume is
adjusted to match the existing peak hour directional volume, eliminating the
negative growth that would otherwise occur.
4. The final future refined peak hour directional segment traffic volumes are
then used as inputs to the turning movement estimation algorithm. The
starting point turn proportion estimates must also be determined. Either
existing peak hour intersection turning movement counts (Option A), or
weighted exiting (leaving the intersection) volumes on intersection legs
(Option B) are used .as the basis for determining the starting point turn
proportion estimates. Existing traffic counts are used if the counts are
representative of future conditions (e.g., no new intersection legs or legs that
are underutilized under existing conditions, which render existing turn
proportions inappropriate for use under future- conditions). Otherwise, -
fi starting point turn proportions are determined as described under Step 5B.
Appendix "B" (to be provided) of this document describes the turning movement
estimation algorithm that is applied once the inputs have been determined. This
algorithm was obtained from the report Highway Traffic Data for Urbanized Area
2-27
■ s
a
Proiect Planning and Design (National Cooperative. Highway Research Program
Report 255, Transportation Research Board, 1982, pp. 105-109), commonly
referred to as NCHRP-255.
if Existing Peak Hour Data Is Available For All Intersection Legs:
i
5A. Use existing turn movement counts as the starting point for the turning
movement algorithm. Apply the NCHRP-255 algorithm. The resulting
turning movement estimates will sum to the outbound volume which was
input for each leg of the intersection. I
If Reliable Existing Turning Movement Is Not Available:
I
I
5B. Use refined model intersection leg outbound volumes to calculate the
starting point turn proportions for the turning movement estimation algorithm.
The potential through movement volume is weighted more heavily (Factor of
3.0) to more accurately -reflect actual observed traffic volume patterns.
Apply the NCHRP 255 algorithm. The resulting turning movement estimates
will again sum to the outbound volume which was input for each leg of the
Intersection.
8
2-28
TO BE
APPENDIX B
NCHRP-255 TURN MOVEMENT ESTIMATION ALGORITHM DISCUSSION
W
ITERATIVE PROCEDURES
This section contains procedures (or producing either directional or non -directional turning
volumes suing an iterative approach. Iteration involves applying a technique repeatedly until she
results converge so an acceptable result. Both procedures derive future year turning movements
from PresPecified link volumes and an initial estimate of turning Percentages. Iteration is required
to balance the volume of traffic entering and leaving the Intersection. Therefore, the number of
iterations necessary to produce an acceptable set of turning volumes is dependent on the ability of
the analyst to make reasonable a Priori estimates of turning percentages. These estimates can be
made by analyzing base year counts at the sane intersection, by reviewing turning movements at
similar Intersections, or by examining adjacent land use intensity and distribution.
Directional Volume Method
Starting with user -estimated turning percentages, the directional volume method proceeds
through an iterative computational technique to produce a (Inal set of future year turning volumes.
The computations involve alternately balancing the rows (inflows) and the columns (outflows) of a
turning movemrnl matrix until an arcepable convergsmro is ohtained. rnuurr year link volumes arc
fixed using dux mrtlnod and the turning movements are adjusted to match.
This procedure is most applicable in cases where the future year turning volume forecasts are
not expected to be radically different from either the base year conditions or from the Initial user -
supplied estimates Of turning Percentages. If large differences occur, several iterations may be
required to reach convergence to the presPecified future year link volumes. Normally, however, six
to ten iterations requiring one or two person -hours should suffice.
Basis for Development
The directional volume method is based on A basic iteration technique developed by Formes&
(30) and modified for Intersection flows by Mckky (64), A similar but more complex formulation
developed by Bacharach (7) involves input-output changes using a biProPortlonal matrix method.
Apart from these iterative techniques, there also exists a nonitersstive method for generating
Intersection directional turning movements. This method, developed by Nostrum At a). (43, 73), may
be substituted for the iterative approach fit eases where the analyst has good initial estimates of the
future year turning movements. However, the mathematical complexity of the formulation, plus the
Probability that negative numbers may result, indicates that the iterative methoo described in this
chapter will produce the most consistent results in a reasonable time frame,
Input Data Requirements
The following Input data are required,
• Future year directional link volumes.
• Either: Base year actual or assigned directional turning movements.
Initial estimate of future year directional turning percentages.
The future year link volumes are obtained directly from the computer forecasts or from the results
of a link refinement or detailing procedure (see Chapters 4 through 7). The base year data wood
preferably be actual turning movement counts, but turning data from a base year assignment could
also be used. In lieu of base year data, the analyst must make an initial estimate of future year
turning percentages based on an examination of adjacent land uses or the turning movements at
similar intersections.
Directions for Use
The directional volume method consists of five steps, as diagrammed in Figure A-60. The
following notations are used in the calcv,latfonss
is = number of links emanating from the intersection;
Onb = base year (b) inflow to the intersection on link 1 (1-1... R);
Oil = future year (0 inflow to the intersmilon on link I (i.l...n);
Djb - base year (b) outflow from the intersection on link j (1-1... n);
Djf = future year (D outflow from the Intersection on link j
Ti)b - base year (b) traffic flow entering through link i and leaving through link j;
Tijt • future year (D traffic flow entering through link f and leaving through link 1;
Pfjf r future year (D estimated Percentage (expressed in decimal torn) of traffic flow fro•n
link i to link j fine in place of T,jb); and
' = represents adjusted values in each iteration.
These notations can be illustrated using the example Intersection diagrammed in Figure A-61.
In this case, the number of links is 4 (n=4). The base year and future year inflows Oib and Oif are
shown for each link, as are the corresponding outflows Djb and Dif. The base year and future year
turning movements Tijb and Tip are diagrammed for each of the 12 movements.
It the base year turns Tyb were not known, estimated future year turn Percentages (Pelf) Could
be substituted, as illustrated for link 1. The Pill, must total to 1,00 (or 100%) for each approach.
Therefore, 1`121 • P13f s P14f = LAID and so forth for each approach link.
The computational steps are described below, followed by an example.
Step I —Construct Initial Turninit Movement Matrix
The first step involves constructing an
initial matrix of turning movements to be used in the Iterations. The construction varies depending
on whether or not base year turning volumes are available. In these and subsequent matrices, the
diagonal elements (f=j) will always be equal to zero unless U-turns are permitted.
Step IA --Base Year Turning Volumes Known. First emstrw.t a tornin
base year turning volumes (T- g movement matrix of
qb)• Next, insert the row and column totals. The'row totals should
represent inflows (O,b) and the column totals should represent outflows (Orb). Thisjis shown below.
0
to
o
aaw �•
fTir 1.
f IA
Mar star
MH tear
irolrMt Noe
H
Ctattrvtt
Initial
relrra
Available
Ar•1teDte
•�
ice Of NJ!
Ilatria
i--- --•!initial
Matti.
tOb
02b
)(_Plff(Tl2n
frl. _.
-
Art a1tt.
I fwivatrnt to
(02f)
(C2n
P12f(Tl71)
Neten
rirat tw
Mr
Iteration
trtela
At r pin I, I ---- riot iue•tlen
-
1
i
I--P)7f(T12n
mati.a
r.I.to
C. Hnl
w !aa Ca
'Srtr f.
rtttera.
Calraw ttult
lauttea Aenpesbtq
A $
riaaa t•lvan
itautloa
rr,,nt.taaae
�.—Tl2b(T72f)
--Tl)b(Tl7n �•—tOln
to
Van
Q7b
fink 7 (b7n
Timu (Tn �— '�
i.ivl 1
gift a,
At. C.I.
(T32fIT72b
rartnn
Saee•A trr
for
Iter.tloa Aeeapuuq
OSs -+
(T7)LIT71b —�
A
Iterative
rcta
rr.
(Olt)—
(T7/_7^ --
-'34b -j
r -g
F h
Ma
To Complain
Ara APr
fTt1 f
Mrtatn
felra Sotala
ab
Saw" eel.
tteratlr Artwteal.
IL.rftlw
M°
rt•c•a.w
ctarl.ta
tan
NO
tan
[batter iteration until
7
,
all aOd Ott air Attaitabl.
.�
t
�
FiEorell•6t1.
IterativepreceduretocmtputedlrectionalturninEvofomes
Fi6ure A•61. Intttseetion notation used to direttitMut iterative protedre.
Coluauts
-Out flows-
Dib Dif
x x x x
Rout Base Year
-Inflows- h x Tij Turn
(01)Oib si x Movements
x x x x
The sum of the Tills across each raw and down each column respectively. should be equal to the Oib and Oils
Finally, display the future year link volume inflows (Oil) and outflows (Dj() in parenthesesadjacent to the corresponding base yeu link volumes. Proceed to Step 2.
If base year turning volumes us not known, omit this step and proceed to Step Ia.
Step IB--Base Year Turning Volumes Not Known. Construct an initial future year turning
movement matrix using estimated turn percentages. The row totals should represent the future year
assigned inflows (Off). The column totals should represent outflows (Olt), The individual turning
movements (i.e., cells of the matrix) are calculated as follows,
Till, . Off . pijf (A-30)
Note that the Fly are applied to the inflows (Oil), not to the outflows (Djl). This cempitatirn is repeated for each cell of the matrix.
To aid in this computation, it is often helpful to construct a separate matrix of estimated
turning percentages, as shown below. Matrix multiplication can then be performed.
Column Totals
Not Usually
Equal. to 1.00
1.00 x x tt x
1.00 x x
1.00 x Pijf x
1.00 x x x x
The row totals must equal 1.00. Except by coincidence, the column totals of Pill will not equal 1,00,
At this point the inflows (Oif) are equal to the desfred future year assigned link volumes. The
adjusted outflows (UPS), rn the other hand, must be Calculated as the sum of all traffic volumes
(Tilt') for the appropriate column O). For Instance, for link 2, till • would equal the sum of all
traffic turning onto Link 2 (e.g., T121 • • T32f • a T42f• • D211• fora four -legged intersection with
no U,turns).
Djf. •
t-1 r. Till• (A-31)
Except in rare cases, this value of Oil* will not equal the desired future year outflow Dil.
Therefore, iterations will be required to enable the Djf• to converge onto the desired Off value. if
the difference between these values is acceptable to the analyst, the procedure is complete.
Typically, a difference of 2 10 percent Is considered to be acceptable. The matrix at this stage is
shown below.
Djf•
x x x x
Oif x Ttjf' x
x
x x x x
Step IB actually represents the first row iteration of the procedure, although derived in a
slightly different fashion from the ease where base year volumes are known. Therefore, if further
iterations are required, the analyst should now skip to Step 3. These relationships are shown in
Figure A-60.
Step 2—Perform the First Row iteration. Perform this step only after Step IA. In the matrix
replace the base year inflows (Olb) with the future year Inflows (Off). Then adjust each individual
turning movement according to the following,
Tijf' a (Oif/Oib) x Tlib (A• 2 7)
where Tijf• is the adjusted future volume lor this Iteration.
Construct a new matrix consisting of the Till. and Off. Now calculate the new Dll• by
summing the TOO in each column j.
N
Dif. s S. T,jl• (A-33)
The matrix at this stage Is shown below. i•t
Djf•
x x x x
Off x Tijf• x
x
x x x x
The Dlf• should be compared with the desired fill from Step IA. It the difference between
these values is acceptable to the analyst, then the procedure Is complete. Typically, a difference of
_ 10 percent is considered to be acceptable. It a larger discrepancy is apparent, then a further
iteration(,) is required.
Step 7-.perform the First Colm, Itc,jtion This step is performed on the adjusted turning
movement matrix from Steps 2 or Ill. Replace the outflows (Olf.) with the o8ginai Dip Ad),,, r-
J
114
each tndiWdual movement according to the following,
T)i{&NEW-(011/0jf')T111'OLD (A-10
WhHH
Tiff -OLD . Tilt. value in the matrix developed in Steps 2 or 101 and
Till.NEW . Adjusted Tilt alter column Iteration.
In subsequent Itcratims, Tip -NEW becomes Ti)POLO and so forth.
Construct a new matrix CWW$ting of the TilfaNEW and Oil. Calculate the adjusted Oil' by
summing the Till -NEW in each row.
Cite - t Tilt -NEW [A-55)
Jet it the dilterrmce between these rallies Is
The Olt& should be compared with the original Olt-
acceptable to the analyst, the procedure is eomplete. TYpically, a value of S t0 percent Is
considered to be aaeptable. It a larger dlscrepanry Is apparent, continue with a further
lteratlonb).
Step 4—Repn� t R�Itetation, it needed, repeat the Step 2 procedure for row Iterations.
Calculate new values for Tilt -NEW and Olt-. Compare D)t- with Dj f.
Steps—Rep&at Column Iteratlen. If needed after Step ►, repeat the Step i column Iteration
procedre. Calculate new values for TI)IaNEW and Cite. Compare oil' with Olt,
Iterations 09"
Continued unt"
f- and Dil-
The row and eTilt* •al,es In the 0naa Iteration maid, acceptable resent the final arc abadjusteobtained. The
directional luring WA though movement. The T{)1' sta.tid be closely reviewed for reasonableness
before wing them In subsequent planing and design studies.
Example
Step 1A A tour -link IntessectIon has bee year turning movements and future year lirk
whines as illustrated In Figure A-62 and displayed in hnatrix form in Figure-A-65. For this
esampte. Step IO is not used, and she analysis moves w Step 2.
Stg2_2t First Row tteratlee Mg. A-64)
Step k First Column iteration (Fig. A-65)
In this example, the Eitlerences in row touts are within 5 percent after the second iteration.
It this difference is acceptable, select th& Tilt'NEW from Step 1 as the final turning nnove,nent
matrix, and subsequent Iterations will not be required.
For comparison. Site six Iterations, the remits in Figure A-66 could be obuined. Therefore.
the additional iterations have reduced the differences twttter still.
s,
S
n n IS
200.0ase Year Value
(250)-lutuce Year Value
$ o
N
t--00
ti n
J L
-100 MM)
(coo) 470 1
e00
(250) 200 40 t
1 250 (10M
r
110 --
1
50-3
Roo
rMm
t
Fngwe A-62. Example of directleml Intersection volumes.
(350D)
(, o)
(300) (�D,,i�Flr.ui
50
10
(Soo)0
so
K0
ZOO
(460)60
JF2oo)
o
too
140 rj�
(2Sot0
40
0
50
so
L70
250
In(Iws
P,gwe A•67. intersection volumes displayed in mama tor.nat.
F i • •�� r • F . r e .. r r • • . • � r
I�
• D- �
335 510D4f O3 r•t Jf
Soo
0
too /so
250
450
q0
0 150
210
Ttj(o
01i 250
139
50' 0
42
O�Xf'f
J
Boo
107
3co 333
0
Olb
con,pare
D�f'
z D;f
J � t
335
300
+12 4.
J•2
510
Soo
+z%
J.3
0S
400
a7.
3 . 4
521
400
- 137•
Total
200o
2010 J
1',gure A•64. First sow iteration.
D;f
�1o9E
JZ44121
10142Z410'f
41 O
129c
353 314,
O
D;f�
i•1
52.) Soo
457.
i•z
444 450
+3%
i.3
2-44 250
-Z%
i•4
1cs Soo
-4�
20
2000 D0 r/
Figure A.65. First column iteration.
D;f
0o
so0
coo
coo
502
0
84
137.
281
453
16
0
135
242
T • ��
Jf
a
0if 250
124
41
0
11
Di_f X �Jf+MO
-,IS
100
3c2
333
D
D;f•
C"arc
af, ^ Dif
d
i.l
502
500
0%
i•2
453
450
•1%
i•3
250
250
0%
i.4
195
ODD_
-1%
2-000
2000 ✓
Figure A=66. Intersection volumes alter 6 iterations.