2015 Travel Demand Model Documentation TAC Approval August 22, 2016
Approved BTPO Policy Board 09,12,2016
TABLE OF CONTENTS
1 Introduction ...... 1 2 Street Network and Traffic Analysis Zone (TAZ) ...... 1 2.1 Street Network Development ...... 2 2.2 Traffic Analysis Zone ...... 6 3 Initialization ...... 6 4 Network Skimming ...... 12 5 Household Disaggregation ...... 12 6 Trip Generation ...... 15 7 Trip Distribution ...... 17 8 Mode Split and Time of Day ...... 19 9 Highway Assignment ...... 25 10 Model Estimation and Calibration ...... 27 1) Household Disaggregation Model...... 27 2) Trip Generation ...... 28 3) Trip Distribution ...... 30 4) Mode Split ...... 33 5) Time of Day Factors ...... 36 6) External to External Trip Distribution ...... 40 11 Model Validation ...... 41 11.1 Network Zones ...... 41 11.2 Socioeconomic Data ...... 41 11.3 Trip Generation ...... 41 11.4 Trip Distribution ...... 42 11.5 Mode Split, Vehicle Occupancy, and Commercial and Other Trips ...... 45 11.6 Network Assignment ...... 46 11.6.1 Daily Traffic Assignment ...... 47 11.6.2 PM Peak Hour Traffic Assignment ...... 49 11.6.3 Mid-Day Peak Hour Traffic Assignment ...... 50 11.6.4 Assessment ...... 51 12 Future Year Modeling or Scenario Modeling ...... 51 12.1 Required File Changes ...... 51
Page | i
12.2 Potential Change for Scenario Modeling ...... 52 Appendix A. Acronyms ...... A-1 Appendix B. Model Structure and Contents ...... B-1
List of Tables
Table 1- Street Network link Attributes ...... 3 Table 2- Input and Output Table of the Initialization Stage ...... 8 Table 3- List of Attributes in the Cross_FT_AT.bin FIle ...... 8 Table 4 - Area Type Definition ...... 8 Table 5 - Cross_FT_AT.bin File ...... 9 Table 6- List of Attributes in the Turn Penalty.dbf File ...... 9 Table 7- List of Attributes Appended to the HighwayNetwrok.dbd File ...... 10 Table 8- List of Input and Output Files to Network Skimming Stage ...... 12 Table 9- List of Input and Outputs Files to the Household Sub Model ...... 13 Table 10- List of Attributes on TAZDATA_2015.bin File ...... 13 Table 11- Percent of one, two, three, and four or more person households based on TAZ Average Household Size ...... 13 Table 12- Percentage of zero, one, two, or three Worker Households Based on Household Size ...... 14 Table 13 - List of Attributes added to the TAZDATA_2015.bin File ...... 15 Table 14- List of Input and Output Files in Trip Generation ...... 16 Table 15- List of Attributes in triprate.csv Table ...... 16 Table 16- Trips Rates by the Number of Persons and Workers ...... 16 Table 17- List of Attributes in tripsBYPurpose.bin File ...... 17 Table 18- List of Input and Output Files to Trip Distribution ...... 18 Table 19- List of Input and Output Files to Mode Split Stage ...... 20 Table 20- List of Attributes in Mode Split (modesplit.csv) ...... 22 Table 21- Model Split ...... 22 Table 22- List of Attributes in Time of Day (TOD_Factors.csv) ...... 24 Table 23- List of Attributes to Vehicle Occupancy (VECOCC_factors.csv) and Factors ...... 24 Table 24- List of Attributes in External Average Daily Traffic (External_ADT..bin) ...... 24 Table 25- List of Calculated Attributes Added to External_ADT.bin File ...... 24 Table 26- List of Attributes and Output Files to Highway Assignment ...... 26 Table 27- Parameters Used in the Highway Assignment ...... 26 Table 28- Highway Assignment Calculated Fields ...... 27 Table 29- Percent of Workers per Household Size Lookup Table ...... 28 Table 30- Home Based Work Trip Rates ...... 29 Table 31- Home Based College Trip Rates ...... 29 Table 32- Home Based School Trip Rates ...... 29 Table 33- Home Based Shopping Trip Rates ...... 30 Table 34- Home Based Other Trip Rates ...... 30 Table 35- Non Home Based Trip Rates ...... 30 Table 36- Destination Choice Model Utilities by Purpose ...... 31
Page | ii
Table 37: Destination Choice Estimated Variable Coefficients ...... 32 Table 38: HBW Auto Share ...... 33 Table 39: HBC Auto Share ...... 33 Table 40: SCH Auto Share ...... 34 Table 41: HBS Auto Share ...... 34 Table 42: HBO Auto Share ...... 35 Table 43: NHB Auto Share ...... 35 Table 44- Time of Day Factors by Mode ...... 40 Table 45: Directional Distribution Factors1 ...... 40 Table 46: Bluetooth External Survey Distribution ...... 40 Table 47: Trip Generation Rate Check ...... 42 Table 48: Trip Generation Adjustment Factor ...... 42 Table 49-Total Trips Check ...... 46 Table 50: Zero Flow Control Delay Adjustment Factors ...... 46 Table 51 - Daily Observed versus Estimated Link Volumes by Facility Type ...... 47 Table 52 - Daily VMT by Facility Type ...... 49 Table 53- PM Peak Hour VMT by Facility Type ...... 50 Table 54- Mid-Day Peak Hour Observed vs Estimated VMT by Facility Type ...... 50
List of Figures
Figure 1 Model Extents ...... 2 Figure 2- Street Network Facility Types ...... 5 Figure 3- Traffic Analysis Zones ...... 7 Figure 4- Number of Lanes, Speed Limit, and Area Type for the Highway Network ...... 11 Figure 5- Sample Contents of dcCoefficient.txt File ...... 18 Figure 6- Output of Trace.txt file ...... 19 Figure 7- Districts ...... 21 Figure 8- External Station Locations ...... 25 Figure 9- Bannock County Households by Household Size ...... 28 Figure 10: All Trip Purposes Combined Trips in Motion Diurnal Distribution ...... 36 Figure 11: HBW Trips in Motion Diurnal Distribution ...... 37 Figure 12: HBC Trips in Motion Diurnal Distribution ...... 37 Figure 13: SCH Trips in Motion Diurnal Distribution ...... 38 Figure 14: HBS Trips in Motion Diurnal Distribution ...... 38 Figure 15: HBO Trips in Motion Diurnal Distribution ...... 39 Figure 16: NHB Trips in Motion Diurnal Distribution ...... 39 Figure 17- Home Base Work Trip Length Frequency ...... 43 Figure 18-Home Base Other Trip Length Frequency...... 43 Figure 19: Home Base Shopping Trip Length Frequency ...... 44 Figure 20: School Trip Length Frequency ...... 44 Figure 21: Home Base College Trip Length Frequency ...... 45 Figure 22: Not Home Base Trip Length Frequency ...... 45 Figure 23- Daily Observed versus Estimated Link Volume...... 47
Page | iii
Figure 24: Daily Assigned Volume ...... 48 Figure 25- PM Peak Hour Volume vs Count- ...... 49 Figure 26 - Mid Day Peak Hour Observed vs Estimated Link Volume ...... 50 Figure 27: Model Structure ...... B-1 Figure 28: Model Code ...... B-2 Figure 29: Bannock.ini File ...... B-2 Figure 30: Bannock_mod.bin File...... B-3 Figure 31: Bannock Toolbox Setup: Creating a UI file ...... B-5 Figure 32: Bannock Toolbox Setup: Creating a UI file ...... B-6 Figure 33: Bannock Toolbox Setup: Setting up Scenarios ...... B-8
Page | iv 1 INTRODUCTION
The purpose of this report is to document the development and validation of a travel demand model (TDM) for the Bannock Transportation Planning Organization (BTPO). The TDM was developed using the TransCAD transportation forecasting software (version 7.0 build 12190). Figure 1.1 displays the model extents. The TDM covers the Portneuf Valley Non-Attainment Area (PVNAA) and including areas in the adjacent vicinity of the PVNAA.
The travel demand models try to reproduce the traffic conditions in a given area based on the road network, population data, employment data, and information on driving characters of that populations. Travel demand models once calibrated and validated can use future road networks, population, and employment data to predict the travel on a specific roadway or the entire network.
The BTPO 2015 Model updates and recalibrates the 2011 travel demand model calibration to a base year of 2015. The calibration of the 2011 model was completed using a household survey of the PVNAA. This report will present many of the calibration inputs, but the basic assumption is the travel demand model is calibrated. Calibration is the process of adjusting the parameters and constants within the various steps of the model to observed data. The calibration data for the model primarily comes from the household survey.
The primary focus of this report is the validation of that 2011 model against 2015 demographic and traffic conditions. Validation was comparing the actual traffic in 2014- 2015 to the estimated data provided by the TDM.
The remaining sections of this report will provide information on seven steps involved in the BTPO TDM. For each step, the inputs, outputs, and information about the processed or model used by the TransCAD software will be described. Appendix B provides a guide to setting up and running the TDM.
The socioeconomic data used in TDM development has the base year of 2015. BTPO 2015 Demographic Update Report details the socioeconomic data, sources, and methodology.
2 STREET NETWORK AND TRAFFIC ANALYSIS ZONE (TAZ)
The street network and the traffic analysis zones are the critical frameworks for the TDM. These two basic frameworks are used to processes trip generation and traffic assignment. The seven model steps use the various data contained in these geographic information system (GIS) based layers. The GIS platform in TransCAD allows modeling of actual conditions and locations. The issues with using the TransCAD model come mainly in the form of compatibility with other GIS platforms used by member agencies namely ARCGIS. The biggest issue in naming attribute fields is ARCGIS’s limitation of a ten-digit field name.
P a g e | 1
Figure 1 Model Extents
2.1 STREET NETWORK DEVELOPMENT BTPO’s street centerline layer was updated too and was used as the 2015 street network. This centerline file represents a snapshot in time what the street network looked like in 2015. Many travel demand models only model those streets classified as a collector or arterial. BTPO’s model includes all public streets within the modeling extents. Local streets are included because the area is relatively small and compact so the location of centroid connectors could alter the traffic flow.
BTPO is in an air quality non-attainment area where the emissions for vehicles on all public streets needs to be determined. Local street traffic volumes are not included in the validation, but Vehicle Miles Traveled VMT will be added to the VMT on centroid connectors to get an estimate of local street
P a g e | 2
traffic. Centroid connectors are a special street type which is added to the public road network and is used to connect the demographic information located in the TAZ layer to the street system. Centroid represents the "center of activity" of a TAZ and is the specific location of the origin and destination for all trips to and from the TAZ. Centroids connectors, as the name suggest connect these centers to the street network. Centroid connectors are not a part of the street network but as indicated previously are included in local VMT.
Travel Demand Models, likes the BTPO model do not follow the traditional GIS convention for public roads. Nodes matter in TDM. The start point or end point of a street segment is called a node. Nodes can have data or information attached such as does this node have a stop sign, traffic signal, or roundabout associated with it. Nodes are also used to link data like trips to the street network. Another difference is directionality. All segments in the network have an AB and BA direction this allows modeling traffic in both travel directions.
Table 1 shows the street network segment attributes. These are the fields associated with the street network.
Table 1- Street Network link Attributes
Field Name Description ID Identification number assigned by software Dir 0 = two way travel; 1 = one-way travel is AB direction; -1=one-way travel in BA direction Length Length of the street segment Prefix Prefix item of the Street name (i.e. N, S, E, W) County The Count the link is located in. CName Address Match Name FName Street Name FDSUF Suffix item of a street FTYPE Street Type (ex. Ln, Dr, Ave) Number Interstate Number (15 or 86) US_route US highway route number 91 or 30 Parity Parity is location of address 0 = address on both side of street 1 or -1 is all address on right or left side
Start Left Street address left beginning Start Right Street Address Right Beginning End Left Street Address Left End End Right Street Address Right End Left ZIP ZIP code left side of street Right ZIP ZIP code right side of street ROADWIDTH Road width ROW Right of Way of the street segment Lane_Quan_AB Number of through lanes AB
P a g e | 3
Field Name Description Lane_Quan_BA Number of through lanes BA URBAN_AREA I if the Street Segment is located within the Pocatello/Chubbuck Urban Area Ln_type Road functional classification Values 1-10. 1=Rural Principal Arterial; 2=Urban Principal Arterial; 3=Rural Interstate; 4=Urban Interstate; 5=Rural Minor Arterial; 6=Urban Minor Arterial; 7 Centroid Connector; 8=Collector; 9=Local; 10= Interstate Ramp
AreaType 1 = Rural, 2 = Urban, 3= CBD FacilityType Centroid Connector = 0, Local=1, Collector = 2, Minor Arterial = 3, Principal Arterial = 4, Ramp = 5, Interstate = 6|
Functional Class Functional Classification of the street segment ONEWAY_DIR Direction of owe-way road SPEEDLIMIT Posted speed limit on the link (This field can be deleted)?
PostedSpeed_AB Posted speed limit in AB direction PostedSpeed_BA Posted speed limit in AB direction PVNAA 1 if link is within the PVNAA Lane Miles The length of the street segment times the number of lanes in each direction. Agency Who owns the street segment 1= Pocatello, 2 = Chubbuck, 3= Bannock County, 4= ITD CCSTYLE The street style number used by TransCAD to assigned a line style
Figure 2 show the BTPO model extents with the facility type for each roadway within the model. This facility type designations are critical in the remaining model setup.
P a g e | 4
86
Facility Type Centroid Local Collector Minor Arterial Principal Arterial 15 Interstate Ramp Interstate 0 1 2 3 Miles
Figure 2- Street Network Facility Types
P a g e | 5
2.2 TRAFFIC ANALYSIS ZONE Traffic Analysis Zones are polygon representations of areas which have similar trip characteristics. The BTPO model has 624 TAZs ranging in size from .01 to 4.0 square miles (Figure 3). 2010 United States Census block information are the basis for the TAZ borders. Most TAZs include several blocks, but sometimes block was split. That occurred when a block had a large area which did not have the same traffic patterns. When that occurred the population and housing units were split proportionally. When housing in a large area TAZ was located in one section of the TAZ, the split occurred to separate the vacant land from the developed land.
The TAZ and street network layers are critical to the modeling process. All other traffic and demographic data are either joined to these layers or calculated from the fields within or joined to the layer.
3 INITIALIZATION
This first stage of the model builds the highway network file which the steps in the modeling process build upon. Table Why are the field names so funny 2 shows the list of input and output files of this stage. The looking? TransCAD allows for long highway database and highway network file have bi- field names but ARCGIS only allows directional fields such as Link Delay, travel time, hourly 10-digits which limited to naming capacity, capacity, and calibration parameter fields. convention. The shorted names are used in the document because they An excel spreadsheet listing the facility type, lanes, and are used in the model. Some are volume-delay function (Cross_FT_AT.bin) provides the lanes more than 10-digits but most are capacity and the calibration parameters. Table 3, 4, and 5 kept to that limit. A description of describe the attributes of this table and the values which are these names is provided in table added to each segment in Table 1 based on the Area Type format. and Facility type of that segment.
Turn penalty database (Table 6) allows the model to identify any turn restrictions within the network. The turn penalties in the BTPO TDM are related to free-flow right turn lanes.
The last added item to the output HighwayNetwork.dbd is a table which the traffic counts from 2015 which will be used in the validation stage of the process. Table 7 describes the attributes appended to the highway network (Table 1).
Figure 4 displays the number of lanes in both the AB and BA direction along with the posted speed limit and Area Types for each street segment identified as a collector, arterial, or interstate.
P a g e | 6
Figure 3- Traffic Analysis Zones
P a g e | 7
Table 2- Input and Output Table of the Initialization Stage
Input/Output File Name Description input Input\HighwayNetwork_
Table 3- List of Attributes in the Cross_FT_AT.bin File
Field Description Example ATFTLookup Area Type * 100 + Facility Type 102 AreaType 1 = Rural, 2 = Urban, and 3 = CBD 1 FacilityType 0 = Centroid Connector, 1 = Local, 2 = Collector , 3 = 2 Minor Arterial, 4 = Principal Arterial, 5 = Ramp and 6 = Interstate Capacity Capacity for this area type 450 Jparameter Volume delay function calibration parameter (Akcelik J 0.08 Parameter) LinkDelay Percent of link control delay (intersection delay + other 0.05 delay)
Table 4 - Area Type Definition
Area Type Density per Square Mile Area Type Description 1 Below 1,000 Rural 2 1,000 - 7,000 Urban 3 Above 7000 CBD
P a g e | 8
Table 5 - Cross_FT_AT.bin File
ATFTLookup AreaType FacilityType Capacity JParameter LinkDelay 300 3 0 9999 0 0 301 3 1 450 1 0.05 302 3 2 500 0.8 0.1 303 3 3 625 0.6 0.04 304 3 4 725 0.4 0.07 305 3 5 850 0.1 0.2 306 3 6 850 0.1 0.09 200 2 0 9999 0 0 201 2 1 500 1 0.05 202 2 2 525 0.8 0.05 203 2 3 700 0.6 0.08 204 2 4 825 0.4 0.06 205 2 5 850 0.1 0.09 206 2 6 850 0.1 0.09 100 1 0 9999 0 0 101 1 1 625 1 0.05 102 1 2 650 0.8 0.05 103 1 3 750 0.6 0.05 104 1 4 850 0.4 0.1 105 1 5 950 0.1 0.15 106 1 6 1250 0.1 0
Table 6- List of Attributes in the Turn Penalty.dbf File
FROM_ID TO_ID PENALTY INTERS_ID LEFT RIGHT STRAIGHT UTURN 579 581 -- 3198 0 1 0 0 581 579 -- 3198 1 0 0 0 1490 1483 -- 4647 0 1 0 0 1484 1493 -- 4466 0 1 0 0 1495 1494 -- 3143 0 1 0 0 4619 1505 -- 1223 0 1 0 0 1784 4619 -- 1223 0 1 0 0 3463 3466 -- 2569 0 1 0 0 1696 1517 -- 1238 0 1 0 0 1695 1674 -- 1239 0 1 0 0 1721 1699 -- 1335 0 1 0 0 1722 1721 -- 3050 0 1 0 0 3110 3109 -- 3050 0 1 0 0 3108 3084 -- 2318 0 1 0 0
P a g e | 9
FROM_ID TO_ID PENALTY INTERS_ID LEFT RIGHT STRAIGHT UTURN 3523 3521 -- 2608 0 1 0 0 1150 4614 -- 886 0 1 0 0 1152 1148 -- 885 0 0 0 1 1353 1351 -- 1099 0 1 0 0 206 325 -- 205 0 1 0 0 923 924 -- 757 0 0 0 1 1160 1161 -- 892 0 0 0 1 1639 1633 -- 2876 0 0 0 1 2379 2375 -- 4121 0 1 0 0
Table 7- List of Attributes Appended to the HighwayNetwrok.dbd File
Field_Name Description Total_15 Average Annual Traffic Volume 24-Hour Both Directions AB_15 Average Annual Traffic Volume 24-Hour AB Direction BA_15 Average Annual Traffic Volume 24-Hour BA Direction AM_Total_15 Average Annual Traffic Volume AM Peak (07:00 - 08:00) Both Directions AM_AB_15 Average Annual Traffic Volume AM Peak (07:00 - 08:00) AB Direction AM_BA_15 Average Annual Traffic Volume AM Peak (07:00 - 08:00) BA Direction Noon_Total_15 Average Annual Traffic Volume Noon Peak (12:00 - 13:00) Both Directions Noon_AB_15 Average Annual Traffic Volume Noon Peak (12:00 - 13:00) AB Direction Noon_BA_15 Average Annual Traffic Volume Noon Peak (12:00 - 13:00) BA Direction PM_Total_15 Average Annual Traffic Volume PM Peak (17:00 - 18:00) Both Directions PM_AB_15 Average Annual Traffic Volume PM Peak (17:00 - 18:00) AB Direction PM_BA_15 Average Annual Traffic Volume PM Peak (17:00 - 18:00) BA Direction VMT_15 Vehicle Miles Traveled computed from Length and Total_15 fields AB_FFTime Free flow time in the AB Direction computed from Length and PostedSpeed_AB fields BA_FFTime Free flow time in the BA Direction computed from Length and PostedSpeed_BA fields ATFT Area Type * 100+ Facility Type AB_Hourly_Capacity Hourly capacity in the AB direction BA_Hourly_Capacity Hourly capacity in the BA direction AB_Daily_Capacity Daily capacity in the AB direction BA_Daily_Capacity Daily capacity in the BA direction VDFParameter Volume delay function parameter AB_ZeroFlowDelay Link control delay * Free flow time in AB direction BA_ZeroFlowDelay Link control delay * Free flow time in AB direction
P a g e | 10
Figure 4- Number of Lanes, Speed Limit, and Area Type for the Highway Network
P a g e | 11
4 NETWORK SKIMMING
Zone to zone travel time and distance skims are built at this stage. Table 8 lists input and output files used in this stage. First, the zone centroid nodes are identified in the highway network node layer based on the CENT field (CENT = 0) and then the shortest paths between these node pairs are built based on highway link travel times. Free flow link travel times are used for the first iteration (before feedback loop) and MSA averaged link travel times are used for the feedback loop iterations. Distance and time skim matrices are generated for each OD pair based on the link travel times. In the second step, intrazonal travel times are computed based on the average travel time to the three closest zones. In the third step, the destination zone terminal times are added to the travel times. In the fourth step, internal and external matrix indices are added to identify rows and columns in the matrices. Terminal travel time is the time it takes a person to get from their residents to their car at the origin and the time it takes to get from the car to the destination for the destination end.
Table 8- List of Input and Output Files to Network Skimming Stage
Input/Output File Name Description input Input\terminal times.mxt Terminal Time Matrix input Output\HighwayNetwork.dbd Highway Database input Output\network.net Highway Network Output Output\spmat.mat Travel Time and Distance Matrices
5 HOUSEHOLD DISAGGREGATION
The household survey conducted in 2011 determined the trip rate (number of trips each household takes) by the size of the household and the number of workers. This stage the model disaggregates households to households by size and the number of workers. Table 9 lists inputs and outputs to this stage.
First, the model computes the average zonal household size (total persons/ total households) which are located in the TAZDATA_2015.bin (Table 10) and then based on that size lookups up in hhsize_lookup.csv the percentage of one, two, three, or four or more person household (Table 11). Next, the model allocates the number of workers based on the household size. Table 12 shows the percentage of households with zero, one, two, or three or more workers based on the size of household.
Table 13 shows the list of attributes for the number of persons which live in various sized households within each TAZ. Table 13 is joined to Table 10 to create a complete demographic data file. Trip productions are generated based on the trip rates by households by size and workers, which are described in detail in the next section.
P a g e | 12
Table 9- List of Input and Outputs Files to the Household Sub Model
Input/Output File Name Description input Input\TAZDATA_2015.bin Zonal land-use data such as households and employment input Input\hhsize_lookup.csv Household size lookup table input Input\hhworker_lookup.csv Household worker lookup table Output Output\tazData_2015.bin Zonal land-use data with trip generation rates
Table 10- List of Attributes on TAZDATA_2015.bin File
Field Description TAZ Traffic Analysis Zone number Persons Number of persons in the zone Households Number of households in the zone Retail Retail employment in the zone (NAICS codes: 44,45) Education Education employment in the zone (NAICS code: 61) Other Other employment in the zone (NAICS codes: 11,21,22,23,31,42,48,49,51, 53,55,56,92) Service Service employment in the zone (NAICS codes: 52,54,62,71,72,81) University University employment in the zone total_emp Total employment in the zone District District number
Table 11- Percent of one, two, three, and four or more person households based on TAZ Average Household Size
Average_Size One_Person Two_Person Three_Person Four_Person 0 0 0 0 0 1 1 0 0 0 1.1 0.8975 0.0689 0.0119 0.0216 1.2 0.7944 0.1487 0.0245 0.0325 1.3 0.6942 0.2254 0.037 0.0434 1.4 0.6038 0.2922 0.0496 0.0544 1.5 0.5309 0.3414 0.0621 0.0655 1.6 0.479 0.3695 0.0746 0.0769 1.7 0.445 0.3795 0.0868 0.0887 1.8 0.4217 0.3785 0.0988 0.101 1.9 0.4019 0.3739 0.1103 0.1139 2 0.3813 0.37 0.1211 0.1277 2.1 0.3589 0.3677 0.131 0.1423 2.2 0.3354 0.3666 0.14 0.1581
P a g e | 13
2.3 0.3118 0.3654 0.1478 0.175 2.4 0.2893 0.3633 0.1543 0.1931 2.5 0.2689 0.3591 0.1594 0.2126 2.6 0.2512 0.3523 0.1631 0.2333 2.7 0.2356 0.3437 0.1654 0.2553 2.8 0.2188 0.3365 0.1663 0.2784 2.9 0.1976 0.3342 0.1659 0.3024 3 0.1753 0.3335 0.1642 0.327 3.1 0.1553 0.3315 0.1613 0.3519 3.2 0.1384 0.3277 0.1573 0.3766 3.3 0.124 0.3229 0.1523 0.4008 3.4 0.1113 0.3181 0.1465 0.4242 3.5 0.0996 0.3138 0.1402 0.4464 3.6 0.0888 0.31 0.1338 0.4673 3.7 0.0791 0.3066 0.1275 0.4868 3.8 0.0703 0.3035 0.1214 0.5048 3.9 0.0626 0.3005 0.1155 0.5214 4 0.056 0.2975 0.1098 0.5367 4.1 0.0503 0.2945 0.1043 0.5509 4.2 0.0455 0.2913 0.099 0.5642 4.3 0.0416 0.2878 0.0936 0.577 4.4 0.0383 0.284 0.0883 0.5894 4.5 0.0353 0.28 0.083 0.6016
Table 12- Percentage of zero, one, two, or three Worker Households Based on Household Size
Person W0 W1 W2 W3 1 0.484 0.516 0 0 2 0.308 0.3 0.393 0 3 0.082 0.365 0.416 0.137 4 0.045 0.281 0.469 0.205
P a g e | 14
Table 13 - List of Attributes added to the TAZDATA_2015.bin File
Field Description Expression Per1Wrk0 One Person Zero Worker Households Households * Person 1 * Worker 0 Per1Wrk1 One Person One Worker Households Households * Person 1 * Worker 1 Per1Wrk2 One Person Two Worker Households Households * Person 1 * Worker 2 Per1Wrk3 One Person Three or more Worker Households Households * Person 1 * Worker 3 Per2Wrk0 Two Person Zero Worker Households Households * Person 2 * Worker 0 Per2Wrk1 Two Person One Worker Households Households * Person 2 * Worker 1 Per2Wrk2 Two Person Two Worker Households Households * Person 2 * Worker 2 Per2Wrk3 Two Person Three or more Worker Households Households * Person 2 * Worker 3 Per3Wrk0 Three Person Zero Worker Households Households * Person 3 * Worker 0 Per3Wrk1 Three Person One Worker Households Households * Person 3 * Worker 1 Per3Wrk2 Three Person Two Worker Households Households * Person 3 * Worker 2 Per3Wrk3 Three Person Three or more Worker Households Households * Person 3 * Worker 3 Per4Wrk0 Four or more Person Zero Worker Households Households * Person 4 * Worker 0 Per4Wrk1 Four or more Person One Worker Households Households * Person 4 * Worker 1 Per4Wrk2 Four or more Person Two Worker Households Households * Person 4 * Worker 2 Per4Wrk3 Four or more Person Three or more Worker Households Households * Person 4 * Worker 3
6 TRIP GENERATION
Person trip productions are generated based on the person trip rates and households by size and workers and then aggregated by trip purpose. Table 14 shows the list of input and output files for this stage. Table 15 and Table 16 show attributes of the trip rate table and actual trip rates by class and trip type. Note that NHB trips are generated at the household home zone, aggregated to the region level, and then re-distributed to production zones in the trip distribution step before being distributed to destination zones. Table 17 shows the attributes for the tripsBYPutpose.bin table is the aggregation of trips for each household class.
P a g e | 15
Table 14- List of Input and Output Files in Trip Generation
Input/Output File Name Description input Output\tazData_2010.bin Zonal land-use data appended with trip generation rates input Input\triprate.csv Trip rates by purpose, size and worker class Output Output\trips.bin Person trip productions by purpose, size, and worker class Output Output\tripsBYPurpose.bin Aggregated person trip productions by purpose
Table 15- List of Attributes in triprate.csv Table
Field Description Class household by persons and worker class label HBW Home based work trip rates HBO Home based other trip rates HBS Home based shop trip rates SCH Home based school trip rates HBC Home based college/ university trip rates NHB Non-home-based trip rates
Table 16- Trips Rates by the Number of Persons and Workers
Purpose HBW HBO HBS SCH HBC NHB Per1Wrk0 0 2.1 0.81 0 0.01 1.19 Per1Wrk1 1.25 1.49 0.34 0 0 1.34 Per1Wrk2 0 0 0 0 0 0 Per1Wrk3 0 0 0 0 0 0 Per2Wrk0 0 4.42 1.25 0.14 0.04 2.37 Per2Wrk1 1.55 3.07 0.84 0.06 0.12 2.69 Per2Wrk2 2.74 2.1 0.83 0 0.07 2.86 Per2Wrk3 0 0 0 0 0 0 Per3Wrk0 0 6.02 1.61 0.41 0.13 6.01 Per3Wrk1 1.47 4.31 1.74 0.48 0.53 2.66 Per3Wrk2 2.46 5.03 0.76 1.72 0.15 3.75 Per3Wrk3 3.69 3.84 1.1 0.36 0.28 7.52 Per4Wrk0 0 9.1 3.08 4.67 0.82 4.54 Per4Wrk1 1.15 12.5 1.82 6.13 0.28 5.6 Per4Wrk2 2.29 8.5 1.2 5.52 0.41 5.98 Per4Wrk3 3.74 11.78 1.3 3.65 1.35 8.12
P a g e | 16
Table 17- List of Attributes in tripsBYPurpose.bin File
Field Description TAZ Traffic Analysis Zone number HBW Home based work trips HBO Home based other trips HBS Home based shop trips SCH Home based school trips HBC Home based college/ university trips NHB Non-home-based trips
7 TRIP DISTRIBUTION
A destination choice model is applied to distribute trips from productions to attractions. In this stage, the model distributes zonal aggregate trip productions by purpose to all zones based on the travel time between the zones and the destination zone size term (which is a measure of the activity in the zones). Table 18 shows the list of input and output files used in this step. When the model is run with feedback, the same destination choice model is called for each loop but uses updated network travel times to distribute the trips. Only the prefix skim file name is used as input, and the full extension gets appended on-the-fly based on the current loop number.
Since the production end of the NHB trip productions are not generated at the home zone, the zonal NHB trip productions are aggregated to the region and then redistributed to zones based on a simple model of the zonal share of household and employment. The destination choice model is then applied to choose destinations for NHB productions. The following equations were estimated from the survey and is used in the NHB production model:
푁퐻퐵_푆𝑖푧푒푖 = log( 0.261 ∗ 퐻퐻푖 + 1.000 ∗ 푅푒푡푎𝑖푙푖 + .0965 ∗ 푆푒푟푣𝑖푐푒푖 + 0.015 ∗ 푂푡ℎ푒푟푖 + 1.240 ∗ 퐸푑푢푐푎푡𝑖표푛푖)
푈푡𝑖푙𝑖푡푦푖 = exp (푁퐻퐵_푆𝑖푧푒푖)
푃푟표푏푎푏𝑖푙𝑖푡푦푖 = exp(푁퐻퐵_푠𝑖푧푒푖) /푠푢푚푎푙푙(exp(푁퐻퐵_푠𝑖푧푒푖)) Where:
HH[i] = Number of households in zone i
Retail [i] = Retail employment in zone i
Service [i] = Service employment in zone i
Other [i] = other employment in zone i
Education[i] = Education employment in zone i
P a g e | 17
The users can also specify a zone pair to trace results for. This is helpful in understanding how the trips are distributed between the zone pairs, and can be used for debugging. The trace output lists the number of trips by purpose along with the computed utilities and size terms. Tracing is turned on in the dcCoefficients.txt file, as shown in Figure 5. The trace output reports all destination choice model computations to the trace.txt file as shown in Figure 6. The output trip distribution matrix is written to dc_trips.mtx.
Due to the special trip generation characteristics of Idaho State University (ISU), the trips to and from ISU can be scaled up after trip distribution by a user-defined ISU Factor specified in the model bin file. This factor is currently set to one (i.e. no scaling).
Table 18- List of Input and Output Files to Trip Distribution
Input/Output File Name Description input Output\spmat_*.mat Travel time skim matrix file input Output\tripsBYPurpose.bin Aggregate trips by purpose input Input\dcCoefficients.txt Destination choice model coefficients input ISU Factor ISU trip adjustment factor Output Output\dc_trips.mtx Trip distribution trip table Output Output\trace.txt Trace output
Figure 5- Sample Contents of dcCoefficient.txt File
P a g e | 18
Figure 6- Output of Trace.txt file
8 MODE SPLIT AND TIME OF DAY
To assist in the development of scenarios during the development of the 2040 Metropolitan Transportation Plan the TDM was developed to allow for other modes to be removed from the trip tables. The household travel survey data was evaluated to determine the percent each mode (vehicle or non-vehicle) contributes to the total number of trips. Eight Districts (Figure 7) were created during the household survey the total model split for each district to district pair.
An additional section of this stage is to load in the external stations to the network. The BTPO TDM covers a small area and trips move from the inside the model domain to outside, outside to inside and through the model domain. The BTPO planning area has two interstates which make accounting for trips on the interstate which never leave the interstate critical. In the 2011 model development, Bluetooth readers were placed on I-86, I-15 north of the region and I-15 south our the region. This data and knowledge of the area allowed the development of a trip table to address external trips.
In this stage, total daily vehicle trip matrices are computed from the trip distribution person trip matrices. Table 19 shows the list of input and output files used in this stage. The following steps describe this stage:
P a g e | 19
The model creates a zonal matrix and fills it with the district to district codes based on the District field in the TAZDATA_2010.bin file (see Table 13 for all attributes). The zonal person trips are multiplied by the input district to district mode shares in modesplit.csv. The time of day factors in Tod_factors.csv and vehicle occupancy factors in VecOcc_factors.csv are applied to split the daily person trips into time-period specific trips and then to convert the person trips to vehicle trips. The model applies the external model to compute external-external (EE), external-internal (EI) and internal–external (IE) vehicle trips. The model applies external shares, shown in Table 24, and input in External_ADT.bin, to split the observed annual daily traffic ((or estimated ADT for future years) to EE_IN, EE_OUT, IE and EI auto trips. Then the model distributes external- external (EE) trips, via a Frater process, with an observed base seed (EE_triptable.mtx) and EE_IN and EE_OUT as control row and column totals. The model further distributes IE and EI trips to the internal zones based on the zonal share of population and employment respectively. Next, it averages the EI and IE trips to balance the results. At last the model combines the IE, EI, and EE trips into a matrix.
All external trips are split by the time of day and then combined with the II trip tables to build trip matrices for all zones. Table 20 to Table 24 show the list of input files to this stage and Table 25 shows the EI and IE trip attributes at the external stations.
Table 19- List of Input and Output Files to Mode Split Stage
Input/Output File Name Description input Output\dc_trips.mtx Trip distribution trip table input Input\modesplit.csv Mode split trip table input Input\Tod_factors.csv Time of day factors input Input\VecOcc_factors.csv Vehicle occupancy factors input Input\EE_triptable.mtx External-External seed trip matrix input Input\External_ADT.bin External station ADT Output Output\Districts.mtx Zonal matrix with district-to-district codes Output Output\mc_trips.mtx Auto person trips Output Output\mc_trips_transposed.mtx Transposed auto person trips Output Output\II_triptable.mtx Internal-Internal auto vehicle trips by time period Output Output\External_ADT.bin External trips by direction Output Output\EE_triptable.mtx External - External trip matrix Output Output\EITrips.mtx External-Internal trips Output Output\EETrips.mtx External-External trips
P a g e | 20
Figure 7- Districts
P a g e | 21
Table 20- List of Attributes in Mode Split (modesplit.csv)
Field Description from_district From district to_district To district year Year HBW_auto Share of HBW auto trips HBO_auto Share of HBO auto trips HBS_auto Share of HBS auto trips SCH_auto Share of SCH auto trips HBC_auto Share of HBC auto trips NHB_auto Share of NHB auto trips
Table 21- Model Split from_ to_ year HBW_ HBO_ HBS_ SCH_ HBC_ NHB_ district district auto auto auto auto auto auto 1 1 2010 0.98 0.82 0.98 0.5 1 0.88 1 2 2010 0.99 0.99 0.99 0.5 1 1 1 3 2010 1 0.99 0.99 0.5 1 1 1 4 2010 0.95 0.99 0.99 0.5 1 1 1 5 2010 1 1 1 0.5 1 1 1 6 2010 1 0.98 0.98 0.5 1 1 1 7 2010 0.99 0.99 0.99 0.5 1 1 1 8 2010 1 1 1 0.5 1 1 2 1 2010 1 0.99 0.99 0.5 1 1 2 2 2010 0.99 0.82 0.98 0.5 1 0.88 2 3 2010 1 0.99 0.99 0.5 1 1 2 4 2010 0.99 0.99 0.99 0.5 1 1 2 5 2010 1 1 1 0.5 1 1 2 6 2010 0.99 0.99 0.99 0.5 1 1 2 7 2010 1 0.99 0.99 0.5 1 1 2 8 2010 1 1 1 0.5 1 1 3 1 2010 0.99 0.99 0.99 0.5 1 1 3 2 2010 1 0.99 0.99 0.5 1 1 3 3 2010 0.99 0.82 0.98 0.5 1 0.88 3 4 2010 0.98 0.99 0.99 0.5 1 1 3 5 2010 1 1 1 0.5 1 1 3 6 2010 0.98 0.99 0.99 0.5 1 1 3 7 2010 0.99 0.99 0.99 0.5 1 1 3 8 2010 1 1 1 0.5 1 1 4 1 2010 0.99 0.99 0.99 0.5 1 1
P a g e | 22
from_ to_ year HBW_ HBO_ HBS_ SCH_ HBC_ NHB_ district district auto auto auto auto auto auto 4 2 2010 1 0.99 0.99 0.5 1 1 4 3 2010 0.99 0.99 0.99 0.5 1 1 4 4 2010 0.9 0.8 0.95 0.5 0.5 0.88 4 5 2010 1 1 1 0.5 1 1 4 6 2010 0.98 0.98 0.98 0.5 1 1 4 7 2010 0.94 0.98 0.98 0.5 0.95 1 4 8 2010 1 1 1 0.5 1 1 5 1 2010 1 1 1 0.5 1 1 5 2 2010 1 1 1 0.5 1 1 5 3 2010 1 1 1 0.5 1 1 5 4 2010 1 1 1 0.5 1 1 5 5 2010 1 0.82 0.99 0.5 1 0.88 5 6 2010 1 1 1 0.5 1 1 5 7 2010 1 1 1 0.5 1 1 5 8 2010 1 1 1 0.5 1 1 6 1 2010 0.99 0.98 0.98 0.5 1 1 6 2 2010 0.99 0.99 0.99 0.5 1 1 6 3 2010 0.99 0.99 0.99 0.5 1 1 6 4 2010 0.95 0.98 0.98 0.5 1 1 6 5 2010 1 1 1 0.5 1 1 6 6 2010 0.99 0.82 0.98 0.5 1 0.88 6 7 2010 0.99 1 1 0.5 1 1 6 8 2010 1 1 1 0.5 1 1 7 1 2010 0.99 0.99 0.99 0.5 1 1 7 2 2010 1 0.99 0.99 0.5 1 1 7 3 2010 0.99 0.99 0.99 0.5 1 1 7 4 2010 0.94 0.98 0.98 0.5 0.95 1 7 5 2010 1 1 1 0.5 1 1 7 6 2010 0.99 1 1 0.5 1 1 7 7 2010 0.98 0.82 0.98 0.5 1 0.88 7 8 2010 1 1 1 0.5 1 1 8 1 2010 1 1 1 0.5 1 1 8 2 2010 1 1 1 0.5 1 1 8 3 2010 1 1 1 0.5 1 1 8 4 2010 1 1 1 0.5 1 1 8 5 2010 1 1 1 0.5 1 1 8 6 2010 1 1 1 0.5 1 1 8 7 2010 1 1 1 0.5 1 1 8 8 2010 1 0.82 1 0.5 1 0.88
P a g e | 23
Table 22- List of Attributes in Time of Day (TOD_Factors.csv)
Field Description Purpose Trip purpose am_pa AM Production-to-Attraction (PA) share of daily trips am_ap AM Attraction-to-Production (AP) share of daily trips md_pa MD PA share of daily trips md_ap AM AP share of daily trips pm_pa PM PA share of daily trips pm_ap PM AP share of daily trips daily_ap Daily PA share of daily trips daily_pa Daily AP share of daily trips
Table 23- List of Attributes to Vehicle Occupancy (VECOCC_factors.csv) and Factors
Field Description Factor HBW HBW vehicle occupancy factor 1.05 HBO HBO vehicle occupancy factor 1.49 HBS HBS vehicle occupancy factor 1.34 SCH SCH vehicle occupancy factor 2.83 HBC HBC vehicle occupancy factor 1.09 NHB NHB vehicle occupancy factor 1.9
Table 24- List of Attributes in External Average Daily Traffic (External_ADT..bin)
Field Description Station External station node number ADT Average annual daily traffic count EE Percent External-External traffic EE_IN Percent of EE traffic flowing into the model region EE_OUT Percent of EE traffic flowing out of the model region EI Percent External-Internal traffic IE Percent Internal- External traffic
Table 25- List of Calculated Attributes Added to External_ADT.bin File
Field Description Station External station node number EE_IN_Trips External trips that travel into the region (may not necessarily end in the region) EE_OUT_Trips External trips that travel out of the region (need not originate in the region) EI_Trips External trips that end in the region IE_Trips External trips that start in the region
P a g e | 24
Figure 8- External Station Locations
9 HIGHWAY ASSIGNMENT
The vehicle trips by time period are assigned to the highway network and link and intersection volumes are calculated at this stage. Table 26 lists the inputs and outputs used in this stage. The model uses the User Equilibrium (UE) traffic assignment method with an Akcelik volume-delay function. The Akcelik function is defined as:
2 2 2 푅 = 푅표 + 퐷0 + 0.25푇 [(푥 − 1) + √(푥 − 1) + 16퐽 ∗ 푋 ∗ 퐿 ÷ 푇 ]
P a g e | 25
Where:
R = Link traversal time
Ro = Free flow link traversal time
Do = Zero flow control delay
T = expected duration of demand
X = Flow to capacity ratio
J = Calibration parameter
L – Link length
Table 27 lists the parameters used in the highway assignment. Table 28 lists the calculated highway assignment fields.
Table 26- List of Attributes and Output Files to Highway Assignment
Input/Output File Name Description Input Output\HighwayNetwork.dbd Highway database Input Input\turn penalty.dbf Turn penalty or prohibition table Input Output\all_trips.mtx All II, EI, IE and EE trips by time period Input Output\network.net Highway network file Output Output\HighwayNetwork.bin Highway database with calculated attributes Output Output\AM_FLOW.bin AM link volume from the final iteration Output Output\MD_FLOW.bin MD link volume from the final iteration Output Output\PM_FLOW.bin PM link volume from the final iteration Output Output\Daily_FLOW.bin Daily link volume from the final iteration Output Output\AM_TurnVolume.bin Intersection volume for AM period Output Output\MD_TurnVolume.bin Intersection volume for MD period Output Output\PM_TurnVolume.bin Intersection volume for PM period Output Output\Daily_TurnVolume.bin Intersection volume for Daily period
Table 27- Parameters Used in the Highway Assignment
Parameters Description Value Assignment Iterations Maximum number of Iterations 50 Assignment Convergence Convergence for highway assignment 0.0001
P a g e | 26
Table 28- Highway Assignment Calculated Fields
Field Description File VDFParameter J quality of service term for the Akcelik VDF HighwayNetwork.bin Con_AB_Time_
10 MODEL ESTIMATION AND CALIBRATION
This section describes the model input parameters, rates, and coefficients used by the trip generation, trip distribution, mode split, and time-of-day factoring models. The section was taken from the BTPO Travel Model User Guide September 2012. The model estimated used for this update is the same as the 2012 TDM.
1) HOUSEHOLD DISAGGREGATION MODEL The household disaggregation model allocates the input households to households by size and the number of workers. The percent of households by size and the number of workers was calculated from the 2010 Census Transportation Planning Package (CTPP) block group level tabulation 47 (total persons)
P a g e | 27
and 75 (households by size, the number of workers, and household income). A lookup table was developed to determine the share of households by household size based on the persons per household by TAZ. As shown in Figure 9, the percent of households by household size by TAZ is looked up from the persons per households for that TAZ. After determining the number of households by household size, a simple disaggregation to the number of workers in the household is done using the percent of each worker category by household size from the Census data as well.
Figure 9- Bannock County Households by Household Size
Table 29- Percent of Workers per Household Size Lookup Table
Number of workers 0 wkr 1 wkr 2 wkr 3+ wkr Total 1 per 0.484 0.516 0 0 1 2 per 0.308 0.300 0.393 0 1
3 per 0.082 0.365 0.416 0.137 1 HH Size HH 4+ per 0.045 0.281 0.469 0.205 1
2) TRIP GENERATION Average weekday person trips are generated for six trip purposes:
P a g e | 28
• HBW – Home-Based Work • HBC – Home-Based College • SCH – Home-Based School • HBS – Home-Based Shopping • HBO – Home-Based Other • NHB – Non-Home-Based
The production rates for all purposes were classified by household size and the number of workers per household Table 30 to Table 35 displays the trip rates by a trip purpose that were estimated from the BTPO household travel survey.
Table 30- Home Based Work Trip Rates
Number of Workers Household Size 0 1 2 3+ Total 1 - 1.43 - - 0.43 2 - 1.53 2.88 - 1.22 3 - 1.49 2.38 4.53 2.24 4+ - 1.10 2.39 4.89 1.95 Total - 1.28 2.50 4.75 1.62
Table 31- Home Based College Trip Rates
Number of Workers Household Size 0 1 2 3+ Total 1 0.01 - - - 0.01 2 0.02 0.13 0.01 - 0.05 3 0.14 0.34 0.11 0.14 0.20 4+ 0.64 0.18 0.23 0.81 0.30 Total 0.10 0.17 0.16 0.55 0.18
Table 32- Home Based School Trip Rates
Number of Workers Household Size 0 1 2 3+ Total 1 - - - - - 2 0.10 0.06 - - 0.06 3 0.27 0.38 1.12 0.24 0.59 4+ 3.19 3.34 3.27 1.76 3.12 Total 0.42 1.92 2.09 1.17 1.53
P a g e | 29
Table 33- Home Based Shopping Trip Rates
Number of Workers Household Size 0 1 2 3+ Total 1 0.67 0.17 - - 0.52 2 1.22 0.69 0.59 - 0.90 3 1.21 1.28 0.51 0.81 0.91 4+ 2.56 1.49 0.92 1.23 1.32 Total 1.16 1.17 0.76 1.07 1.04
Table 34- Home Based Other Trip Rates
Number of Workers Household Size 0 1 2 3+ Total 1 1.66 1.17 - - 1.51 2 3.10 2.37 1.50 - 2.46 3 4.61 3.00 4.14 2.29 3.43 4+ 6.51 8.91 6.20 7.42 7.61 Total 3.06 5.96 4.71 5.43 4.84
Table 35- Non Home Based Trip Rates
Number of Workers Household Size 0 1 2 3+ Total 1 0.97 0.91 - - 0.95 2 1.73 1.97 2.12 - 1.91 3 4.59 1.96 2.76 4.68 3.10 4+ 2.82 3.75 3.95 4.88 3.90 Total 1.80 2.84 3.29 4.80 2.89
3) TRIP DISTRIBUTION The destination choice trip distribution models were developed from the household travel survey data using a multinomial logit estimation procedure. In a destination choice model, trip productions are distributed to destination zones by calculating a utility of each destination that is a function of the impedance (i.e. travel time) and size of the destination (i.e. employment at the location). The variables used in the model are:
A) Auto travel time - Autott = auto travel time matrix B) Employment in attraction zones: a. RetEmp = retail trade employees b. SvcEmp = service employees c. EduEmp = education employees, are not including Idaho State University employees d. ISUEmp = Idaho State University employees e. OthEmp = all other employees f. Households = households
P a g e | 30
Table 36 and Table 37 show the destination choice utility expressions and estimated model coefficients used in the model.
Table 36- Destination Choice Model Utilities by Purpose
Purpose Utility Expressions HBWORK U = exp ( -0.2975 *autott + 0.0139*(autott2) – 0.0002*(autott3)+ ln (1.0*RetEmp + 2.1441*SvcEmp + 3.5078*EduEmp + 0.2789*ISUEmp + 0.3736*OthEmp + 0.4027*Households)) HBOTHER U = exp ( -0.8897 *autott + 0.0584*(autott2) – 0.0015*(autott3)+ ln (1.0*RetEmp + 4.9431*SvcEmp + 5.8299*EduEmp + 0.8409*ISUEmp + 2.5577*Households)) HBSHOP U = exp ( -1.0320*autott + 0.0719*(autott2) – 0.0018*(autott3)+ ln (1.0*RetEmp + 0.1175*SvcEmp + 0.0001*OthEmp + 0.0103*Households)) HBSCHOOL U = exp ( -1.2490*autott + 0.0902*(autott2) – 0.0024*(autott3)+ ln (1.0*EduEmp + 0.0616*SvcEmp + 0.0239*OthEmp)) HBCOLLEGE U = exp (ln (1.0 * ISUEmp))
NHB U = exp ( -0.6339*autott + 0.0355*(autott2) – 0.0008*(autott3)+ ln (1.0*RetEmp + 0.7724*SvcEmp + 0.6340*EduEmp + 0.0686*ISUEmp + 0.0841*OthEmp + 0.3116*Households)) NHB Production Model U = exp ( ln(1.0*RetEmp + 0.965*SvcEmp + 1.240*EduEmp + 0.105*OthEmp + 0.261*Households))
P a g e | 31
Table 37: Destination Choice Estimated Variable Coefficients
Purpose Work College School Shopping Other Non-Home Based Coeff t-stat Coeff t-stat Coeff t-stat Coeff t-stat Coeff t-stat Coeff t-stat Time -0.2975 -3.2 -1.4050 -10.5 -1.032 -8.0 -0.8897 -15.8 -0.6339 -9.5 Time Squared 0.0139 1.6 0.1063 6.9 0.0719 5.2 0.0584 9.6 0.0355 4.4 Time Cubed -0.0020 -1.0 -0.0029 -5.5 -0.0018 -4.1 -0.0015 -7.6 -0.0008 -2.9 Size Variables Other Employment 0.3736 -3.9 0.0239 -24.1 0.0001 -2.8 0.0841 -12.4 Households 0.4027 -3.9 0.0103 -17.3 2.5577 5.2 0.3116 -11.6 Retail Employment 1.0000 1.0000 1.0000 1.0000 Education (not including ISU) 1.1850 5.4 1.0000 1.0000 5.8299 9.0 0.6340 -2.7 Service Employment 2.1441 3.5 0.0616 -26.0 0.1175 -14.5 4.9431 8.5 0.7724 -2.4 ISU Employees 0.2789 -2.5 0.8409 -0.6 0.0686 -6.4
P a g e | 32
4) MODE SPLIT The mode split model is based on a simple lookup table of auto mode shares by district production- attraction pair as calculated from the household survey. The two modes defined in the model are auto (which includes drive alone, drive with a passenger, and being a passenger), and non-auto (which includes transit, school bus, walk, and bike). The mode distribution factors also vary by trip purpose and production and attraction district. Table 38 to Table 43 shows the auto share by production and attraction district for each trip purpose. See Figure 8 for the district definitions.
Table 38: HBW Auto Share Production District 1 2 3 4 5 6 7 8
1 100.0% 100.0% 100.0% 95% 100.0% 100.0% 100.0% 100.0%
2 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
3 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
4 100.0% 100.0% 100.0% 95% 100.0% 100.0% 100.0% 100.0%
5 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
6 100.0% 100.0% 100.0% 95% 100.0% 100.0% 100.0% 100.0%
7 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
8 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
Table 39: HBC Auto Share Production District 1 2 3 4 5 6 7 8
1 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
2 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
3 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
4 100.0% 100.0% 100.0% 50% 100.0% 100.0% 100.0% 100.0%
5 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
6 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
7 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
8 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
2015 Travel Demand Model Documentation P a g e | 33
Table 40: SCH Auto Share Production District 1 2 3 4 5 6 7 8
1 55% 55% 55% 55% 55% 55% 55% 55%
2 55% 55% 55% 55% 55% 55% 55% 55%
3 55% 55% 55% 55% 55% 55% 55% 55%
4 55% 55% 55% 55% 55% 55% 55% 55%
5 55% 55% 55% 55% 55% 55% 55% 55%
6 55% 55% 55% 55% 55% 55% 55% 55%
7 55% 55% 55% 55% 55% 55% 55% 55%
8 55% 55% 55% 55% 55% 55% 55% 55%
Table 41: HBS Auto Share Production District 1 2 3 4 5 6 7 8
1 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
2 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
3 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
4 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
5 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
6 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
7 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
8 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
2015 Travel Demand Model Documentation P a g e | 34
Table 42: HBO Auto Share Production District 1 2 3 4 5 6 7 8
1 84% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
2 100.0% 84% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
3 100.0% 100.0% 84% 100.0% 100.0% 100.0% 100.0% 100.0%
4 100.0% 100.0% 100.0% 84% 100.0% 100.0% 100.0% 100.0%
5 100.0% 100.0% 100.0% 100.0% 84% 100.0% 100.0% 100.0%
6 100.0% 100.0% 100.0% 100.0% 100.0% 84% 100.0% 100.0%
7 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 84% 100.0%
8 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 84%
Table 43: NHB Auto Share Production District 1 2 3 4 5 6 7 8
1 89% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
2 100.0% 89% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
3 100.0% 100.0% 89% 100.0% 100.0% 100.0% 100.0% 100.0%
4 100.0% 100.0% 100.0% 89% 100.0% 100.0% 100.0% 100.0%
5 100.0% 100.0% 100.0% 100.0% 89% 100.0% 100.0% 100.0%
6 100.0% 100.0% 100.0% 100.0% 100.0% 89% 100.0% 100.0%
7 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 89% 100.0%
8 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 89%
2015 Travel Demand Model Documentation P a g e | 35
5) TIME OF DAY FACTORS Travel by time of day is estimated for auto trips, and the factors are direction-specific (production to attraction or attraction to production direction). The time of day factors was estimated by hour using the trip start and end time data from the household travel survey. Figure 9 to Figure 16 show the trips- in-motion distribution by time of day for all trips and then each trip purpose. A trip was counted in each half hour time-period if its start or end time fell in the particular half hour bin. These figures were used to determine the peak hours (5:00 – 6:00 pm) of the day that was modeled. The external trip time-of- day factors were calculated based on the hourly share of daily traffic counts.
Figure 10: All Trip Purposes Combined Trips in Motion Diurnal Distribution
30,000
25,000
20,000
of Auto Trips Auto of 15,000
10,000 Number
5,000
-
1:00PM 2:00PM 3:00PM 4:00PM 5:00PM 6:00PM 7:00PM 8:00PM 9:00PM
1:00AM 2:00AM 3:00AM 4:00AM 5:00AM 6:00AM 7:00AM 8:00AM 9:00AM
12:00PM 10:00PM 11:00PM
12:00AM 10:00AM 11:00AM
2015 Travel Demand Model Documentation P a g e | 36
Figure 11: HBW Trips in Motion Diurnal Distribution 6,000
5,000
4,000
of Auto Trips Auto of 3,000
Number 2,000
1,000
-
1:00 PM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM 7:00 PM 8:00 PM 9:00
1:00 AM 1:00 AM 2:00 AM 3:00 AM 4:00 AM 5:00 AM 6:00 AM 7:00 AM 8:00 AM 9:00
12:00 PM 12:00 PM 10:00 PM 11:00
12:00 AM 12:00 AM 10:00 AM 11:00
Figure 12: HBC Trips in Motion Diurnal Distribution 800
700
600
500
400
of Auto Trips Auto of 300
200 Number 100
-
1:00PM 2:00PM 3:00PM 4:00PM 5:00PM 6:00PM 7:00PM 8:00PM 9:00PM
1:00AM 2:00AM 3:00AM 4:00AM 5:00AM 6:00AM 7:00AM 8:00AM 9:00AM
12:00PM 10:00PM 11:00PM
12:00AM 10:00AM 11:00AM
2015 Travel Demand Model Documentation P a g e | 37
Figure 13: SCH Trips in Motion Diurnal Distribution 12,000
10,000
8,000
6,000
of Auto Trips Auto of 4,000
2,000 Number
-
1:00PM 4:00PM 2:00PM 3:00PM 5:00PM 6:00PM 7:00PM 8:00PM 9:00PM
1:00AM 2:00AM 3:00AM 4:00AM 5:00AM 6:00AM 7:00AM 8:00AM 9:00AM
12:00PM 10:00PM 11:00PM
12:00AM 10:00AM 11:00AM
Figure 14: HBS Trips in Motion Diurnal Distribution 3,000
2,500
2,000
1,500 of Auto Trips Auto of 1,000
Number 500
-
6:00PM 1:00PM 2:00PM 3:00PM 4:00PM 5:00PM 7:00PM 8:00PM 9:00PM
9:00AM 1:00AM 2:00AM 3:00AM 4:00AM 5:00AM 6:00AM 7:00AM 8:00AM
12:00PM 10:00PM 11:00PM
12:00AM 10:00AM 11:00AM
2015 Travel Demand Model Documentation P a g e | 38
Figure 15: HBO Trips in Motion Diurnal Distribution 12,000
10,000
8,000
6,000 of Auto Trips Auto of
4,000 Number 2,000
-
1:00PM 2:00PM 3:00PM 4:00PM 5:00PM 6:00PM 7:00PM 8:00PM 9:00PM
1:00AM 2:00AM 3:00AM 4:00AM 5:00AM 6:00AM 7:00AM 8:00AM 9:00AM
11:00 PM 11:00 12:00PM 10:00PM
11:00 AM 11:00 12:00AM 10:00AM
Figure 16: NHB Trips in Motion Diurnal Distribution 8,000
7,000
6,000
5,000
4,000
of Auto Trips Auto of 3,000
2,000 Number 1,000
-
3:00PM 1:00PM 2:00PM 4:00PM 5:00PM 6:00PM 7:00PM 8:00PM 9:00PM
1:00AM 2:00AM 3:00AM 4:00AM 5:00AM 6:00AM 7:00AM 8:00AM 9:00AM
12:00PM 10:00PM 11:00PM
12:00AM 10:00AM 11:00AM
The time of day factors shown in Table 44 and Table 45 are used to convert production to attraction (PA) matrices into OD matrices by multiplying the daily PA trip matrix first by the time-of-day percent for each time period and then by the share of trips that are PA or AP direction in that time period. An
2015 Travel Demand Model Documentation P a g e | 39
example of a trip in the PA direction is a trip from home to work. An example of a trip in the AP direction is a trip from work to home.
Table 44- Time of Day Factors by Mode
Total HBW HBColl HBSch HBShop HBOth NHB Total 7:00 to 8:00am 18.2% 18.3% 31.5% 1.0% 6.7% 3.3% 8.3% 12:00 to 1:00pm 5.8% 6.9% 2.3% 13.9% 8.6% 12.9% 9.6% 4:30 to 5:30pm 16.8% 3.5% 6.8% 8.4% 12.4% 11.3% 11.7%
Table 45: Directional Distribution Factors1 Time of Day HBW HBColl HBSch HBShop HBOth NHB AM_PA 17.4% 18.3% 31.5% 0.8% 5.1% 3.3% AM_AP 0.9% 0.0% 0.0% 0.2% 1.5% 0.0% MD_PA 2.3% 2.3% 1.0% 5.9% 4.1% 12.9% MD_AP 3.5% 4.6% 1.3% 8.0% 4.4% 0.0% PM_PA 1.2% 1.0% 1.3% 2.1% 4.8% 11.3% PM_AP 15.6% 2.6% 5.5% 6.3% 7.6% 0.0% Daily_AP 50.0% 50.0% 50.0% 50.0% 50.0% 50.0% Daily_PA 50.0% 50.0% 50.0% 50.0% 50.0% 50.0% 1Share of trips oriented A->P by trip purpose and time-of-day. Share of trips P->A is (1-factor).
6) EXTERNAL TO EXTERNAL TRIP DISTRIBUTION The external to external (EE) trips are distributed from origins to productions based on the findings from the Bluetooth external OD survey. The trip destination shares for the four Bluetooth count locations are shown in the table below. The distribution patterns to/from the other external stations were asserted based on local knowledge.
Table 46: Bluetooth External Survey Distribution
Trip Destination I-15 N US 91 N I-15 E I-86 W Internal
I-15 N 0% 0% 36% 19% 45%
US 91 N 0% 0% 4% 8% 88% I-15 E 49% 0% 0% 10% 41%
I-86 W 29% 0% 14% 0% 57% Trip Origin Trip Internal 35% 7% 37% 21% 0%
2015 Travel Demand Model Documentation P a g e | 40
11 MODEL VALIDATION
The model validation effort focused on analyzing the model results across a number of dimensions to ensure the model is producing reasonable results. A number of validation summaries are presented and discussed below.
11.1 NETWORK ZONES As described in the Section 2 Street Network and Traffic Analysis Zones, the TDM network was built using all the public street that were determined using GPS technology. The distance of the link are reasonable, and the centroid connectors are located in appropriate locations. Even if the location of centroids were incorrect, the density of the street network would compensate for any issues in the location of the centroids. The TAZ’s likewise were built on the 2010 US Census Block data. The size of the TAZs is consistent with current state of the practice.
11.2 SOCIOECONOMIC DATA This report does not specifically cover the socioeconomic date generation and validation. For specific information on the based 2015 and future socioeconomic data see the 2015 Demographic Update Report (Bannock Transportation Planning Organization, 2016). In summary, the model requires the number of households and employment by type.
The 2015 household data was developed by taking the 2010 Census data and adding the number of housing units built during the time from May 2010 to May 2015. The vacancy rate for the zone in 2010 was used to determine the households in 2015. The base 2015 employment data by section came from the Bureau of Economic Analysis (BEA) employment data for 2001 to 2013. 2015 employment data was calculated by taking the growth rate from 2010 to 2013 and predicting the 2015 employments data. This is the control number the actual number of employees by type was calculated from the Idaho Department of Labor employment data which was geo-located and aggregated by TAZ. The percentage of the total which each zone represented was used to grow the Idaho Department of Employment data to the Bureau of Economic Analysis data.
11.3 TRIP GENERATION The household trip generation model produces 9.4 person trips per household, which is comparable to published national trip generation rates. The rates by HBW, HBNW (non-work) and NHB are presented in the table below. However, during model calibration and validation, it was determined that the model was not producing enough daily vehicle trips per household. The previous BTPO model approximately matched daily VMT by producing 9.4 vehicle trips per household. Given the reasonable trip distribution, mode split, and vehicle occupancy results described below, it was decided to review the trip generation model and the household travel survey analysis for possible revisions. When compared to the 2010 Census for Bannock County, the surveyed distribution of householder age was skewed toward older households, which tend to generate fewer trips. As a result, a trip generation adjustment factor was calculated by purpose and used to scale the trip generation rates. It was calculated by normalizing the household distribution by the age of head to match the 2010 Census and then calculating trip rates by purpose. Table 48 lists the calculated adjustment factor.
2015 Travel Demand Model Documentation P a g e | 41
Table 47: Trip Generation Rate Check
Purpose Model NCHRP 716 NCHRP 365 2009 NHTS HBW 1.6 1.8 1.8 1.4 HBNW 5.4 4.9 5.2 4.9 NHB 2.5 3.0 2.1 2.8 Total 9.4 9.7 9.2 9.1
Table 48: Trip Generation Adjustment Factor
Purpose Factor HBW 1.00 HBO 1.38 HBS 1.26 SCH 1.74 HBC 1.30 NHB 1.22
11.4 TRIP DISTRIBUTION The distribution results were reviewed according to their trip length frequencies and district to district flows by trip purpose. Due to some discrepancies between the observed and estimated HBShop trips by destination zone, small, less than one minute of equivalent travel time, calibration constants were added to the HBShop distribution model to encourage trips to districts 2 and 6. A similar, less than one minute of equivalent travel time, calibration constant was added to district 3 for SCH trips. These constants can be thought of as employment correction factors to make certain areas more or less attractive as destinations. Overall the model is matching the trip length frequencies by purpose quite well, and the district to district flows by purpose are reasonable. The trip length frequencies we taken from the 2015 model runs. There were some differences in the percentage of the 2012 model, but the curves were very similar.
2015 Travel Demand Model Documentation P a g e | 42
HWB Trip Length Frequency 25.00%
20.00%
15.00%
10.00% Percentage Percentage 5.00%
0.00% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Distance
Observed Percentage Modeled Percentage
Figure 17- Home Base Work Trip Length Frequency
HBO Trip Length Frequency 35.00% 30.00% 25.00% 20.00% 15.00%
Percentage Percentage 10.00% 5.00% 0.00% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Distance
Observed Percentage Modeled Percentage
Figure 18-Home Base Other Trip Length Frequency
2015 Travel Demand Model Documentation P a g e | 43
HBS Trip Length Frequncy 35.00% 30.00% 25.00% 20.00% 15.00%
Percentage Percentage 10.00% 5.00% 0.00% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Distance
Observed Percentage Modeled Percentage
Figure 19: Home Base Shopping Trip Length Frequency
SCH Trip Length Frequency
40.00% 35.00%
30.00%
25.00% 20.00% 15.00% Percentage Percentage 10.00%
5.00% 0.00% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Distance
Observed Percentage Modeled Percentage
Figure 20: School Trip Length Frequency
2015 Travel Demand Model Documentation P a g e | 44
HBC Trip Length Frequency 25.00%
20.00%
15.00%
10.00% Percentage Percentage 5.00%
0.00% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Distance
Observed Percentage Modeled Percentage
Figure 21: Home Base College Trip Length Frequency HBC Trip Length Frequency 25.00%
20.00%
15.00%
10.00% Percentage Percentage 5.00%
0.00% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Distance
Observed Percentage Modeled Percentage
Figure 22: Not Home Base Trip Length Frequency
11.5 MODE SPLIT, VEHICLE OCCUPANCY, AND COMMERCIAL AND OTHER TRIPS The mode split percent and vehicle occupancy rates are applied to the distribution matrix in order to create daily vehicle trips. The model produces 12.1 person trips per household which are converted to 8.2 vehicle trips per household when accounting for the mode of travel and vehicle occupancy.
The NHB trips were factored up after the vehicle occupancy step to account for commercial vehicle and other missing trips such as ISU student travel, In the Calgary business establishments survey, which was used in the development of the Calgary commercial movement model (Hunt, 2003), commercial vehicles were estimated to be 15% of regional traffic. This 15% commercial vehicle factor was applied to boost NHB trips in the BTPO model. It was applied by calculating the number of NHB trips needed to add 15%
2015 Travel Demand Model Documentation P a g e | 45
more trips to the model. This resulted in about 40k more trips. An additional 35k NHB trips were added to account for other missing trips, most notably ISU student travel. These additional trips were calculated by comparing the daily assignment results to the daily traffic counts. Adjusting the NHB vehicle occupancy factor from 1.43 to 0.6 resulted in 75k additional trips. Table 49 provides a summary of the total number of trips at each step of the model.
Table 49-Total Trips Check
Trip Type Trips Person Trips Post Distribution 334,982 Auto Trips Post Mode Choice 302,459 Auto Trips Post Vehicle Occupancy 225,409 Auto Trips External Model 49,175 Commercial and Other Trips 76,186 Auto Trips Daily Assignment 350,770
11.6 NETWORK ASSIGNMENT The network assignment step was calibrated and validated by comparing traffic counts to assigned link volumes, as well as by adjusting the zero flow control delay parameter used in the Akcelik volume-delay function. This zero flow control delay term can be used to account for intersection control delay and is calculated by area type and facility type in the model. It is calculated as a reduction in link travel time as a percent of link speed. The values were calculated from the BTPO speed study and averaged across area type and facility type. The reductions in link travel time are not applied to connectors or the interstate freeway.
Table 50: Zero Flow Control Delay Adjustment Factors
Average Intersection Delay (as a % of link speed) Area Type Facility Class Rural Urban CBD Centroid Connector 0.00 0.00 0.00 Local 0.05 0.05 0.05 Collector 0.05 0.05 0.10 Minor Arterial 0.05 0.08 0.05 Principal Arterial 0.10 0.07 0.08 Ramp 0.15 0.09 0.14 Interstate 0.00 0.00 0.00
The network free-flow speeds were adjusted in a few places to model trip route choice better. The link speeds were adjusted as follows: 1) speeds on in-town interstate ramps were reduced to encourage trips to stay on the arterial roads, 2) speeds on links along Yellowstone, Hawthorne, Hiline, and Olympus were reduced to discourage usage, and 3) speeds on Gould, Center, and Benton, were adjusted to deal with competing routes.
2015 Travel Demand Model Documentation P a g e | 46
11.6.1 Daily Traffic Assignment Figure 23 is a comparison of the observed link volumes (i.e. traffic counts) and the estimated link volumes. A scatterplot of the daily assignment results is presented below. It shows a good goodness-of- fit between the traffic counts and the assigned link volumes. The R-squared between the assigned link volumes and the traffic counts is 0.8962 for daily. The percent error between the counts and assigned volumes by facility type are presented in Table 5 below. The included target percent errors by facility type are from the 1990 FHWA publication Calibrating and Adjustment of System Planning Models. The model is within the acceptable percent errors. Figure 23 and Table 51 shows the complete daily assignment and VMT for facility type and area type.
Figure 23- Daily Observed versus Estimated Link Volume
Table 51 - Daily Observed versus Estimated Link Volumes by Facility Type
Facility Type Count Assigned % Error Target % Error Collector 24,839 24,315 2.11% 25.00% Minor Arterial 78,436 75,348 3.94% 10.00% Principal Arterial 116,460 116,497 -0.03% 10.00% Interstate 126,302 122,682 2.87% 7.00% Total 2,169,243 2,134,560 1.60% 5.00%
2015 Travel Demand Model Documentation P a g e | 47
Figure 24: Daily Assigned Volume
The final assignment validation summary is network vehicle miles traveled (VMT). Table 52 shows total network VMT by facility and area type. These results are similar to the previous 2011 modeled daily VMT of 1.372 million.
2015 Travel Demand Model Documentation P a g e | 48
Table 52 - Daily VMT by Facility Type
Facility Type Rural Urban CBD Total Local* 61,454 73,257 37,679 172,390 Collector 29,861 46,483 13,407 89,751 Minor Arterial 47,046 120,565 50,933 218,544 Principal Arterial 55,453 137,983 182,152 375,588 Ramp 3,096 18,353 11,945 33,394 Interstate 338,744 123,287 110,335 572,366 Total 535,654 519,928 406,451 1,462,033 * Includes centroid connectors
11.6.2 PM Peak Hour Traffic Assignment Figure 25 is a comparison of the observed link volumes (i.e. traffic counts) and the estimated link volumes for the PM Peak hour. A scatterplot of the daily assignment results is presented below. It shows a good goodness-of-fit between the traffic counts and the assigned link volumes. The R-squared between the assigned link volumes and the traffic counts is 0.8799 for PM Peak hour. The percent error between the counts and assigned volumes by facility type are presented in Table 53.
Figure 25- PM Peak Hour Volume vs. Count-
2015 Travel Demand Model Documentation P a g e | 49
Table 53- PM Peak Hour VMT by Facility Type
Facility Type Count Assigned % Error Target % Error Collector 2,175 2,275 -4.60% 25.00% Minor Arterial 7,005 6,471 7.62% 10.00% Principal Arterial 9,514 10,091 -6.06% 10.00% Interstate 10,271 9,107 11.33% 7.00% Total 28,965 27,944 3.52% 5.00%
11.6.3 Mid-Day Peak Hour Traffic Assignment Figure 26 is a comparison of the observed link volumes (i.e. traffic counts) and the estimated link volumes for the Mid-day Peak hour. A scatterplot of the daily assignment results is presented below. It shows a good goodness-of-fit between the traffic counts and the assigned link volumes. The R-squared between the assigned link volumes and the traffic counts is 0.8852 for Mid-Day Peak hour. Table 54 contains the percent error between the counts and assigned volumes by facility type.
Figure 26 - Mid-Day Peak Hour Observed vs. Estimated Link Volume
Table 54- Mid-Day Peak Hour Observed vs. Estimated VMT by Facility Type
Facility Type Count Assigned % Error Target % Error Collector 2,175 2,275 -4.60% 25.00% Minor Arterial 7,005 6,471 7.62% 10.00% Principal Arterial 9,514 10,091 -6.06% 10.00% Interstate 10,271 9,107 11.33% 7.00% Total 28,965 27,944 3.52% 5.00%
2015 Travel Demand Model Documentation P a g e | 50
11.6.4 Assessment The BTPO TDM for the 24-hour day and for PM and Mid-Day peak hour is a good representation of the regions traffic flows. Table 51, 52, and 53 all show the assignment is within the allowable error of for that facility class. The scatter plot also show that the assigned volumes are a good fit to the counts. A closer look at the scatter plots demonstrate that there are some areas where the model does not perform well. A case in point is the tri-corridors or Yellowstone Ave, Jefferson, and McKinley from Alameda to Oak. None of these corridor perform very well and many links are outside the percent error recommendations. The model does show the relative distribution and the preferred traffic flow well. Some of the issue with this and other problem areas in the region is compact development and the closely spaced street network. This make travel times very similar and even with a change in the volume-delay function capacity does not affect travel choice very much.
For overall modeling and to determine the impact of development to predict future travel patterns the model is a valid tool. For sub-area are smaller analysis some care should be observed and additional modeling conducted.
12 FUTURE YEAR MODELING OR SCENARIO MODELING
The primary purpose of the travel demand model is to predict future travel patterns given assumptions about future demographic or changes to the street network. This document’s first eleven chapters provide an overview of the current year model and demonstrate that the TDM is valid for 2015 base year. This chapter will provide an overview of the types of changes are made and how that change might be reflected in the travel demand model.
12.1 REQUIRED FILE CHANGES The base year is based on inputs that provide information about the demographics data, street network and external traffic. To model any future year the following files needs to be changed to represent the anticipated future conditions.
Street Network – The highway network provides the information about the speed, capacity, and delay factors for each link. Changes anticipated in the future need to be made to the highway network and that file referenced in the model setup. Required fields:
Length Dir Lane Quantity AB and BA (both directions) Posted Speed (both directions) Area Type Facility Type County
Demographics – The 2015 demographic was used to validate the Model. For model runs for future years such as 2020 to 2045 the 2015 Demographic Update Report provide the needed household and employment data by TAZ for every five year period from 2020 – 2045. For years between milestone years,’ a growth factor can be estimated between milestone year periods.
2015 Travel Demand Model Documentation P a g e | 51
12.2 POTENTIAL CHANGE FOR SCENARIO MODELING While changes to the demographic information over time and changes to the highway network are the most likely changes to model other potential changes from policy changes or changes in land use can be changed by modifications certain files within the model setup. Those files include:
Model Split (modesplit.csv) - In this file, the auto mode split is specified by district to district. The user can modify the factor depending on the scenario. The factors range from 0 to 1 where 0 is zero auto share, and 1 is 100 percent auto share between the districts.
External Trips (External_ADT.bin) - The share of external traffic, external-internal and internal- external are specified in this file. The users can modify the shares depending on the scenario.
Time of Day (tod_factors.bin) - The time of day factors are specified by PA and AP directions, by purpose. The users can also modify these to test a different peak hour.
Vehicle Occupancy Factors (VecOcc_factors.csv) - The vehicle occupancy factors by purpose are specified here. The user can modify the factors here to model an increase in carpooling for example.
2015 Travel Demand Model Documentation P a g e | 52 Appendix A -Acronyms
Appendix A. Acronyms
ADT – Average Daily Traffic BTPO – Bannock Transportation Planning Organization CBD – Central Business District GIS – Geographic Information System TAZ – Traffic Analysis Zones TDM - Transportation Demand Model VMT – Vehicle Miles Traveled
2015 Travel Demand Model Documentation P a g e | A-1 September 12, 2016 Appendix B- Model Structure and Contents
Appendix B. Model Structure and Contents
The Bannock Transportation Planning Organization Travel Demand Model was updated in 2012 by Parsons Brinckerhoff. Appendix B is the model structure and contents of chapter 1 and 2 from the BTPO Travel Model Users Guide. These two chapters cover the model structure and how to create scenarios for modeling future years or different development patterns.
Chapter 1
The Bannock model is delivered in a zip file called BANNOCK.zip. Figure 27 shows the list of files and folders contained in the model zip folder.
Figure 27: Model Structure
The following are the descriptions of the contents of the BANNOCK.zip file
1) Model Logo (bannock.bmp) - This is the model user interface image which is displayed on the Bannock toolbox, explained in detail in Model Setup. The user must copy this file to C:\Program Files\TransCAD6.0\bmp for TransCAD to display the logo in the Bannock toolbox.
2) 2010: Scenario folder - This is the base year scenario for which the model was developed and calibrated. In a few sections of the documentation, the model refers to the year 2011, since 2011 was the year of much of the data collection. However, the land use and socio-economic data were built from the 2010 Census data. The main model directory may also contain additional scenarios such as 2007, 2015, 2020 Alternative 1, 2020 Alternative 2 and so on. Each scenario contains two sub-folders: Input and Output, where the Output is an empty folder and Input has all the required model inputs such as highway networks, terminal times, zonal and land use data. 3) Rsc Folder - This folder contains the model GISDK code. Depending on the location of the model directory, the hard coded file location path inside the model GISDK code file needs to be updated as shown in Figure 28.
2015 Travel Demand Model Documentation P a g e | B-1 September 12, 2016 Appendix B- Model Structure and Contents
Figure 28: Model Code
4) TAZ Folder - This folder contains the zonal files that are used in the model. Although these files are not used in the model application, they are provided in the same package in order to produce thematic maps such as land use densities, area types, zone delineations or to display other geographic information.
5) Bannock.ini File - This file contains the list of file locations and the model directory path that are referenced throughout the model GISDK code. Figure 29 shows the contents of this file. Two of the listed files are not included in the package since they are created during the Bannock toolbox setup. The following files are defined:
A) Model Table (bannock_mod.bin): This file contains the list of stages (step), macros, the number of feedback loops, input files, and output files. The location of this file is specified here. See section 6) below for more information. B) UI File (Bannock_ui.dbd): This file is a compiled user interface file and is not provided in the package but is created during the Bannock toolbox setup. C) Scenario File (Bannock_scen.arr): This is a binary file that contains the list of scenarios and their input files. This file is also created during the Bannock toolbox setup. D) Data Directory: Location of the default model scenario.
Figure 29: Bannock.ini File
2015 Travel Demand Model Documentation P a g e | B-2 September 12, 2016 Appendix B- Model Structure and Contents
6) Bannock_mode.bin File - As mentioned above, this file contains the list of all model stages, macros, inputs, outputs, and model parameters. Figure 30 shows the structure and contents of this file.
Figure 30: Bannock_mod.bin File
A B
C
D
E
The structure of the file is divided into five sections, and each section contains six fields: ID, NAME, VALUE, DESCRIPTION, IN, and OUT. The sections are:
A) MODEL: This section defines the Bannock model user interface. A.1. NAME: BANNOCK Model is the name of the model and also the name of the bannock toolbox that is explained in B). A.2. VALUE: This is the Bannock logo file name. This is the same file that was described above under 1). A.3. DESCRIPTION: This field is not used in the application but only listed for description purposes. A.4. IN: The input to this section is a model compiled user interface which is described in detail in A). A.5. OUT: There are no output files for this stage B) STAGE: This section describes the model stages and order of stages. B.1. NAME: List of stage names
2015 Travel Demand Model Documentation P a g e | B-3 September 12, 2016 Appendix B- Model Structure and Contents
B.2. VALUE: Lists the associated image name that is displayed next to the steps in the Bannock toolbox (see Figure 33) B.3. DESCRIPTION: A brief description of the stage and its function B.4. IN: There are no inputs to this section B.5. OUT: There are no outputs to this section C) MACRO: This contains the list of macros and associated stages numbers C.1. NAME: List of macros that are coded in the rsc\bannock
2015 Travel Demand Model Documentation P a g e | B-4 September 12, 2016 Appendix B- Model Structure and Contents
Model Setup
1) Path Specification - The following files are required to be checked before setting up the model: A) Bannock.ini: The paths for the first item (model table) and last item (working directory) need to be checked. See Figure 29 for more information. B) Bannock_v
Figure 31: Bannock Toolbox Setup: Creating a UI file
2015 Travel Demand Model Documentation P a g e | B-5 September 12, 2016 Appendix B- Model Structure and Contents
B) Setup Add-ins: Once the model code is converted to the UI, it needs to be setup as an external plug-in in TransCAD. Figure 31 shows the steps to add the UI to TransCAD. B.1. Select Setup Add-Ins under the Tools menu. This opens a new window as shown in Figure 31 B.2. Click on the Add button to display a list of settings: B.2.1. Type: select Dialog box B.2.2. Description: type BANNOCK Model B.2.3. Name: type BANNOCK Model B.2.4. UI database: select file bannock_ui.dbd B.3. Click OK to create and add the Bannock toolbox to TransCAD B.4. Copy bannock.bmp to C:\Program Files\TransCAD 6.0\bmp
Figure 32: Bannock Toolbox Setup: Creating a UI file
2015 Travel Demand Model Documentation P a g e | B-6 September 12, 2016 Appendix B- Model Structure and Contents
C) Accessing the Bannock Toolbox: The following steps describe how to setup up a scenario using the Bannock toolbox. C.1. Scenario Settings: C.1.1. The newly added plug-in is accessed via Tool > Add-Ins..> BANNOCK Model as shown in Figure 33. C.1.2. When opened for the first time, all the stages under the toolbox are grayed out except for Setup. C.1.3. When Setup is clicked, TransCAD shows a message that there are no valid scenarios and asks if the user would like to set one up. Click YES to set up the scenario. C.1.4. Click on Add to setup a scenario. Then type a scenario name under Name. Click on Dir to browse to the scenario directory, for example, 2010. C.1.5. Select Stage and click on Contents to check all the required files. A list of input and output files are displayed for the selected stage. Column status shows whether that file exists or is missing in the directory. After everything is checked, click Ok to enable the Bannock toolbox stages.
C.2. Run Settings: There are three radio buttons under Run in the Bannock toolbox to select the type of model run to complete:
2015 Travel Demand Model Documentation P a g e | B-7 September 12, 2016 Appendix B- Model Structure and Contents
C.2.1. Stage: This option runs only the selected stage. For example, when the user selects this run type and clicks on Generation, the model runs only trip generation and then stops. C.2.2. Loop: This option runs the entire model for the selected loop. When the user selects this run type, they also need to specify the loop in the Start Feedback Loop drop down menu. The user need to remember two additional things: C.2.2.1. The macros are run only if the selected loop’s value is 1 for that macro (see C.2 or bannock_mod.bin). C.2.2.2. The model runs from the selected (clicked) stage onwards. C.2.3. All Loops: This option runs the entire model. However, the user can limit the number of loops by specifying the Max. Feedback Loops. The model stops when it reaches the set maximum loop.
Figure 33: Bannock Toolbox Setup: Setting up Scenarios
2015 Travel Demand Model Documentation P a g e | B-8 September 12, 2016