Method Selection for Travel Forecasting (2017)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

A-1 Methods Reference A P P E N D I X A Passenger Vehicle Models ..................................................................................................................A-2 Short Description...........................................................................................................................A-2 Methods ........................................................................................................................................A-2 Methods for Passenger Vehicle Models .......................................................................................A-2 Other Travel Demand Methods .........................................................................................................A-39 Short Description.........................................................................................................................A-39 Methods ......................................................................................................................................A-39 Methods for Other Travel Demand Methods ..............................................................................A-39 Commercial Vehicle Models ..............................................................................................................A-46 Short Description.........................................................................................................................A-46 Methods ......................................................................................................................................A-46 Methods for Commercial Vehicle Models ...................................................................................A-47 Assignment and Microsimulation Models ..........................................................................................A-72 Short Description.........................................................................................................................A-72 Methods ......................................................................................................................................A-72 Methods for Assignment and Microsimulation Models ...............................................................A-73 References ......................................................................................................................................A-107

A-2 Method Selection for Travel Forecasting: User Guide This reference guide replicates the online reference guide that is available within the online TFGuide tool. Each element of the TFGuide tool is described for reference and the relationships provided within the tool are provided. Each method also includes a longer description and a set of references. PASSENGER VEHICLE MODELS Short Description Passenger vehicle models include trip/tour generation, distribution, mode, and time of day model components for the movement of people. Methods Daily Activity Pattern Model Destination Choice Mode Choice (Disaggregate) Mode Choice (Fixed Factors) Parking Location Choice Population Synthesis Time of Day (Disaggregate) Time of Day (Fixed Factors) Tour Frequency Model Transit Pass Ownership Trip Attraction (Regression) Trip Distribution (Gravity) Trip Production (Cross-classification) Trip Production (Regression) Methods for Passenger Vehicle Models Daily Activity Pattern Model Short Description Person daily activity pattern models are a key component of activity-based modeling systems designed to realistically represent how people schedule activities in time and space. Long Description Person daily activity pattern models are often adopted because they consistently represent policy response across several dimensions. In daily activity pattern models, discrete choice models are used to represent the number of activities by purpose, their sequence, and their start or end time, among others. Daily activity pattern models include a component that simultaneously simulates the primary activity and general tour pattern for that activity, and the presence and number of other (secondary) tours by purpose. Subsequent components are used to schedule the tours

Methods Reference A-3 generated, populate details regarding exact numbers of stops, and so on. Recent approaches to daily activity pattern generation and schedule include simultaneous consideration and coordination between multiple household members in activity participation and scheduling. An alternative approach to the daily activity pattern model formulation uses a sequential decision tree process to generate daily activity schedules of individuals in a household. Hazard-based duration models and multiple continuous-discrete extreme value models have also been utilized in the context of activity scheduling to represent time along a continuous spectrum. Daily activity pattern models offer greater consistency in the generation and scheduling of activities, and greater sensitivity to land-use and transport network inputs than 4-step or tour- based models. However, daily activity pattern models are more computationally intensive than aggregate methods. These models also require a greater level of familiarity with disaggregate simulation procedures to develop and apply. Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Proven or validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Congestion Management Plan Energy Use Study Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Long-Range Transportation Plan Major Highway Corridor Age of Head of Household City City/County County Detailed Planning Highway Network Supply Household Size Household Socioeconomic Characteristics Income Groups Investment Grade Megaregion Microzone Number of Workers in Household Operational Planning Project Design Region Region Resident Passenger Travel Roadway Segment or Passenger Demand

A-4 Method Selection for Travel Forecasting: User Guide Study Major Transit Corridor Study Pricing Study Project Prioritization Traffic Impact Study Transit-Oriented Development Study Transportation Improvement Program Travel Demand Management Program Parcel State Subarea or Corridor Subregion Time Period Transit Network Supply Travel Analysis Zone Travel Distance Traveler Demographics Resources Lower-class Methods Method Relationships Activity-based Software Advanced Travel Forecasting Background Census Household Travel Survey Sociodemographic Estimates Standard Hardware Statistics Background Tour Frequency Model Trip Attraction (Regression) Trip Production (Cross- classification) Trip Production (Regression) Population Synthesis Transit Pass Ownership Destination Choice Trip Distribution (Gravity) Budget $40,000–$60,000 Schedule 3–6 months References Arentze, T., and H. Timmermans. 2004. “Albatross: A Learning-based Transportation Oriented Simulation System.” Transportation Research Part B Methodological, Vol. 38, No. 7, pp. 613– 633. Bhat, C.R., Guo, J.Y., Srinivasan, S., and A. Sivakumar. 2004. “Comprehensive Econometric Microsimulator for Daily Activity-Travel Patterns. ” Transportation Research Record: Journal of the Transportation Research Board, No. 1894, pp. 57–66. Bowman, J., and M. Ben-Akiva. 1997. “Activity-based Travel Forecasting.” Presented at the Activity-based Travel Forecasting Conference, Washington, DC. Bradley, M., and P. Vovsha. 2005. “A Model for Joint Choice of Daily Activity Pattern Types of Household Members.” Transportation, Vol. 32, No. 5, pp. 545–571.

Methods Reference A-5 RDC, Inc. 1995. “Activity-Based Modeling System for Travel Demand Forecasting.” US Department of Transportation, Washington, DC, Report DOT-T-96-02. Destination Choice Short Description Destination choice models are discrete choice models where choice alternatives are destination zones; take trips or tours by market segment and origin or production zone and choose a destination. Long Description Destination choice models are a class of discrete choice models in which the choice alternatives are destination zones. A destination choice model takes trips or tours by market segment and origin or production zone and chooses a destination. The most common form of destination choice model is the multinomial logit model. In the destination choice model, a utility is calculated for each potential destination as a function of its accessibility from the origin zone and its size (Utilityj = β*Accessibilityij + log[Sizej]). The size of the destination is typically calculated as a function of land-use characteristics that represent the number of opportunities to engage in each activity in the destination. For example, zone size for shopping trips might be a function of the number of retail jobs in the zone. Because the zone size is a quantity attribute of the zone, its log is taken in the utility calculation; this ensures that the probability of selection is proportional to the size of the zone, all else being equal. The accessibility term in the destination choice model is typically based on the mode choice logsum, which represents the utility of travel across all transport modes considered in the mode choice model. A model of this form can be considered as a nested mode and destination choice model where the mode choice is nested under the destination choice. A model formulated in this way has the advantage of demonstrating appropriate cross-elasticities across mode and destination. For example, the increase in utility for a specific mode between an origin and destination will result in an increase in trips on this mode equal to the sum of the trips switching from other modes in this zone pair and trips switching from other destinations to the destination whose utility was improved. Destination choice models provide a better behavioral basis for trip distribution by allowing for a wider range of explanatory variables than gravity models. These models are better able to reproduce observed travel patterns than gravity models. Destination choice models can be applied in either an aggregate or a disaggregate framework, and are appropriate for both trip- based models and activity-based models. In aggregate trip-based models, the number of zones and market segments (purposes and socioeconomic categories) considered by the model must be balanced with the computational burden of the model and runtime. The key advantage of disaggregate models is that a virtually unlimited number of market segments can be considered, though number of zones can still be an issue. However, sampling techniques are typically implemented in disaggregate destination choice models to reduce computational burden, allowing them to consider tens of thousands of zonal alternatives in a more systematic manner.

A-6 Method Selection for Travel Forecasting: User Guide Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Proven or validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Congestion Management Plan Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Long-Range Transportation Plan Major Highway Corridor Study Major Transit Corridor Study Pricing Study Project Prioritization Traffic Impact Study Transit-Oriented Development Study Transportation Improvement Program Travel Demand Management Program Age of Head of Household City Complex Transit Networks County Detailed Planning Employment Categories Establishment Characteristics Establishment Size Groups External Passenger Travel Fares Fixed Tolls General Purpose Lanes Highway Network Supply Household Size Household Socioeconomic Characteristics Income Groups Investment Grade Long-Distance Passenger Travel Managed Lanes Megaregion Microzone Number of Workers in Household Operating Costs Operational Planning Project Design Region Resident Passenger Travel Roadway Segment or Parcel Simple Transit Networks Accessibility Economic Development Passenger Mobility Schools User Benefits

Methods Reference A-7 Special Markets State Subarea or Corridor Time Period Transit Network Supply Travel Analysis Zone Travel Distance Traveler Demographics Travel Purpose Travel Time Value-of-Time Segments Resources Lower-class Methods Method Relationships Advanced Travel Forecasting Background Highway Networks Household Travel Survey Standard Hardware Statistics Background Transportation Network Planning Package Trip Distribution (Gravity) Transit Pass Ownership Trip Attraction (Regression) Trip Production (Cross- classification) Trip Production (Regression) Tour Frequency Model Daily Activity Pattern Model Mode Choice (Disaggregate) Time of Day (Disaggregate) Dynamic Traffic Assignment Multiclass Equilibrium Traffic Assignment Integrated Multiresolution Model Mode Choice (Fixed Factors) Time of Day (Fixed Factors) Traffic Microsimulation Stochastic User Equilibrium Assignment

A-8 Method Selection for Travel Forecasting: User Guide Budget $25,000–$60,000 Schedule 2–5 months References Ben-Akiva, M., and S. R. Lerman. 1985. Discrete Choice Analysis: Theory and Application to Travel Demand. The MIT Press. Bernardin, V. L., Koppelman, F., and D. Boyce. 2009. “Enhanced Destination Choice Models Incorporating Agglomeration Related to Trip Chaining While Controlling for Spatial Competition.” Transportation Research Record: Journal of the Transportation Research Board, No. 2132, pp. 143–151. Bhat, C., Govindarajan, A., and V. Pulugata. 1998. “Disaggregate Attraction-End Choice Modeling: Formulation and Empirical Analysis.” Transportation Research Record: Journal of the Transportation Research Board 1645, pp. 60–68. Borgers, A., and H. Timmermans. 1987. “Choice Model Specification, Substitution and Spatial Structure Effects: A Simulation Experiment.” Regional Science and Urban Economics, Vol. 17, No. 1, pp. 29–47. Chow, L.-F., Zhao, F., Li, M.-T., and S.-C. Li. 2005. “Development and Evaluation of Aggregate Destination Choice Models for Trip Distribution in Florida.” Transportation Research Record: Journal of the Transportation Research Board, Vol. 1931, pp. 18–27. Daly, A. 1982. “Estimating Choice Models Containing Attraction Variables.” Transportation Research Part B: Methodological, Vol. 16, No. 1, pp. 5–15. Fotheringham, A. S. 1983. “Some Theoretical Aspects of Destination Choice and Their Relevance to Production-Constrained Gravity Models.” Environment and Planning, Vol. 15, No. 8, pp. 1121–1132. Jonnalagadda, N., Freedman, J., Davidson, W. A., and J. D. Hunt. 2001. “Development of Microsimulation Activity-Based Model for San Francisco: Destination and Mode Choice Models.” Transportation Research Record: Journal of the Transportation Research Board, No. 1777, pp. 25–35. Mishra, S., Wang, Y., Zhu, X., Moeckel, R., and S. Mahaparta. 2013. “Comparison between Gravity and Destination Choice Models for Trip Distribution in Maryland.” Presented at Transportation Research Board 92nd Annual Meeting, Washington, DC.

Methods Reference A-9 Mode Choice (Disaggregate) Short Description Person mode choice (disaggregate) models estimate the demand for and number of trips and tours on various modes of transportation—auto, walk, bike, and transit are the most common. Long Description Person mode choice (disaggregate) models estimate the demand for and number of trips and tours on various modes of transportation—auto, walk, bike, and transit are the most common. These models date back to the 1960s, and are based on discrete choice theory. The theory states that travelers are rational decision-makers who ascribe a utility to each choice alternative and choose the alternative with the highest utility. The models are probabilistic rather than deterministic since there are errors in the calculation of utility; these models calculate a choice probability for each alternative. In mode choice models, the alternatives are transport modes (e.g., drive alone, share a ride, walk or bike, walk to transit, drive to transit). The utility of each mode is a function of its attributes (the time and cost of using the mode), the attributes of the decision-maker (e.g., income, auto ownership, age) and the attributes of the trip itself (e.g., purpose, time of day). The most common form of the discrete choice model used for mode choice is the logit model. In the multinomial version of the logit model, each alternative competes equally; a change in the utility of one mode will cause equal changes in the probability of each competing mode. A more mathematically advanced form of the model—the nested logit model—permits unequal competition across modes. In a nested logit model, modes are grouped into nests containing similar modes. The cross-elasticities across nested modes are higher than the cross-elasticities with non-nested modes, resulting in a higher percentage change in probability for similar modes. Mode choice models can be applied in either an aggregate or a disaggregate framework, and are appropriate for both trip-based models and activity-based models. In aggregate trip-based models, the number of market segments (purposes and socioeconomic categories) considered by the model must be balanced with the computational burden of the model and runtime. The key advantage of disaggregate models is that a virtually unlimited number of explanatory variables can be considered. Since most transportation planning studies involve the analysis of multimodal alternatives, mode choice models are a key component of most travel demand models. Mode choice models are an essential tool for estimation of transit demand, mode shifts due to congestion, introduction of new modes, parking and other pricing policies, and many other aspects of transportation demand analysis. Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Proven or validated

A-10 Method Selection for Travel Forecasting: User Guide Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Bicycle and Pedestrian Capital Investments Bicycle and Pedestrian Plan Congestion Management Plan Economic Impact Analysis Energy Use Study Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Greenhouse Gas Mitigation Study Health and Physical Activity Plan Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Long-Range Transportation Plan Major Highway Corridor Study Major Transit Corridor Study Pricing Study Project Prioritization Sustainable Community Strategies Traffic Impact Study Transit Operations Study City Complex Transit Networks County Daily Detailed Planning Dynamic Tolls Fares Fixed Tolls General Purpose Lanes Highway Network Supply Household Size Household Socioeconomic Characteristics Income Groups Investment Grade Managed Lanes Megaregion Microzone Operating Costs Operational Planning Pedestrian/Bike Facilities Pedestrian/Bike Network Supply Project Design Region Resident Passenger Travel Simple Transit Networks Special Markets State Time Period Transit Network Supply Travel Analysis Zone Travel Demand Management Policies Travel Distance Traveler Demographics Travel Purpose Travel Time Active Transportation Passenger Mode Shares Transit Measures

Methods Reference A-11 Transit-Oriented Development Study Transportation Improvement Program Travel Demand Management Program Value-of-Time Segments Vehicle Occupancy Groups Resources Lower-class Methods Method Relationships Census Highway Networks Standard Hardware Standard Travel Forecasting Background Statistics Background Transit Networks Transit On-Board Survey Transportation Network Planning Package Direct-Demand Model Mode Choice (Fixed Factors) Multiclass Equilibrium Traffic Assignment Parking Location Choice Destination Choice Trip Distribution (Gravity) Time of Day (Disaggregate) Dynamic Capacity- Constrained Transit Assignment Dynamic Traffic Assignment Frequency-based Transit Assignment Transit Pass Ownership Integrated Multiresolution Model Traffic Microsimulation Schedule-based Transit Assignment Integrated Multiresolution Model Pedestrian/Bicycle Simulation Time of Day (Fixed Factors) Stochastic User Equilibrium Assignment Budget $25,000–$100,000 Schedule 3–6 months

A-12 Method Selection for Travel Forecasting: User Guide References Ben-Akiva, M., and S. R. Lerman. 1985. Discrete Choice Analysis: Theory and Application to Travel Demand. The MIT Press. Bhat, C.R. 2000. “Incorporating Observed and Unobserved Heterogeneity in Urban Work Mode Choice Modeling.” Transportation Science, Vol. 34, No. 2, pp. 228–238. Hensher, D., Rose, J. M., and W. H. Greene. 2005. Applied Choice Analysis: A Primer. Cambridge University Press. Koppelman, F.S., and C. Bhat. 2006. A Self-Instructing Course in Mode Choice Modeling: Multinomial and Nested Logit Models. US Department of Transportation, Federal Transit Administration. Train, K. 2003. Discrete Choice Methods with Simulation. Cambridge University Press. Mode Choice (Fixed Factors) Short Description Person mode choice models (fixed factors) estimate mode choice based on current observed shares. Long Description Person mode choice models (fixed factors) assume that mode shares do not change over time. Fixed factors are estimated based on observed mode shares, segmented by trip purpose and sometimes by time period. These models may be suitable for areas where ridesharing and transit use is low, or where transit is primarily used by captive riders. Model factors can be derived from household travel surveys. Method Type and Performance Consistency Method type: Aggregate Performance consistency: Proven or validated

Methods Reference A-13 Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Congestion Management Plan Energy Use Study Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Major Highway Corridor Study Traffic Impact Study Transit Operations Study Transit-Oriented Development Study Travel Demand Management Program City City/County County Daily High-Level Planning Income Groups Operational Planning Region Region Resident Passenger Travel Subarea or Corridor Subregion Time Period Traveler Demographics Travel Purpose Value-of-Time Segments Passenger Mode Shares Resources Lower-class Methods Method Relationships Census Standard Hardware Standard Travel Forecasting Background Traffic Counts Transit Counts Transportation Network Planning Package Multiclass Equilibrium Traffic Assignment Time of Day (Disaggregate) Time of Day (Fixed Factors) Dynamic Capacity- Constrained Transit Assignment Frequency-based Transit

A-14 Method Selection for Travel Forecasting: User Guide Assignment Trip Distribution (Gravity) Schedule-based Transit Assignment Destination Choice Single-class Equilibrium Traffic Assignment Stochastic User Equilibrium Assignment Budget $10,000–$20,000 Schedule 1–3 months References Cambridge Systematics, Inc., Vanasse Hangen Brustlin, Inc., Gallop Corporation, Bhat, C., Shapiro Transportation Consulting, LLC, Martin/Alexiou/Bryson, PLLC. 2012. NCHRP Report 716: Travel Demand Forecasting: Parameters and Techniques. Transportation Research Board, Washington, DC. Parking Location Choice Short Description Parking location choice models are used to determine the parking location of auto trips, typically in a parking-constrained part of the region. Long Description Parking location choice models are used to determine the parking location of auto trips, typically in a parking-constrained part of the region. These models are particularly useful for analyzing policies related to parking price and supply, area congestion pricing, or transit services, such as “downtown people movers” or parking lot shuttles. Parking lot choice models typically take the form of discrete choice models, where a probability is estimated for each potential parking lot based on its relative utility. The utility of the lot can be based on the price and availability of parking at the lot, the auto time or distance from the origin to the lot, the walk or transit time from the lot to the destination, and attributes of the decision- maker and trip purpose. In the context of airport parking location choice, the trip duration (number of days for parking) can also be considered, particularly if implemented within a disaggregate framework. In the context of university parking lot choice, the status of the traveler (student versus faculty\staff) is important to consider given parking supply restrictions.

Methods Reference A-15 The advantages of parking location choice are more accurate network loadings in parking- constrained areas and model sensitivities to parking price and supply. They are particularly useful when analyzing demand response to the location and price of parking, or transit services that serve parking lots. However, parking location choice models require observed input data on parking price and quantity; demand data on current parking behavior is useful for model calibration, but it is not required. Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Proven or validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Arterial Operations and Management Study Comprehensive Plans Congestion Management Plan Environmental Clearance and Preliminary Design for Transportation Projects Environmental Justice Plan Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Long-Range Transportation Plan Major Highway Corridor Study Major Transit Corridor Study Pricing Study Traffic Impact Study Transportation Improvement Program Travel Demand 15-minutes or less City County Daily Detailed Planning Establishment Characteristics Establishment Size Groups General Purpose Lanes Highway Network Supply Hour Household Socioeconomic Characteristics Megaregion Microzone Operational Planning Project Design Region Resident Passenger Travel State Subarea or Corridor Time Period Transit Network Supply Travel Analysis Zone Travel Distance Traveler Demographics Travel Purpose

A-16 Method Selection for Travel Forecasting: User Guide Management Program Resources Lower-class Methods Method Relationships Activity-based Software Custom Programming Model Estimation Skills Parking Supply Inventory Standard Hardware Standard Travel Forecasting Background Statistics Background Mode Choice (Disaggregate) Dynamic Traffic Assignment Traffic Microsimulation Budget $15,000–$30,000 Schedule 2–5 months References Alsup, R., Freedman, J., Bettinardi, A., and S. Payne. 2014. “A University Student Tour-based Travel Demand Model.” Presented at the Innovations in Travel Demand Forecasting Conference, Baltimore, MD. Gosling, G. 2008. ACRP Synthesis of Airport Practice 5: Airport Ground Access Mode Choice Models. Transportation Research Board, Washington, DC. Hunt, J. D., and S. Teply. 1993. “A Nested Logit Model of Parking Location Choice.” Transportation Research Part B: Methodological, Vol. 27, No. 4, pp. 253–265. Population Synthesis Short Description Population synthesis is a method used to generate a database of individual households and people that are the foundation of disaggregate model applications such as tour and activity-based models. Long Description Population synthesis is a method used to generate a database of individual households and people that are the foundation of disaggregate model applications such as tour and activity-based models. The household and person agents that make up the synthetic population make decisions in disaggregate models. The results of these decisions are aggregated to provide information to decision-makers regarding the costs and benefits of transportation, land-use, and other investments and policies considered by the models.

Methods Reference A-17 In population synthesis, a synthetic population is created where key characteristics of the population match control totals specified by an analyst. These control totals can be specified at the household level and, in newer population synthesis procedures, at the person level. Often the control totals are specified for each Transportation Analysis Zone, but new procedures also allow the user to specify controls at varying levels of geography (regional, tract, zonal, and/or subzonal). Data sources for control totals are like, but perhaps more extensive than, aggregate 4- step model inputs. Base-year data sources for controls include American Community Survey three- and five-year averages, Census Public Use Microdata Samples (PUMS), and local, regional, and state agency demographic and housing data. Future-year data sources include city, regional, and state government forecasts of population and housing, as well as socioeconomic and demographic models. Population synthesis procedures can take several mathematical forms, but typically involve two separate but related processes—fitting and allocation. The fitting procedure estimates the number of households required to meet the geographic and socioeconomic controls. The allocation procedure involves drawing those populations from disaggregate data. Early fitting procedures utilize a process known as Iterative Proportional Fitting (IPF). In this procedure, a seed table of households by socioeconomic attributes is created from observed data (typically PUMS) and “balanced” to control totals for each TAZ using row and column balancing. The allocation procedure utilizes a Monte Carlo simulation technique to choose a representative set of households from an input data source (typically PUMS data) without replacement to populate the synthetic household and person database. List-based approaches have been used in population synthesis; this permits greater flexibility in the specification of controls, the extension of controls to consider both household and person constraints, and multiple geographies. Linear optimization algorithms have been introduced to speed up the convergence of the fitting algorithm. Combinatorial optimization methods have also been proposed in which the IPF and allocation processes are combined into one method of iteratively generating a population and swapping out households and people to improve goodness of fit to control totals. The advantage of population synthesis (and disaggregate models in general) is that model results (e.g., number of tours/trips generated, destinations and modes chosen, times of day of travel) can be aggregated along any dimension available in the synthetic population. For example, if an analyst wishes to summarize model simulations of trips by mode made by low-income single- parent households, s/he can do so if these variables are included in the synthetic population. This provides significant opportunities for performing equity and “winners and losers” analysis. However, disaggregate models are stochastic as opposed to deterministic like aggregate 4-step models. A deterministic model provides an expected or average value, while a stochastic model produces an outcome that is one point along a distribution of potential results. The variance, or difference, between the stochastic outcome and the expected or average value is referred to as Monte Carlo error, and this error is inversely related to the number of agents simulated. The user must consider this error when summarizing results, or else run the model several times and average the results before reporting.

A-18 Method Selection for Travel Forecasting: User Guide Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Proven or validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Congestion Management Plan Economic Impact Analysis Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Greenhouse Gas Mitigation Study Health and Physical Activity Plan Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Long-Range Transportation Plan Major Highway Corridor Study Major Transit Corridor Study Pricing Study Project Prioritization Sustainable Community Strategies Traffic Impact Study Transit-Oriented Age of Head of Household Annual City City/County County Detailed Planning High-Level Planning Household Size Household Socioeconomic Characteristics Income Groups Investment Grade Long-Distance Passenger Travel Megaregion Microzone Number of Workers in Household Operational Planning Project Design Region Region Resident Passenger Travel Roadway Segment or Parcel Special Markets State Subarea or Corridor Travel Analysis Zone Traveler Demographics Environmental Justice/Social Equity

Methods Reference A-19 Development Study Transportation Improvement Program Travel Demand Management Program Resources Lower-class Methods Method Relationships Advanced Travel Forecasting Background Census Custom Programming Sociodemographic Estimates Standard Hardware Statistics Background Daily Activity Pattern Model Transit Pass Ownership Tour Frequency Model Budget $20,000–$60,000 Schedule 2–8 months References Abraham, J.E., Stefan, K.J., and J.D. Hunt. 2012. “Population Synthesis Using Combinatorial Optimization at Multiple Levels.” Presented at the Transportation Research Board 91st Annual Meeting, Washington, DC. Beckman, R.J., K.A. Baggerly, and M.D. McKay. 1996. “Creating Synthetic Baseline Populations.” Transportation Research Part A: Policy and Practice, Vol. 30, No. 6, pp. 415–429. Müller, K., and K. W. Axhausen. 2010. “Population Synthesis for Microsimulation: State of the Art.” ETH Zürich, Institut für Verkehrsplanung, Transporttechnik, Strassen-und Eisenbahnbau (IVT). Vovsha, P., Hicks, J. E., Paul, B. M., Livshits, V., Maneva, P., and K. Jeon. 2014. “New Features of Population Synthesis.” Presented at the Transportation Research Board 94th Annual Meeting, Washington, DC. Williamson, P., Birkin, M., and P.H. Rees. 1998. “The Estimation of Population Microdata by Using Data from Small Area Statistics and Samples of Anonymised Records.” Environment and Planning A, Vol. 30, No. 5, pp. 785–816.

A-20 Method Selection for Travel Forecasting: User Guide Time of Day (Disaggregate) Short Description Person time of day choice models are used to simulate the share of trips or tours by time period or hour. Long Description Person time of day choice models are used to simulate the share of trips or tours by time period or hour. The periods could be aggregations of multiple hours (peak vs. off-peak, or AM, Midday, PM, and evening), discrete hours or half-hours, or even continuous time. Early examples of time of day choice models were largely focused on peak spreading (i.e., trips were shifted between the peak hour and shoulders of the peak based on congestion levels). Similar approaches have been applied in other trip-based modeling contexts. With the emergence of practical large-scale activity-based models in the late 1990s, time-of-day choice models evolved to consider many more time periods and even schedule activities in continuous time. A key disadvantage of trip- based time-of-day choice models, which has limited their usefulness, is that the duration of activities is not considered. By considering both the start and end time of the activity, the utility of engaging in an activity for a certain period and the institutional and other constraints imposed by start and end times of specific activities (such as the need to work eight hours per day during regular office hours) can be considered. A key advantage of time-of-day choice modeling in the context of tour- and activity-based simulation models is that they can consider many more variables—including not only the attributes of the traveler but also the constraints imposed by other household members. Time-of-day choice models are important for testing policies that may result in a change to the timing of travel (e.g., tolls that vary by time-of-day, variable parking pricing policies, transit services that only operate in peak periods). Further, time-of-day choice models are useful in large regions where congestion causes some travelers to shift their departure times. Time-of-day models are important when linking travel demand model output to a traffic microsimulation or dynamic traffic assignment model. However, trip-based time-of-day choice models are fairly limited in the variables that they can consider and do not consider activity duration. Time of day choice model estimation and calibration requires data on travel patterns by the time periods of interest. Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Proven or validated

Methods Reference A-21 Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Arterial Operations and Management Study Comprehensive Plans Congestion Management Plan Energy Use Study Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Freeway Operations and Management Study Greenhouse Gas Mitigation Study Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Long-Range Transportation Plan Major Highway Corridor Study Major Transit Corridor Study Pricing Study Project Prioritization Safety Program Sustainable Community Strategies Traffic Impact Study Transit Operations Study Transit-Oriented 15-minutes or less Age of Head of Household City City/County Complex Transit Networks County Detailed Planning Dynamic Tolls Fares Fixed Tolls General Purpose Lanes Highway Network Supply Highway Operations Hour Household Size Household Socioeconomic Characteristics Income Groups Investment Grade Long-Distance Passenger Travel Managed Lanes Megaregion Microzone Number of Workers in Household Operating Costs Operational Planning Project Design Region Region Resident Passenger Travel Roadway Segment or Parcel Simple Transit Networks Special Markets State Subregion Passenger Demand

A-22 Method Selection for Travel Forecasting: User Guide Development Study Transportation Improvement Program Travel Demand Management Program Time Period Travel Analysis Zone Travel Distance Traveler Demographics Travel Purpose Travel Time Resources Lower-class Methods Method Relationships Advanced Travel Forecasting Background Household Travel Survey Sociodemographic Estimates Standard Hardware Statistics Background Traffic Counts Transportation Network Planning Package Time of Day (Fixed Factors) Multiclass Equilibrium Traffic Assignment Mode Choice (Disaggregate) Mode Choice (Fixed Factors) Destination Choice Trip Distribution (Gravity) Dynamic Capacity- Constrained Transit Assignment Dynamic Traffic Assignment Frequency-based Transit Assignment Integrated Multiresolution Model Traffic Microsimulation Integrated Multiresolution Model Pedestrian/Bicycle Simulation Schedule-based Transit Assignment Budget $20,000–$50,000 Schedule 2–5 months References Allen Jr., W.G., and G.W. Schultz. 1996. “Congestion-Based Peak Spreading Model.” Transportation Research Record 1556, pp. 8–15.

Methods Reference A-23 Bhat, C., and J. Steed. 2002. “A Continuous-Time Model of Departure Time Choice for Urban Shopping Trips.” Transportation Research Part B: Methodological, Vol. 36, No. 3, pp. 207–224. Cambridge Systematics, Inc. 1999. “Time-of-Day Modeling Procedures: State-of-the-Art, State- of-the-Practice,” DOT-T-99-01, US Department of Transportation, Washington, DC. https://www.fhwa.dot.gov/planning/tmip/publications/other_reports/tod_modeling_procedures/c h01.cfm (Accessed on Dec. 21, 2016). McFadden, D., Talvitie, A., Cosslett, S., Hasan, I., Johnson, M., Reid, A., and K. Train. 1977. “Demand Model Estimation and Validation.” The Urban Travel Demand Forecasting Project, Volume V. Vovsha, P., and M. Bradley. 2004. “Hybrid Discrete Choice Departure Time and Duration Model for Scheduling Travel Tours.” Transportation Research Record: Journal of the Transportation Research Board, No. 1894, pp. 46–56. Vovsha, P., Petersen, E., and R. Donnelly. 2003. “Explicit Modeling of Joint Travel by Household Members: Statistical Evidence and Applied Approach.” Transportation Research Record: Journal of the Transportation Research Board, No. 1831, pp. 1–10. Time of Day (Fixed Factors) Short Description Person time-of-day models percent daily trips active factors and are not sensitive to congestion. Long Description Person time-of-day models convert production-attraction trip tables to origin-destination trip tables by time of day. These models require observed data from household travel surveys to calculate time-of-day factors by trip purpose. Peak-period factors are typically estimated by counting percentage of "trips in motion" during the peak. These models can also be estimated from traffic counts (applicable to all trip purposes). Method Type and Performance Consistency Method type: Aggregate Performance consistency: Proven or validated

A-24 Method Selection for Travel Forecasting: User Guide Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Arterial Operations and Management Study Congestion Management Plan Energy Use Study Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Freeway Operations and Management Study Greenhouse Gas Mitigation Study Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Long-Range Transportation Plan Major Highway Corridor Study Major Transit Corridor Study Safety Program Sustainable Community Strategies Traffic Impact Study Transit Operations Study Transit-Oriented Development Study Travel Demand Management Program City County External Passenger Travel General Purpose Lanes High-Level Planning Hour Long-Distance Passenger Travel Managed Lanes Operational Planning Region Resident Passenger Travel Subarea or Corridor Time Period Travel Analysis Zone Travel Purpose Passenger Demand

Methods Reference A-25 Resources Lower-class Methods Method Relationships Standard Hardware Standard Travel Forecasting Background Traffic Counts Transportation Network Planning Package Multiclass Equilibrium Traffic Assignment Mode Choice (Fixed Factors) Dynamic Capacity- Constrained Transit Assignment Frequency-based Transit Assignment Schedule-based Transit Assignment Mode Choice (Disaggregate) Destination Choice Trip Distribution (Gravity) Origin-Destination Matrix Estimation Single-class Equilibrium Traffic Assignment Stochastic User Equilibrium Assignment Budget $10,000–$20,000 Schedule 1–2 months References Cambridge Systematics, Inc., Vanasse Hangen Brustlin, Inc., Gallop Corporation, Bhat, C., Shapiro Transportation Consulting, LLC, Martin/Alexiou/Bryson, PLLC. 2012. NCHRP Report 716: Travel Demand Forecasting: Parameters and Techniques. Transportation Research Board, Washington, DC. Tour Frequency Model Short Description Person tour frequency models simulate the number of tours, typically by purpose, for each household or person in a model area. A tour is a sequence of trips that start and end at home or at work.

A-26 Method Selection for Travel Forecasting: User Guide Long Description Tour frequency models can be implemented as a component of an activity-based travel modeling system, or as a stand-alone component of a tour-based modeling system. The key difference between a tour-based modeling system and an activity-based modeling system is that an activity- based modeling system explicitly recognizes temporal and spatial constraints on the generation and scheduling of activities. In other words, the activity-based model recognizes that an individual can be in only one place at one time and will not schedule overlapping activities. A tour-based model does not necessarily observe this constraint. Most tour frequency models are based on utility maximization theory and implemented within a discrete choice modeling framework. One of the earliest tour frequency models attempted to represent the number, purpose, destination and “main” mode of tours, but used aggregate categories for each choice to make the model computationally tractable. One of the first operational large-scale tour frequency models was developed for the Dutch Ministry of Transport and Public Works in the mid-1980s and was then expanded and applied to several other regions in Europe. In the United States, one of the first operational tour frequency models was developed for New York Metropolitan Transportation Commission. They have subsequently been implemented in the context of activity-based models or within simpler tour-based modeling systems. Tour frequency models offer the ability to consider many more socioeconomic, network, and land-use variables on the generation of travel than do aggregate trip-based models. These models also provide the opportunity to consistently model the location, sequencing, and mode choice of stops on tours. When coupled with a daily activity pattern model, tour frequency models offer the most consistent treatment for the generation of travel. Some of the advantages of activity- based models, such as consistent scheduling of multiple tours, are not realized when these models are developed as stand-alone components. The development of tour frequency models requires more resources than simple trip generation rates. The application of these models within a simulation paradigm introduces Monte Carlo variation in the results. They are less complex than activity-based models and are occasionally used to represent auxiliary travel such as long- distance, truck, visitor, or university travel. Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Proven or validated

Methods Reference A-27 Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Congestion Management Plan Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Long-Range Transportation Plan Major Highway Corridor Study Major Transit Corridor Study Pricing Study Project Prioritization Traffic Impact Study Transit-Oriented Development Study Transportation Improvement Program Travel Demand Management Program Age of Head of Household City City/County County Detailed Planning Highway Network Supply Household Size Household Socioeconomic Characteristics Income Groups Investment Grade Megaregion Microzone Number of Workers in Household Operational Planning Project Design Region Region Resident Passenger Travel Roadway Segment or Parcel Special Markets State Subarea or Corridor Subregion Transit Network Supply Travel Analysis Zone Travel Distance Traveler Demographics Travel Purpose Passenger Demand Resources Lower-class Methods Method Relationships Advanced Travel Forecasting Background Census Household Travel Survey Trip Attraction (Regression) Trip Production (Cross- classification) Destination Choice Population Synthesis Trip Distribution (Gravity)

A-28 Method Selection for Travel Forecasting: User Guide Sociodemographic Estimates Standard Hardware Statistics Background Transportation Network Planning Package Trip Production (Regression) Budget $20,000–$40,000 Schedule 1–3 months References Adler, T., and M. Ben-Akiva. 1977. “A Theoretical and Empirical Model of Trip Chaining Behavior.” Transportation Research B, Vol. 13B, pp. 243–257. Fox, J., Daly, A., and H. Gunn. 2005. “Review of RAND Europe’s Transport Demand Model Systems.” Prepared for TRL Limited by RAND Europe. https://www.rand.org/content/dam/rand/pubs/monograph_reports/2005/MR1694.pdf (As of Dec. 12, 2016). Gunn, H. 1994. “The Netherlands National Model: A Review of Seven Years of Application.” International Association of Operational Research, Vol.1, No.2 pp. 125–133. Vovsha, P., and K. A. Chiao. 2008. “Development of New York Metropolitan Transportation Council Tour-based Model.” TRB Conference Proceedings 42: Innovations in Travel Demand Modeling, Vol. 2. Transportation Research Board, Washington, DC, pp. 21–23. Transit Pass Ownership Short Description Transit pass ownership models are used to simulate whether each traveler has a transit pass and can take transit at a discounted price (or for free) compared to a per-trip cash fare. Long Description Transit pass ownership models can improve the accuracy of mode choice models to changes in transit fare, and can result in more accurate estimates of transit revenue. Transit pass ownership models are typically implemented as a long- or medium-term “mobility” component within a tour or activity-based travel modeling system. These models are usually discrete choice models whose alternatives may be a binary choice of transit pass ownership, a multinomial choice of various types of transit passes, or combinations of transit pass ownership, automobile ownership, and/or bicycle ownership. Transit pass ownership models typically consider the cost of mobility options and the accessibility of the household or traveler to destinations via each option.

Methods Reference A-29 Transit pass ownership models provide long- and medium-term elasticity responses to the cost of transit passes. However, these models require aggregate data on transit pass ownership for calibration and disaggregate data for model estimation. These requirements add to the complexity of the model system. Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Not proven/not validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Comprehensive Plans Environmental Clearance and Preliminary Design for Transportation Projects Environmental Justice Plan Long-Range Transportation Plan Major Transit Corridor Study Pricing Study Project Prioritization Transit Operations Study Transportation Improvement Program Travel Demand Management Program Age of Head of Household City County Detailed Planning Household Socioeconomic Characteristics Income Groups Megaregion Microzone Number of Workers in Household Operational Planning Project Design Region Resident Passenger Travel Roadway Segment or Parcel Season/Month Simple Transit Networks State Subarea or Corridor Transit Network Supply Travel Analysis Zone Traveler Demographics Resources Lower-class Methods Method Relationships Activity-based Software Advanced Travel Forecasting Background Daily Activity Pattern Model Destination Choice

A-30 Method Selection for Travel Forecasting: User Guide Household Travel Survey Standard Hardware Statistics Background Transit On-Board Survey Population Synthesis Mode Choice (Disaggregate) Budget $5,000–$10,000 Schedule 1–3 months References Litman, T. 2004. “Transit Price Elasticities and Cross-Elasticities.” Journal of Public Transportation, Vol. 7, No. 2., pp. 37–58. Scott, D. M., and K. W. Axhausen. 2006. “Household Mobility Tool Ownership: Modeling Interactions Between Cars and Season Tickets.” Transportation, Vol. 33, No. 4, pp. 311–328. Vovsha, P., and E. Petersen. 2009. “Model for Person and Household Mobility Attributes.” Transportation Research Record: Journal of the Transportation Research Board, No. 2132, pp. 95–105. Trip Attraction (Regression) Short Description Person trip attraction (regression) models estimate trip attractions as a function of activity in a traffic analysis zone. Long Description Person trip attraction (regression) models rely on household travel survey data to estimate parameters. These models are better statistically if zones are grouped together to achieve enough samples in the household travel survey required for model estimation. Method Type and Performance Consistency Method type: Aggregate Performance consistency: Somewhat proven

Methods Reference A-31 Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Congestion Management Plan Economic Development Plan Economic Impact Analysis Energy Use Study Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Greenhouse Gas Mitigation Study Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Long-Range Transportation Plan Major Highway Corridor Study Major Transit Corridor Study Pricing Study Project Prioritization Sustainable Community Strategies Traffic Impact Study Transit-Oriented Development Study Transportation Improvement Program City County Detailed Planning Dynamic Tolls Employment Categories Establishment Size Groups External Passenger Travel Fixed Tolls High-Level Planning Household Socioeconomic Characteristics Income Groups Investment Grade Megaregion Operating Costs Operational Planning Project Design Region Resident Passenger Travel State Subarea or Corridor Travel Analysis Zone Travel Demand Management Policies Travel Purpose Passenger Demand

A-32 Method Selection for Travel Forecasting: User Guide Travel Demand Management Program Resources Lower-class Methods Method Relationships Census Employment Data Sociodemographic Estimates Standard Hardware Standard Travel Forecasting Background Transportation Network Planning Package Destination Choice Trip Distribution (Gravity) Trip Production (Regression) Trip Production (Cross- classification) Budget $10,000–$20,000 Schedule 1–2 months References Meyer, M., and E. J. Miller. 2000. Urban Transportation Planning. McGraw-Hill. Trip Distribution (Gravity) Short Description Person trip distribution (gravity) models produce a trip table by production and attraction zone. Long Description Person trip distribution (gravity) models produce a trip table by production and attraction zone. These models illustrate the macroscopic relationships between places (e.g., homes, workplaces). The interaction between two locations declines with increasing distance, time, and cost between them, but is positively associated with the amount of activity at each location. Gravity models distribute trips between zones based on the following formula: number of productions multiplied by the number of attractions multiplied by a "friction factor" that represents travel times between zones. Gravity model estimation procedure consists of estimating friction factors as a function of travel time. Travel times are typically based solely on auto travel times, but these can be based on logsums from a mode choice model (travel times weighted by mode).

Methods Reference A-33 Method Type and Performance Consistency Method type: Aggregate Performance consistency: Proven or validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Comprehensive Plans Congestion Management Plan Energy Use Study Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Greenhouse Gas Mitigation Study Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Long-Range Transportation Plan Major Highway Corridor Study Major Transit Corridor Study Pricing Study Project Prioritization Sustainable Community Strategies Traffic Impact Study Transit-Oriented Development Study Transportation Improvement Program Travel Demand Management Program City City/County County Detailed Planning Fixed Tolls High-Level Planning Highway Network Supply Household Socioeconomic Characteristics Income Groups Megaregion Operating Costs Operational Planning Region Region Resident Passenger Travel State Subarea or Corridor Subregion Time Period Travel Analysis Zone Travel Distance Travel Purpose Travel Time Economic Development Passenger Mobility User Benefits

A-34 Method Selection for Travel Forecasting: User Guide Resources Lower-class Methods Method Relationships Highway Networks Standard Hardware Standard Travel Forecasting Background Statistics Background Transportation Network Planning Package Multiclass Equilibrium Traffic Assignment Mode Choice (Disaggregate) Time of Day (Disaggregate) Trip Attraction (Regression) Trip Production (Regression) Trip Production (Cross- classification) Mode Choice (Fixed Factors) Tour Frequency Model Daily Activity Pattern Model Time of Day (Fixed Factors) Single-class Equilibrium Traffic Assignment Stochastic User Equilibrium Assignment Budget $10,000–$20,000 Schedule 1–3 months References Allen, Jr., W.G. 1983. “Trip Distribution Using Composite Impedance.” Transportation Research Record 944, pp. 118–127. Meyer, M., and E.J. Miller. 2000. Urban Transportation Planning. McGraw-Hill. Wilson, A.G. 1967. “A Statistical Theory of Spatial Distribution Models.” Transportation Research, Vol. 1, pp. 253–269.

Methods Reference A-35 Trip Production (Cross-classification) Short Description Person trip production (cross-class) models classify households by one or more socioeconomic factors and then estimate household trip productions by trip purpose for each classification. Long Description Person trip production (cross-class) models classify households by one or more socioeconomic factors and then estimate household trip productions by trip purpose for each classification. Examples of categories include the following: Household size Household income Auto ownership Age of head of household Number of workers in household Household travel surveys are the most common data source. Cross-classifications should be chosen so that there are enough (minimum of 50) observations in each cross-classification cell. Cells may be combined if there are insufficient numbers of observations in each cell. Method Type and Performance Consistency Method type: Aggregate Performance consistency: Proven or validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Bicycle and Pedestrian Capital Investments Bicycle and Pedestrian Plan Comprehensive Plans Congestion Management Plan Economic Development Plan Economic Impact Analysis Energy Use Study Age of Head of Household City County Detailed Planning High-Level Planning Household Size Household Socioeconomic Characteristics Income Groups Investment Grade Long-Distance Passenger Travel Megaregion Microzone Passenger Demand

A-36 Method Selection for Travel Forecasting: User Guide Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Greenhouse Gas Mitigation Study Health and Physical Activity Plan Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Long-Range Transportation Plan Major Highway Corridor Study Major Transit Corridor Study Pricing Study Project Prioritization Sustainable Community Strategies Traffic Impact Study Transit-Oriented Development Study Transportation Improvement Program Travel Demand Management Program Number of Workers in Household Operational Planning Project Design Region Resident Passenger Travel State Subarea or Corridor Time Period Travel Analysis Zone Travel Purpose Resources Lower-class Methods Method Relationships Census Sociodemographic Estimates Standard Hardware Standard Travel Forecasting Background Transportation Network Planning Package Destination Choice Trip Distribution (Gravity) Trip Attraction (Regression)

Methods Reference A-37 Budget $10,000–$20,000 Schedule 1–2 months References Cambridge Systematics, Inc. 2010. “Travel Model Validation and Reasonableness Checking Manual - Second Edition.” Prepared for the Federal Highway Administration Travel Model Improvement Program. FHWA-HEP-10-042. https://www.fhwa.dot.gov/planning/tmip/publications/other_reports/validation_and_reasonablen ess_2010/ (Accessed on Dec. 21, 2016). Meyer, M., and E.J. Miller. 2000. Urban Transportation Planning. McGraw-Hill. Trip Production (Regression) Short Description Person trip production (regression) models estimate trip productions as a function of zonal aggregate variables. Long Description Person trip production (regression) models are largely superseded by cross-classification trip production models. These regression models typically require that users aggregate zonal data from household travel surveys since there are often insufficient observations in a single zone. As a result, this method can produce misleading results from regression on aggregate values. Method Type and Performance Consistency Method type: Aggregate Performance consistency: Somewhat proven Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Congestion Management Plan Energy Use Study Environmental Clearance Age of Head of Household City County Establishment Characteristics High-Level Planning Household Size Passenger Demand

A-38 Method Selection for Travel Forecasting: User Guide and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Long-Range Transportation Plan Major Highway Corridor Study Major Transit Corridor Study Pricing Study Project Prioritization Traffic Impact Study Transit-Oriented Development Study Transportation Improvement Program Travel Demand Management Program Household Socioeconomic Characteristics Income Groups Number of Workers in Household Region Resident Passenger Travel Subarea or Corridor Travel Analysis Zone Traveler Demographics Resources Lower-class Methods Method Relationships Census Sociodemographic Estimates Standard Hardware Standard Travel Forecasting Background Transportation Network Planning Package Destination Choice Trip Distribution (Gravity) Trip Attraction (Regression) Budget $10,000–$20,000 Schedule 1–2 months

Methods Reference A-39 References Meyer, M., and E.J. Miller. 2000. Urban Transportation Planning. McGraw-Hill. Zenina, N., and A. Borisov. 2013. “Regression Analysis for Transport Trip Generation Evaluation.” Information Technology and Management Science, Vol. 16, No. 1, pp. 89–94. OTHER TRAVEL DEMAND METHODS Short Description Travel demand forecasting methods that can be applied for a single purpose to support passenger and freight and assignment methods. Methods Direct-Demand Model Origin-Destination Matrix Estimation Pivot-Point and Incremental Models Methods for Other Travel Demand Methods Direct-demand Model Short Description Direct-demand models directly estimate travel demand for a single mode using modal networks and socioeconomic data. Long Description Direct-demand models replace the conventional trip generation, distribution, and mode choice model components with an estimation of total travel demand and mode choice. Direct-demand models are closely related to general econometric models of demand. These models can be segmented by trip purposes and are often used to simulate transit ridership, although there are other applications for bicycle, pedestrian, or other modes. Direct-demand models are flexible and can provide a better fit with observed data than conventional 4-step models for transit ridership. However, these models are not as behaviorally robust as 4-step or activity-based travel models because they are sensitive to a limited number of factors, whereas 4-step or activity-based models are sensitive to a wider array of factors and interactions between these factors Method Type and Performance Consistency Method type: Not applicable Performance consistency: Proven or validated

A-40 Method Selection for Travel Forecasting: User Guide Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Bicycle and Pedestrian Capital Investments Bicycle and Pedestrian Plan Environmental Impact Study Environmental Justice Plan Health and Physical Activity Plan Major Transit Corridor Study Transit Operations Study Transit-Oriented Development Study Travel Demand Management Program Age of Head of Household City Complex Transit Networks County Daily Detailed Planning Fares Federal Standards High-Level Planning Household Size Household Socioeconomic Characteristics Income Groups Number of Workers in Household Project Design Region Resident Passenger Travel Route Simple Transit Networks Special Markets Subarea or Corridor Time Period Transit Network Supply Transit Operations Travel Analysis Zone Travel Distance Traveler Demographics Travel Purpose Travel Time Active Transportation Revenues Transit Measures User Benefits Resources Lower-class Methods Method Relationships Census Limited Travel Forecasting Background Standard Hardware Transit Counts Transit Networks Single-class Equilibrium Traffic Assignment Frequency-based Transit Assignment Schedule-based Transit Assignment

Methods Reference A-41 Transit On-Board Survey Transportation Network Planning Package Pedestrian/Bicycle Simulation Stochastic User Equilibrium Assignment Budget $20,000–$35,000 Schedule 2–4 months References de Dios Ortúzar, J. and L.G. Willumsen. 1990. Modelling Transport. Wiley. Fagnant, D. J., and K. Kockelman. 2016. A Direct-Demand Model for Bicycle Counts: The Impacts of Level of Service and Other Factors. Environment and Planning B: Planning and Design. Vol. 43, pp. 93–107. Origin-destination Matrix Estimation Short Description Origin-destination matrix estimation cost-effectively estimates a trip matrix directly from traffic counts. Long Description Origin-destination matrix estimation (ODME) cost-effectively estimates a trip matrix directly from traffic counts. ODME is a static mathematical estimation process to derive an origin- destination (O-D) matrix from traffic counts. This process can be applied for any time period (with counts from that time period) or any vehicle class (with vehicle class counts). A seed matrix of O-D travel can be used to improve the estimation process. Loaded flows are compared with counted volumes to calculate an adjustment to the matrix that, when loaded again, will improve the match between assigned flows and counts. This procedure continues iteratively until the match between the flows loaded from the estimated matrix and the counts cannot be improved further. ODME is not behavioral and does not provide insight into the underlying characteristics of the resulting trip tables. It is most often used when speed and simplicity of the application are required and aggregate results are sufficient for the analysis. The quality of the traffic counts and seed matrix are important to the quality of the ODME solution.

A-42 Method Selection for Travel Forecasting: User Guide Method Type and Performance Consistency Method type: Not applicable Performance consistency: Proven or validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Arterial Operations and Management Study Congestion Management Plan Emergency Evacuation Plan Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Freeway Operations and Management Study Freight Plan Greenhouse Gas Mitigation Study Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Long-Range Transportation Plan Major Highway Corridor Study Safety Program Traffic Impact Study Transit Operations Study Transit-Oriented Development Study Transportation Improvement Program 15-minutes or less Annual City City/County County Daily Detailed Planning General Purpose Lanes High-Level Planning Hour Internal Commercial Travel Megaregion Microzone Operational Planning Project Design Region Region Resident Passenger Travel Season/Month State Subarea or Corridor Time Period Travel Analysis Zone Travel Distance Truck Lanes Economic Development Freight Mobility Passenger Mobility

Methods Reference A-43 Resources Lower-class Methods Method Relationships Highway Networks Limited Travel Forecasting Background Observed Origin- Destination Trip Table Standard Hardware Statistics Background Traffic Counts Transportation Network Planning Package Single-class Equilibrium Traffic Assignment Dynamic Capacity- Constrained Transit Assignment Dynamic Traffic Assignment Frequency-based Transit Assignment Integrated Multiresolution Model Multiclass Equilibrium Traffic Assignment Pedestrian/Bicycle Simulation Time of Day (Fixed Factors) Stochastic User Equilibrium Assignment Traffic Microsimulation Budget $10,000–$35,000 Schedule 1–3 months References Florian, M., He, S., and S. Velan. 2010. “A Practical Method for Adjusting Temporal Origin- Destination Matrices for Dynamic Traffic Assignment.” Submitted to TRB Innovations in Travel Demand Forecasting. Morgan, D., and R. Mayberry. 2010. “Application of a Combined Travel Demand and Microsimulation Model for a Small City.” http://www.caliper.com/PDFs/Morgan_Paper.pdf (As of Dec. 9, 2016). Willumsen, L.G. 1978. “Estimation of O-D Matrix from Traffic Counts: A Review.” Working Paper 99, Institute for Transport Studies, University of Leeds.

A-44 Method Selection for Travel Forecasting: User Guide Pivot-point and Incremental Models Short Description Pivot-point and incremental models are used to estimate growth compared to a baseline condition and then add this growth to an observed trip table to produce a future trip table. Long Description Pivot-point and incremental models are used to estimate growth compared to a baseline condition and then add this growth to an observed trip table to produce a future trip table. The main advantage of this approach is that data are needed only for the attributes that change. Pivot-point and incremental methods can be applied in several ways. Select link analysis can be used to obtain a relationship between zonal activity and traffic levels. Forecasts of zonal activity from time series methods can then be directly related to traffic levels for a chosen facility. This method is particularly well suited for project-level analysis, where only a few links are being analyzed and where highly accurate forecasts are essential for each facility. The method applies to any situation where traffic will not be redistributed due to major network changes or capacity restraint. An advantage of this technique is that is requires only one base-case run of a 4-step model. The model does not need to be adjusted or rerun for future scenarios. Another pivot-point method is to run a synthetic travel model for a baseline and future scenario, subtract to obtain the growth, and then add this growth to an observed trip table or set of traffic counts. This method is often used to improve upon the accuracy of the synthetic forecasts for a subarea. Pivot-point and incremental models require an observed dataset to pivot off; this can be a trip table or loaded network. Method Type and Performance Consistency Method type: Not applicable Performance consistency: Proven or validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Arterial Operations and Management Study Bicycle and Pedestrian Capital Investments Bicycle and Pedestrian Plan Congestion Management Plan City Complex Transit Networks County Detailed Planning External Commercial Travel External Passenger Travel Fares Auto Congestion/Traffic Measures Economic Development Passenger Demand Passenger Mobility Transit Measures User Benefits

Methods Reference A-45 Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Freeway Operations and Management Study Health and Physical Activity Plan Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Major Highway Corridor Study Major Transit Corridor Study Pricing Study Project Prioritization Safety Program Traffic Impact Study Transit Operations Study Transit-Oriented Development Study Transportation Improvement Program Travel Demand Management Program Fixed Tolls General Purpose Lanes High-Level Planning Highway Network Supply Highway Operations Household Socioeconomic Characteristics Income Groups Internal Commercial Travel Long-Distance Passenger Travel Managed Lanes Microzone Operating Costs Region Resident Passenger Travel Roadway Segment or Parcel Route Simple Transit Networks Special Markets Subarea or Corridor Time Period Transit Network Supply Transit Operations Travel Analysis Zone Travel Distance Traveler Demographics Travel Purpose Truck Lanes Turn Lanes/Auxiliary Lanes Value-of-Time Segments Vehicle Occupancy Groups Resources Lower-class Methods Method Relationships Observed Origin- Destination Trip Table Standard Hardware Standard Travel Forecasting Background Schedule-based Transit Assignment Single-class Equilibrium Traffic Assignment Multiclass Equilibrium

A-46 Method Selection for Travel Forecasting: User Guide Transportation Network Planning Package Traffic Assignment Volume-Delay Functions Dynamic Capacity- Constrained Transit Assignment Dynamic Traffic Assignment Frequency-based Transit Assignment Stochastic User Equilibrium Assignment Budget $25,000–$50,000 Schedule 3–6 months References Center for Urban Transportation Studies, University of Wisconsin – Milwaukee and Wisconsin Department of Transportation. 1999. Guidebook on Statewide Travel Forecasting. https://www.fhwa.dot.gov/planning/processes/statewide/forecasting/swtravel.pdf (As of Dec. 12, 2016). de Dios Ortúzar, J. and L.G. Willumsen. 1990. Modelling Transport. Wiley. COMMERCIAL VEHICLE MODELS Short Description Commercial vehicle models include supply chain, logistics, truck touring model components for simulating the movement of goods and services. Methods Commodity Flow Models Firm Synthesis Freight and Services Tours and Stops Mode and Shipment Size Mode Choice (Fixed Factors) Stop Sequence and Duration Supplier Selection Model Supply Chain Model Time of Day (Fixed Factors)

Methods Reference A-47 Trip Attraction (Regression) Trip Distribution (Gravity) Trip Production (Cross-classification) Vehicle Types and Tour Patterns Methods for Commercial Vehicle Models Commodity Flow Models Short Description Commercial vehicle commodity flow models allocate commodity flows between zones in an aggregate freight or truck model based on employment by industry in each zone and distance between the zones. Long Description Commercial vehicle commodity flow models use a commodity flow allocation model to apportion commodity flows between zones based on measures of activity such as employment by industry in each zone and distance between the zones. This simple approach disaggregates observed commodity flow data such as the Freight Analysis Framework (FAF) database. Allocated commodity flows can then be converted to truck trips using payload factors (average loading by commodity). Method Type and Performance Consistency Method type: Aggregate Performance consistency: Proven or validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Economic Development Plan Economic Impact Analysis Freight Plan Greenhouse Gas Mitigation Study Long-Range Transportation Plan Annual City/County Commodity Group County Employment Categories Establishment Characteristics Establishment Size Groups External Commercial Travel High-Level Planning Internal Commercial Commercial Vehicle Demand Economic Development Freight Mobility

A-48 Method Selection for Travel Forecasting: User Guide Major Highway Corridor Study Project Prioritization Sustainable Community Strategies Transportation Improvement Program Travel Megaregion Region Season/Month State Travel Analysis Zone Travel Distance Resources Lower-class Methods Method Relationships Commodity Flow Data Custom Programming Employment Data Freight Background Highway Networks Standard Hardware Standard Travel Forecasting Background Transportation Network Planning Package Trip Distribution (Gravity) Mode and Shipment Size Mode Choice (Fixed Factors) Budget $20,000–$40,000 Schedule 2–6 months References Cambridge Systematics, Inc., Global Insight, Cohen, H., Horowitz, A., and R. Pendyala. 2008. NCHRP Report 606: Forecasting Statewide Freight Toolkit. Transportation Research Board. Washington, DC. Kuzmyak, J.R. 2008. NCHRP Synthesis of Highway Practice 384: Forecasting Metropolitan Commercial and Freight Travel. Transportation Research Board. Washington, DC. Firm Synthesis Short Description Commercial vehicle firm synthesis creates a database of business establishments that are producers and consumers of freight.

Methods Reference A-49 Long Description Commercial vehicle firm synthesis describes the process of creating individual firm objects to represent establishments in simulation-based freight models. This process uses observed employment data for the modeling region, which may be available in different forms. For a fully disaggregate approach, the ideal form of data is an employment database with records on individual establishments, including addresses for physical locations, number of employees, and NAICS industry and commodity codes. Employment databases with this level of detail are produced by commercial vendors, such as InfoGroup, or can be developed from publicly available datasets such as the Longitudinal Employer-Household Dynamics (LEHD) and County Business Patterns (CBP) datasets, both published by the US Census Bureau. LEHD and CBP provide summaries of establishment-based employment, aggregated by US counties and broken down by NAICS industry codes and firm-size groupings. These more aggregate datasets are transformed into synthetic firms for simulation modeling uses through the following steps: 1. Develop joint distributions of the number of establishments by NAICS codes and employee-size groupings. Start with the most disaggregate groupings of NAICS (e.g., six digits for CBP) and firm size available in the source data, and aggregate as necessary to the groupings needed for the model. County-level aggregation is typical for CBP. 2. Enumerate firms. Create a firm record/object for the simulation (enumeration) for each count of an establishment by firm size and category. This should provide both a NAICS code attribute and a firm-size attribute. Distribute the synthesized firms to traffic analysis zones (TAZ) or similar geography using local employment data if locational attributes are needed for a finer geographic resolution than the county level. Data can be used to create synthetic firms in more precise geographic locations if point business data are available. 3. Add Production. The make table in input-output accounts can be used to estimate the dollar quantities of commodities produced by synthetic firms and can be differentiated by industry and firm size. For some industries that produce multiple commodities, a simple approach can be to select a single production commodity and assume the amount produced is proportional to the firm size. 4. Add Consumption. Given a production commodity and quantity, the input-output Use or Direct Requirements table may be used to generate consumption commodities. The Direct Requirements table shows the dollar-amount of each input commodity needed to produce a dollar of the output (production) commodity. Simplifying assumptions may be necessary to limit the number of modeled input commodities because most production commodities use multiple input commodities; this may also require rescaling total quantities to adequately represent flows. The firm synthesis model produces a list of establishments with location, firm-size, industry, production, and consumption details that aggregate to meet the joint distributions of establishments by industry and firm size.

A-50 Method Selection for Travel Forecasting: User Guide Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Somewhat proven Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Economic Development Plan Economic Impact Analysis Freight Plan Greenhouse Gas Mitigation Study Long-Range Transportation Plan Major Highway Corridor Study Project Prioritization Sustainable Community Strategies Transportation Improvement Program Annual City/County Commodity Group Detailed Planning Employment Categories Establishment Characteristics Establishment Size Groups External Commercial Travel High-Level Planning Internal Commercial Travel Investment Grade Megaregion Microzone Region Roadway Segment or Parcel State Travel Analysis Zone Resources Lower-class Methods Method Relationships Advanced Travel Forecasting Background Custom Programming Employment Data Establishment Survey Sociodemographic Estimates Standard Hardware Statistics Background Supplier Selection Model

Methods Reference A-51 Budget $20,000–$40,000 Schedule 2–3 months References RSG, University of Illinois at Chicago, and J. Bowman. 2012. Tour-based and Supply Chain Freight Forecasting Framework: Final Report (BAA DTFH61-10-R-00013). Federal Highway Administration. Freight and Services Tours and Stops Short Description Commercial vehicle tours and stops models simulate the complexity of a multistop tour and group delivery/pick-up locations of shipments into truck tours. Long Description First, the commercial vehicle tours and stops model determines the number of truck tours for each shipment. Second, the model clusters stops for the shipments in multitour patterns per the number of truck tours category. This type of model can be estimated using a truck diary survey that records truck movements, including pick-up or delivery location details at each stop, over the course of a day or more. The Texas Commercial Vehicle Survey is an example; data from this survey have been used for developing tour-based truck models. The number of tours model is a discrete choice model with alternatives for the category of number of truck tours (e.g., all stops in one tour, two tours, three tours, and four tours). All-in- one tour means the shipment belongs to a tour category in which a truck covers all the stops assigned to it in a single tour. The model’s second stage simulates clusters of shipments based on the number of truck tours simulated; this simulation is made only for shipments in multitour patterns. An algorithm, such as hierarchical clustering, is used to cluster shipments based on proximity. The measures of proximity can be travel time or distance or other impedance measures based on highway network skims. Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Somewhat proven

A-52 Method Selection for Travel Forecasting: User Guide Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Economic Development Plan Economic Impact Analysis Freight Plan Greenhouse Gas Mitigation Study Long-Range Transportation Plan Major Highway Corridor Study Project Prioritization Sustainable Community Strategies Transportation Improvement Program 15-minutes or less City Commodity Group County Daily Detailed Planning External Commercial Travel Fixed Tolls High-Level Planning Highway Network Supply Hour Internal Commercial Travel Investment Grade Megaregion Microzone Region Roadway Segment or Parcel State Time Period Travel Analysis Zone Travel Distance Travel Time Truck Lanes Commercial Vehicle Demand Resources Lower-class Methods Method Relationships Advanced Travel Forecasting Background Custom Programming Employment Data Freight Background Standard Hardware Statistics Background Truck Surveys Stop Sequence and Duration Vehicle Types and Tour Patterns

Methods Reference A-53 Budget $20,000–$40,000 Schedule 2–3 months References RSG, University of Illinois at Chicago, and J. Bowman. 2012. Tour-based and Supply Chain Freight Forecasting Framework: Final Report (BAA DTFH61-10-R-00013). Federal Highway Administration. Ruan, M., Lin, J., and K. Kawamura. 2011. “Modeling Commercial Vehicle Daily Tour Chaining.” Presented at the Transportation Research Board 90th Annual Meeting, Washington, DC. Smith, C., Chen, J., Sana, B., and M. Outwater. 2013. “A Disaggregate Tour-Based Truck Model with Simulation of Shipment Allocation to Trucks.” Presented at the Transportation Research Board 92nd Annual Meeting, Washington, DC. Mode and Shipment Size Short Description Simulates primary modes and shipment sizes for shipments with complex supply chains using inputs from modal networks, including descriptions of transfer facilities. Long Description A discrete choice model and path choice model assigns a mode and shipment size for shipments transported between buyer-supplier pairs. The mode choice is between several primary modes (e.g., road, rail, air, and water) The shipment size choice is determined based on the travel time, cost, characteristics of the shipment (e.g., bulk natural resources, finished goods), characteristics of the distribution channel (e.g., whether the shipment is routed via a warehouse, consolidation, or distribution center), and whether the shipment includes an intermodal transfer (e.g. truck-rail- truck). A mode and shipment size that would have the lowest annual transport and logistics cost is chosen from a set of feasible modes. Annual goods flow between buyer-supplier firm pairs is broken down into individual shipments, revealing the shipment size (weight) and the corresponding number of shipments per year. Shipment size affects the mode used to transport the shipment. A discrete choice model is used to estimate choice of shipment size, typically estimated using an establishment survey dataset. Alternatives in the model are shipment size categories, with specific shipment sizes within categories simulated using data such as that from the Commodity Flow Survey, by commodity. An annual delivery frequency is calculated using the annual commodity flow (in tons) and the individual shipment size for all the buyer-supplier firms.

A-54 Method Selection for Travel Forecasting: User Guide Applying the model involves a two-step process: 1) a set of feasible paths between each origin- destination pair is enumerated using modal networks and nodal information about locations where transfers between modes can take place; and 2) a reasonable set of parameters is applied to generate total annual transport and logistics costs for each combination of path and mode. Supply chain and inventory control costs are considered and incorporated to account for the inventory- associated costs when calculating the total annual costs for each pair of seller and buyer. Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Somewhat proven Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Economic Development Plan Economic Impact Analysis Freight Plan Greenhouse Gas Mitigation Study Long-Range Transportation Plan Major Highway Corridor Study Project Prioritization Sustainable Community Strategies Transportation Improvement Program Annual City/County Commodity Group Daily Detailed Planning Employment Categories Establishment Characteristics Establishment Size Groups External Commercial Travel Fixed Tolls High-Level Planning Highway Network Supply Internal Commercial Travel Investment Grade Megaregion Microzone Operating Costs Region Roadway Segment or Parcel Season/Month State Travel Analysis Zone Travel Distance Travel Time Commercial Vehicle Demand Freight Mobility Freight Mode Shares

Methods Reference A-55 Truck Lanes Resources Lower-class Methods Method Relationships Advanced Travel Forecasting Background Custom Programming Employment Data Establishment Survey Freight Background Freight Rail/Intermodal Network Highway Networks Standard Hardware Statistics Background Truck Surveys Mode Choice (Fixed Factors) Supply Chain Model Vehicle Types and Tour Patterns Commodity Flow Models Budget $50,000–$100,000 Schedule 4–10 months References Cambridge Systematics, Inc. 2011. A Working Demonstration of a Mesoscale Freight Model for the Chicago Region Final Report and User’s Guide. Prepared for the Chicago Metropolitan Agency for Planning. de Jong, G., and M. Ben-Akiva. 2007. A Micro-Simulation Model of Shipment Size And Transport Chain Choice. Transportation Research Part B: Methodological, Vol. 41, No. 9, pp.950–965. Leachman, R., Prince, T., Brown, T., and G. Fetty. 2005. Final Report: Port and Modal Elasticity Study. Southern California Association of Governments. RSG, University of Illinois at Chicago, and J. Bowman. 2012. Tour-based and Supply Chain Freight Forecasting Framework: Final Report (BAA DTFH61-10-R-00013). Federal Highway Administration. Mode Choice (Fixed Factors) Short Description Commercial vehicle mode choice allocates commodity flows to modes using fixed factors derived from observed data.

A-56 Method Selection for Travel Forecasting: User Guide Long Description An aggregate commercial vehicle model uses a fixed factor mode choice model that fixes mode shares over time. Commodity flows are allocated to modes by commodity group and zone pair based on observed commodity flow data from sources such as the Freight Analysis Framework (FAF) or other commodity flow data sources. The modes may include primary modes (e.g., truck or rail), or multiple mode combinations (such as truck-rail-truck); detail within modes (e.g., truck type) may also be provided. Method Type and Performance Consistency Method type: Aggregate Performance consistency: Proven or validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Economic Development Plan Economic Impact Analysis Freight Plan Greenhouse Gas Mitigation Study Long-Range Transportation Plan Major Highway Corridor Study Project Prioritization Sustainable Community Strategies Transportation Improvement Program Annual Daily Detailed Planning Employment Categories External Commercial Travel Fixed Tolls High-Level Planning Internal Commercial Travel Megaregion Region State Travel Analysis Zone Commercial Vehicle Demand Freight Mobility Freight Mode Shares Resources Lower-class Methods Method Relationships Commodity Flow Data Freight Background Standard Hardware Standard Travel Forecasting Background Time of Day (Fixed Factors) Commodity Flow Models Trip Distribution (Gravity) Stochastic User

Methods Reference A-57 Transportation Network Planning Package Equilibrium Assignment Single-class Equilibrium Traffic Assignment Multiclass Equilibrium Traffic Assignment Budget $10,000–$20,000 Schedule 1–2 months References Cambridge Systematics, Inc., Global Insight, Cohen H., Horowitz A., and R. Pendyala. 2008. NCHRP Report 606: Forecasting Statewide Freight Toolkit. Transportation Research Board, Washington, DC. Kuzmyak, J.R. 2008. NCHRP Synthesis of Highway Practice 384: Forecasting Metropolitan Commercial and Freight Travel. Transportation Research Board. Washington, DC. Stop Sequence and Duration Short Description A commercial vehicle stop sequence and duration model sequences the stops in a commercial vehicle’s tour and estimates the time spent at each stop. Long Description A greedy algorithm—or “traveling salesman” algorithm—can be applied to sequence all stops within a tour. A greedy algorithm makes the closest stop to the tour’s starting point the first stop on the tour. A greedy algorithm then adds the next closest stop from the previous stop to the trip sequence until all stops in a tour are assigned a sequence. The greedy algorithm estimates the shortest path based on a search of alternative sequences. Typically, this search is simplified (nonexhaustive) because this is computationally tractable and better matches suboptimal sequencing observed in truck diary data. Commercial vehicle stop sequence and duration models use discrete choice models to choose from alternative time bins for stop duration (e.g., 15 minutes or less, 15–30 minutes, 30–45 minutes, 45–60 minutes, 60–75 minutes, and more than 75 minutes). Stop duration models include variables such as shipment weight (if the shipment is a larger load it is likely to increase the duration of the stop), number of stops on the tour (if there are more stops involved in the tour, the stops are likely to be shorter in duration), and vehicle size (if the vehicle is larger, it will likely take longer to unload).

A-58 Method Selection for Travel Forecasting: User Guide Commercial vehicle stop sequence and duration models can be estimated using a truck diary survey that records truck movements. These diaries can occur over the course of one day or more and can include each stop’s shipment details, including type of pick-up and delivery items. The Texas Commercial Vehicle Survey is an example of this type of survey; data from this survey have been used for developing tour-based truck models. Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Somewhat proven Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Economic Development Plan Economic Impact Analysis Freight Plan Greenhouse Gas Mitigation Study Long-Range Transportation Plan Major Highway Corridor Study Project Prioritization Sustainable Community Strategies Transportation Improvement Program 15-minutes or less City Commodity Group County Daily Detailed Planning External Commercial Travel Fixed Tolls High-Level Planning Highway Network Supply Hour Internal Commercial Travel Investment Grade Megaregion Microzone Region Roadway Segment or Parcel State Time Period Travel Analysis Zone Travel Distance Truck Lanes Commercial Vehicle Demand

Methods Reference A-59 Resources Lower-class Methods Method Relationships Advanced Travel Forecasting Background Custom Programming Employment Data Establishment Survey Freight Background Highway Networks Standard Hardware Statistics Background Truck Surveys Freight and Services Tours and Stops Budget $20,000–$40,000 Schedule 2–3 months References RSG, University of Illinois at Chicago, and J. Bowman. 2012. Tour-based and Supply Chain Freight Forecasting Framework: Final Report (BAA DTFH61-10-R-00013). Federal Highway Administration. Ruan, M., Lin, J., and K. Kawamura. 2011. “Modeling Commercial Vehicle Daily Tour Chaining.” Presented at the Transportation Research Board 90th Annual Meeting, Washington, DC. Smith, C., Chen, J., Sana, B., and M. Outwater. 2013. “A Disaggregate Tour-Based Truck Model with Simulation of Shipment Allocation to Trucks.” Presented at the Transportation Research Board 92nd Annual Meeting, Washington, DC. Supplier Selection Model Short Description Commercial vehicle supplier selection models match buyers and suppliers among synthesized firms based on the size, industry, proximity, and production and consumption needs of each firm. Long Description A supplier is selected from the buyers and suppliers dataset produced by a firm synthesis model for each buyer or supplier firm. The probability of a supplier firm being chosen increases with its size (employment) and its proximity to the buyer firm. The goods demand model then allocates commodity flows to the buyer-supplier pairs.

A-60 Method Selection for Travel Forecasting: User Guide The supplier selection model creates a choice set of suppliers for each buyer firm based on the commodity inputs the buyer firm requires and the corresponding NAICS of the suppliers. A supplier firm can be excluded from the choice set if no flows for the commodity being traded are observed in the Freight Analysis Framework (FAF) commodity flow database or alternative commodity flow database between the relevant zones (if a commodity flow database is being used as a control). A score for each buyer and potential supplier pair is then calculated by using the model’s coefficients and by adding a random value for stochasticity. For each buyer firm, the supplier firm with the highest score is selected. The goods demand model uses a commodity flow database, such as FAF, and the buyer-supplier pairs estimated in the supplier firm selection model as inputs. The volume of commodities shipped on an annual basis between each pair of firms is apportioned based on the number of employees at the buyer firm and the buyer firm’s industry; this is done to match the total commodity flows by zone pair and commodity in the commodity flow database. Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Somewhat proven Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Economic Development Plan Economic Impact Analysis Freight Plan Greenhouse Gas Mitigation Study Long-Range Transportation Plan Major Highway Corridor Study Project Prioritization Sustainable Community Strategies Transportation Improvement Program Annual City/County Commodity Group Detailed Planning Employment Categories Establishment Characteristics Establishment Size Groups External Commercial Travel High-Level Planning Internal Commercial Travel Investment Grade Megaregion Microzone Region Roadway Segment or Parcel State Economic Development Freight Mobility User Benefits

Methods Reference A-61 Travel Analysis Zone Travel Distance Resources Lower-class Methods Method Relationships Advanced Travel Forecasting Background Commodity Flow Data Custom Programming Employment Data Establishment Survey Freight Background Standard Hardware Statistics Background Firm Synthesis Supply Chain Model Budget $20,000–$40,000 Schedule 2–3 months References Cambridge Systematics, Inc. 2011. A Working Demonstration of a Mesoscale Freight Model for the Chicago Region Final Report and User’s Guide. Prepared for the Chicago Metropolitan Agency for Planning. RSG, University of Illinois at Chicago, and J. Bowman. 2012. Tour-based and Supply Chain Freight Forecasting Framework: Final Report (BAA DTFH61-10-R-00013). Federal Highway Administration. Samimi, A., Mohammadian, A., and K. Kawamura. 2010. A Behavioral Freight Movement Microsimulation Model: Method and Data. Transportation Letters: The International Journal of Transportation Research, Vol. 2, pp. 53–62. Supply Chain Model Short Description Commercial vehicle supply chain models allocate commodity flows to a distribution channel for shipments between buyer-supplier pairs simulated in supplier selection models. Long Description Supply chains are represented by distribution channels, which refers to the logistics that a shipment follows from the supplier to the buyer (e.g., whether it is shipped directly or via one or more intermediary locations such as distribution centers, intermodal transshipment locations, or

A-62 Method Selection for Travel Forecasting: User Guide other transfer or temporary storage locations). The distribution channel will affect the mode, cost, shipment size, or frequency of shipments between a buyer-supplier firm pair. A distribution channel model uses discrete choice methods to allocate the distribution channel used for each buyer-supplier pair. The choice set in the model includes alternatives, such as the direct distribution channel and distribution channels, with one or more stops. The variables that affect the choice of distribution channel include firm size, commodity, and the industry type of the firms involved. Distribution channel models are estimated using disaggregate shipment data sources, such as the Commodity Flow Survey data. Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Somewhat proven Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Economic Development Plan Economic Impact Analysis Freight Plan Greenhouse Gas Mitigation Study Long-Range Transportation Plan Major Highway Corridor Study Project Prioritization Sustainable Community Strategies Transportation Improvement Program Annual Commodity Group Detailed Planning Employment Categories Establishment Characteristics Establishment Size Groups External Commercial Travel Fixed Tolls High-Level Planning Highway Network Supply Internal Commercial Travel Investment Grade Megaregion Microzone Region Roadway Segment or Parcel Season/Month State Travel Analysis Zone Travel Distance Travel Time Freight Mobility

Methods Reference A-63 Resources Lower-class Methods Method Relationships Advanced Travel Forecasting Background Commodity Flow Data Custom Programming Establishment Survey Freight Background Standard Hardware Statistics Background Truck Surveys Mode and Shipment Size Supplier Selection Model Budget $20,000–$40,000 Schedule 2–3 months References Cavalcante, R., and M.J. Roorda. 2010. “A Disaggregate Urban Shipment Size/Vehicle-Type Choice Model.” Presented at the Transportation Research Board 89th Annual Meeting, Washington, DC. Pourabdollahi, Z., Karimi, B., Mohammadian, A., and K. Kawamura. 2014. “Shipping Chain Choices in Long-Distance Supply Chains: Descriptive Analysis and a Decision Tree Model.” Submitted to the Transportation Research Board 93rd Annual Meeting, Washington, DC. RSG, University of Illinois at Chicago, and J. Bowman. 2012. Tour-based and Supply Chain Freight Forecasting Framework: Final Report (BAA DTFH61-10-R-00013). Federal Highway Administration. Samimi, A., Mohammadian, A., and K. Kawamura. 2010. A Behavioral Freight Movement Microsimulation Model: Method and Data. Transportation Letters: The International Journal of Transportation Research, Vol. 2, pp. 53–62. Time of Day (Fixed Factors) Short Description Commercial vehicle time of day (fixed factors) models allocate daily commercial vehicle trips to time periods or hours using fixed factors derived from observed data. Long Description Allocation of the commercial vehicle trips by time of day occurs after commodity flows have been converted to daily truck trips, or a daily truck trip table has been directly developed from

A-64 Method Selection for Travel Forecasting: User Guide truck trip generation rates. The fixed factors can also be estimated from truck counts (by truck class if separate factors are required for different types of truck) or from truck diary survey or establishment surveys that include information on truck trip start and end times. Method Type and Performance Consistency Method type: Aggregate Performance consistency: Proven or validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Economic Development Plan Economic Impact Analysis Freight Plan Greenhouse Gas Mitigation Study Long-Range Transportation Plan Major Highway Corridor Study Project Prioritization Sustainable Community Strategies Transportation Improvement Program 15-minutes or less City County Detailed Planning External Commercial Travel High-Level Planning Hour Internal Commercial Travel Megaregion Region State Time Period Travel Analysis Zone Commercial Vehicle Demand Resources Lower-class Methods Method Relationships Establishment Survey Standard Hardware Standard Travel Forecasting Background Transportation Network Planning Package Truck Counts Multiclass Equilibrium Traffic Assignment Stochastic User Equilibrium Assignment Single-class Equilibrium Traffic Assignment Mode Choice (Fixed Factors) Trip Distribution (Gravity)

Methods Reference A-65 Budget $10,000–$20,000 Schedule 1–2 months References Cambridge Systematics, Inc., Global Insight, Cohen H., Horowitz A., and R. Pendyala. 2008. NCHRP Report 606: Forecasting Statewide Freight Toolkit. Transportation Research Board, Washington, DC. Kuzmyak, J. R. 2008. NCHRP Synthesis of Highway Practice 384: Forecasting Metropolitan Commercial and Freight Travel. Transportation Research Board. Washington, DC. Trip Attraction (Regression) Short Description Commercial vehicle trip attraction models estimate commercial vehicle attractions as a function of activity in a zone. Long Description Commercial vehicle trip attraction models estimate commercial vehicle attractions as a function of activity in a zone. These models estimate trip attractions as a function of zone activity, usually employment by industrial sector (e.g., agricultural, manufacturing, retail). The data used to estimate these regression models is typically an establishment survey that collects data on the number of deliveries to an establishment and includes relevant information about the deliveries (e.g., size, weight, commodity), the type of truck used, and information about the establishment (e.g., industry, size, number of employees). Method Type and Performance Consistency Method type: Aggregate Performance consistency: Proven or validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Economic Development Annual Commodity Group County Daily Detailed Planning Commercial Vehicle Demand

A-66 Method Selection for Travel Forecasting: User Guide Plan Economic Impact Analysis Freight Plan Greenhouse Gas Mitigation Study Long-Range Transportation Plan Major Highway Corridor Study Project Prioritization Sustainable Community Strategies Transportation Improvement Program Employment Categories Establishment Characteristics Establishment Size Groups External Commercial Travel High-Level Planning Highway Network Supply Internal Commercial Travel Megaregion Region Season/Month State Time Period Travel Analysis Zone Resources Lower-class Methods Method Relationships Census Employment Data Standard Hardware Standard Travel Forecasting Background Statistics Background Transportation Network Planning Package Trip Production (Cross- classification) Trip Distribution (Gravity) Budget $20,000–$30,000 Schedule 2–3 months References Cambridge Systematics, Inc., Global Insight, Cohen H., Horowitz A., and R. Pendyala. 2008. NCHRP Report 606: Forecasting Statewide Freight Toolkit. Transportation Research Board, Washington, DC. Holguin-Veras, J., Jaller, M., Sanchez-Diaz, I., Wojtowicz, J., Campbell, S., Levinson, H., Lawson, C., Powers, E. L., and L. Tavasszy. 2012. NCHRP Report 739/NCFRP Report 19: Freight Trip Generation and Land Use. Transportation Research Board. Washington, DC.

Methods Reference A-67 Trip Distribution (Gravity) Short Description Commercial vehicle trip distribution models distribute trips between production zones and attraction zones using a gravity model. Long Description Commercial vehicle trip distribution models distribute trips between production zones and attraction zones using a gravity model. These models allocate commodity flows or truck trips based on the number of productions multiplied by the number of attractions multiplied by a “friction factor” that represents travel times between zones. Estimation procedures consist of estimating friction factors as a function of travel time or other measures of impedance that include costs. Travel times are usually based on vehicle (truck) travel times, but these may include a composite travel time if multiple freight transportation modes are being considered. The gravity model can be calibrated using observed data from establishment or truck surveys that collects truck trip length or shipment length information. Method Type and Performance Consistency Method type: Aggregate Performance consistency: Proven or validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Economic Development Plan Economic Impact Analysis Freight Plan Greenhouse Gas Mitigation Study Long-Range Transportation Plan Major Highway Corridor Commodity Group County Daily Detailed Planning Employment Categories Establishment Characteristics Establishment Size Groups External Commercial Travel High-Level Planning Highway Network Supply Internal Commercial Commercial Vehicle Demand Economic Development User Benefits Kuzmyak, J. R. 2008. NCHRP Synthesis of Highway Practice 384: Forecasting Metropolitan Commercial and Freight Travel. Transportation Research Board. Washington, DC.

A-68 Method Selection for Travel Forecasting: User Guide Study Project Prioritization Sustainable Community Strategies Transportation Improvement Program Travel Megaregion Region State Time Period Travel Analysis Zone Travel Distance Travel Time Resources Lower-class Methods Method Relationships Highway Networks Standard Hardware Standard Travel Forecasting Background Statistics Background Transportation Network Planning Package Trip Production (Cross- classification) Time of Day (Fixed Factors) Trip Attraction (Regression) Mode Choice (Fixed Factors) Budget $20,000–$30,000 Schedule 2–3 months References Cambridge Systematics, Inc., Global Insight, Cohen H., Horowitz A., and R. Pendyala. 2008. NCHRP Report 606: Forecasting Statewide Freight Toolkit. Transportation Research Board. Kuzmyak, J. R. 2008. NCHRP Synthesis of Highway Practice 384: Forecasting Metropolitan Commercial and Freight Travel. Transportation Research Board. Washington, DC. Trip Production (Cross-classification) Short Description Commercial vehicle trip production (cross-class) models estimate the number of truck trips by vehicle type (light, medium, heavy) and commodity group (8–12 types) as a function of employment types. Long Description Commercial vehicle trip production (cross-class) models estimate the number of truck trips by vehicle type (light, medium, heavy) and commodity group (8–12 types) as a function of different

Methods Reference A-69 types of employment. These models include the dimensions of the cross-classification industrial categories into which employment locations in the zone can be categorized, vehicle (truck) types, and commodity group (usually using aggregations of one of the common commodity classifications such as the Standard Classification of Transported Goods [SCTG]). Truck trip rates by vehicle type and industrial category are typically derived from an establishment survey, and can be based on measures such as the employment of the establishment, its building square footage, or other measures (such as the number of truck bays). Method Type and Performance Consistency Method type: Aggregate Performance consistency: Proven or validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Economic Development Plan Economic Impact Analysis Freight Plan Greenhouse Gas Mitigation Study Long-Range Transportation Plan Major Highway Corridor Study Project Prioritization Sustainable Community Strategies Transportation Improvement Program Annual City Commodity Group County Daily Detailed Planning Employment Categories Establishment Characteristics Establishment Size Groups General Purpose Lanes High-Level Planning Internal Commercial Travel Megaregion Operating Costs Region State State Standards Subarea or Corridor Travel Analysis Zone Truck Lanes Commercial Vehicle Demand

A-70 Method Selection for Travel Forecasting: User Guide Resources Lower-class Methods Method Relationships Employment Data Standard Hardware Standard Travel Forecasting Background Statistics Background Transportation Network Planning Package Trip Attraction (Regression) Trip Distribution (Gravity) Budget $20,000–$30,000 Schedule 2–3 months References Cambridge Systematics, Inc., Global Insight, Cohen H., Horowitz A., and R. Pendyala. 2008. NCHRP Report 606: Forecasting Statewide Freight Toolkit. Transportation Research Board, Washington, DC. Kuzmyak, J. R. 2008. NCHRP Synthesis of Highway Practice 384: Forecasting Metropolitan Commercial and Freight Travel. Transportation Research Board. Washington, DC. Holguin-Veras, J., Jaller, M., Sanchez-Diaz, I., Wojtowicz, J., Campbell, S., Levinson, H., Lawson, C., Powers, E. L., and L. Tavasszy. 2012. NCHRP Report 739/NCFRP Report 19: Freight Trip Generation and Land Use. Transportation Research Board. Washington, DC. Vehicle Types and Tour Patterns Short Description Commercial vehicle types and tour-pattern models simulate the choices of vehicle and tour patterns for the delivery or pick-up of each shipment in a disaggregate truck touring model. Long Description The vehicle types and tour-pattern models jointly simulate the type of truck and the tour pattern for each of the shipments due to the interrelationship that exists between the choices of vehicle or truck types and tour patterns. These models can be estimated using a truck diary survey that records the daily (or more) movements of trucks and includes pick-up or delivery details of shipments at each stop. The Texas Commercial Vehicle Survey is an example; data from this survey have been used for developing tour-based truck models. The choices in the vehicle types and tour-pattern models are combinations of tour patterns (e.g., direct and multistop) and vehicle types (e.g., 2-axle single-unit trucks, 3- or 4-axle single-unit

Methods Reference A-71 trucks, and semi/trailer). Explanatory variables include the type of commodity, pick-up/drop-off weights (shipment sizes), and the type of industry at the stop location of the shipment. Typically, trucks get larger with heavier shipments and heavier shipments are more likely to be delivered on direct tours as opposed to multistop tours. Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Somewhat proven Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Comprehensive Plans Economic Development Plan Economic Impact Analysis Freight Plan Greenhouse Gas Mitigation Study Long-Range Transportation Plan Major Highway Corridor Study Project Prioritization Sustainable Community Strategies Transportation Improvement Program 15-minutes or less City Commodity Group County Daily Detailed Planning Employment Categories Establishment Characteristics Establishment Size Groups External Commercial Travel Fixed Tolls High-Level Planning Highway Network Supply Hour Internal Commercial Travel Investment Grade Megaregion Microzone Operating Costs Region Roadway Segment or Parcel State Time Period Travel Analysis Zone Travel Distance Truck Lanes Commercial Vehicle Demand

A-72 Method Selection for Travel Forecasting: User Guide Resources Lower-class Methods Method Relationships Advanced Travel Forecasting Background Custom Programming Employment Data Establishment Survey Freight Background Standard Hardware Statistics Background Truck Surveys Freight and Services Tours and Stops Mode and Shipment Size Budget $20,000–$40,000 Schedule 2–3 months References RSG, University of Illinois at Chicago, and J. Bowman. 2012. Tour-based and Supply Chain Freight Forecasting Framework: Final Report (BAA DTFH61-10-R-00013). Federal Highway Administration. Ruan, M., Lin, J., and K. Kawamura. 2011. “Modeling Commercial Vehicle Daily Tour Chaining.” Presented at the Transportation Research Board 90th Annual Meeting, Washington, DC. Smith, C., Chen, J., Sana, B., and M. Outwater. 2013. “A Disaggregate Tour-Based Truck Model with Simulation of Shipment Allocation to Trucks.” Presented at the Transportation Research Board 92nd Annual Meeting, Washington, DC. ASSIGNMENT AND MICROSIMULATION MODELS Short Description Route, path models of traffic, transit, and bike/pedestrian microsimulation. Methods Dynamic Capacity-Constrained Transit Assignment Dynamic Traffic Assignment Frequency-based Transit Assignment Integrated Multiresolution Model Intersection Delay Multiclass Equilibrium Traffic Assignment

Methods Reference A-73 Pedestrian/Bicycle Simulation Route-Path Choice Models Schedule-based Transit Assignment Single-class Equilibrium Traffic Assignment Stochastic User Equilibrium Assignment Traffic Microsimulation Volume-Delay Functions Methods for Assignment and Microsimulation Models Dynamic Capacity-constrained Transit Assignment Short Description Dynamic capacity-constrained transit assignment adjusts the path chosen for each vehicle dynamically based on the capacity limitations of the transit network. Long Description Dynamic capacity-constrained transit assignment can employ several methods for assigning transit origin-destination trip tables to transit routes. The assignment methods differ in how they find the shortest path and in how they allocate passengers across multiple routes serving the same links. Frequency-based transit assignment methods consider only the relative average frequency of each transit route serving the same path. Schedule-based methods consider the exact timetable of transit services. Frequency-based transit assignment methods work best for moderate to high frequency transit services with regular passenger arrivals at the stop. Schedule-based methods often perform better for infrequent transit service and when considering the impacts of surges of passengers within a single hour (e.g., bus stops outside of a train station). Most frequency- and schedule-based methods do not consider capacity limitations on the transit service. Dynamic capacity-constrained transit assignment considers the service schedules, the punctuality of the service, and crowding on the transit vehicle. Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Not proven/not validated

A-74 Method Selection for Travel Forecasting: User Guide Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Economic Impact Analysis Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Major Transit Corridor Study Safety Program Transit Operations Study Transit-Oriented Development Study 15-minutes or less City City/County Complex Transit Networks County Daily Detailed Planning Fares Federal Standards Hour Microzone Operational Planning Project Design Region Region Roadway Segment or Parcel Route Subarea or Corridor Subregion Time Period Transit Network Supply Transit Operations Travel Analysis Zone Travel Distance Travel Time Accessibility Economic Development Revenues Schools Transit Measures User Benefits Resources Lower-class Methods Method Relationships Advanced Travel Forecasting Background High-performance Hardware Highway Networks Transit Counts Transit Networks Transportation Network Planning Package Frequency-based Transit Assignment Schedule-based Transit Assignment Origin-Destination Matrix Estimation Mode Choice (Disaggregate) Mode Choice (Fixed Factors) Time of Day (Fixed Factors) Time of Day (Disaggregate) Pivot-Point and

Methods Reference A-75 Incremental Models Integrated Multiresolution Model Budget $50,000–$100,000 Schedule 3–6 months References Fu, Q., Liu, R., and S. Hess. 2012. “A Review on Transit Assignment Modelling Approaches to Congested Networks: A New Perspective.” Procedia—Social and Behavioral Sciences, Vol. 54, No. 4, pp. 1145–1155. Gentile, G., and K. Noekel (Eds.). 2016. Modeling Public Transport Passenger Flows in the Era of Intelligent Transport Systems. Springer. Papola, N., Francesco, F., Gentile, G., and L. Meschini. 2008. “Schedule-Based Transit Assignment: New Dynamic Equilibrium Model with Vehicle Capacity Constraints. ” Schedule- Based Modeling of Transportation Networks: Operations Research/Computer Science Interfaces Series, Vol. 46, pp. 1–26. Schmöcker, J-D. 2006. “Dynamic Capacity-Constrained Transit Assignment.” PhD dissertation, Centre for Transport Studies, Department for Civil and Environmental Engineering, Imperial College London. Dynamic Traffic Assignment Short Description Dynamic traffic assignment reflects the impact of time-varying network delays on departure time and path choice. Long Description Metropolitan planning organizations (MPOs) and other transportation agencies in the United States use dynamic traffic assignment (DTA) models to assess the impact of potential transportation projects. These time-variant traffic flow models provide valuable insights on the macroscopic or mesoscopic impacts of a change to the system through analysis of queuing, route assignment, and travel-time reliability on a link (or network) level. FHWA’s Traffic Analysis Toolbox Volume XIV: Guidebook on the Utilization of Dynamic Traffic Assignment in Modeling provides direction on the appropriate applications of DTA tools for transportation decision-making. DTA models allow practitioners to test various network structures and schemes, including alternative capacity, traffic control (e.g., signals or ramp

A-76 Method Selection for Travel Forecasting: User Guide meters), pricing, and evacuation planning since these models capture the interactions between travelers and the network. Some of the recommended DTA applications include the following: Bottleneck removal studies. Active Transportation and Demand Management (ATDM) strategies. Integrated Corridor Management (ICM) strategies. Operational strategies. Incident response management scenarios. Special events. Work-zone impacts and construction diversion. Practitioners must be aware of the following requirements for the proper utilization of DTA before project application: Origin-destination (O-D) data—sorted into peak periods/hours—is a fundamental input to any DTA; without this information, developing an accurate DTA model is difficult. Extensive temporal data collection efforts are required for the compulsory calibration of the model. Additional requirements include access to software with DTA capabilities and staff members with expertise in transportation modeling. DTA requires extensive data collection, model calibration, and expertise. These models can be useful for MPOs and state DOTs conducting operational planning projects. DTA models are especially useful when these planning projects involve changes to roadway configuration, freeway expansions, development of a city bypass, the addition of high- occupancy toll (HOT)/high-occupancy vehicle (HOV) lanes, integrated corridor improvements, and travel demand management strategies (e.g., congestion pricing). DTA models are also recommended for engineers working on large-scale, real-time traffic management or information provision issues because of their ability to address these issues in a systematic manner integrated into the demand forecasting process. Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Somewhat proven

Methods Reference A-77 Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Arterial Operations and Management Study Comprehensive Plans Congestion Management Plan Energy Use Study Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Freeway Operations and Management Study Greenhouse Gas Mitigation Study Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Long-Range Transportation Plan Major Highway Corridor Study Major Transit Corridor Study Pricing Study Project Prioritization Safety Program Transit-Oriented Development Study Transportation Improvement Program Travel Demand Management Program 15-minutes or less City City/County County Daily Detailed Planning Dynamic Tolls Fixed Tolls General Purpose Lanes Highway Network Supply Highway Operations Hour Managed Lanes Microzone Operational Planning Project Design Region Region Road Geometry and Controls Roadway Segment or Parcel Route Subarea or Corridor Subregion Time Period Travel Analysis Zone Travel Demand Management Policies Travel Distance Travel Time Truck Lanes Turn Lanes/Auxiliary Lanes Value-of-Time Segments Accessibility Air Quality Auto Congestion/Traffic Measures Economic Development Energy Consumption Noise Pollution Revenues Safety Schools Truck Congestion/Traffic Measures User Benefits

A-78 Method Selection for Travel Forecasting: User Guide Resources Lower-class Methods Method Relationships Dynamic Traffic Assignment Dynamic Traffic Assignment Package High-performance Hardware Highway Networks Intersection Configurations and Signal Timings Standard Travel Forecasting Background Traffic Counts Traffic Speeds Transportation Network Planning Package Intersection Delay Multiclass Equilibrium Traffic Assignment Single-class Equilibrium Traffic Assignment Stochastic User Equilibrium Assignment Volume-Delay Functions Parking Location Choice Origin-Destination Matrix Estimation Mode Choice (Disaggregate) Destination Choice Time of Day (Disaggregate) Pivot-Point and Incremental Models Intersection Delay Route-Path Choice Models Integrated Multiresolution Model Budget $150,000–$200,000 Schedule 6–12 months References Chiu, Y-C., Bottom, J., Mahut, M., Paz, A., Balakrishna, R., Waller, T., and J. Hicks. 2011. Transportation Research Circular E-C153: Dynamic Traffic Assignment: A Primer. http://onlinepubs.trb.org/onlinepubs/circulars/ec153.pdf (As of Dec. 9, 2016). Sloboden, J., Lewis, J., Alexiadis, V., Chiu, Y., and E. Nava. 2012. “Guidebook on the Utilization of Dynamic Traffic Assignment in Modeling.” Traffic Analysis Toolbox, Vol. XIV, http://ops.fhwa.dot.gov/publications/fhwahop13015/fhwahop13015.pdf (As of Dec. 9, 2016). Frequency-based Transit Assignment Short Description Frequency-based transit assignment splits demand based on relative frequency among transit routes serving the same origin-destination pair. Long Description When there are multiple transit routes available to the transit passenger to get from point A to point B (all the same mode and the same travel time), then the frequency-based transit assignment method assigns the boarding passengers to each route in proportion to its frequency

Methods Reference A-79 of service. If one transit route (Route A) arrives twice an hour, and the other route (Route B) arrives once an hour, then (and if there are no capacity constraints) this assignment method will put twice as many riders on Route A as on Route B. Frequency-based transit assignments are affected by assumptions for service regularity, passenger demographics and boarding, and alighting locations. Recent research has developed and elaborated on capacity-constrained, frequency-based transit assignment. Method Type and Performance Consistency Method type: Aggregate Performance consistency: Proven or validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Economic Impact Analysis Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Major Transit Corridor Study Transit Operations Study Transit-Oriented Development Study Travel Demand Management Program City City/County Complex Transit Networks County Daily Detailed Planning Fares Federal Standards Hour Megaregion Microzone Operational Planning Region Region Roadway Segment or Parcel Route Simple Transit Networks State Subarea or Corridor Subregion Time Period Transit Network Supply Transit Operations Travel Analysis Zone Accessibility Economic Development Revenues Schools Transit Measures User Benefits

A-80 Method Selection for Travel Forecasting: User Guide Resources Lower-class Methods Method Relationships Highway Networks Standard Hardware Standard Travel Forecasting Background Transit Counts Transit Networks Transportation Network Planning Package Origin-Destination Matrix Estimation Mode Choice (Disaggregate) Mode Choice (Fixed Factors) Time of Day (Fixed Factors) Time of Day (Disaggregate) Direct-Demand Model Pivot-Point and Incremental Models Budget $25,000–$30,000 Schedule 0–3 months References Cepeda, M., Cominetti, R., and M. Florian. 2006. “A Frequency-Based Assignment Model for Congested Transit Networks with Strict Capacity Constraints: Characterization and Computation of Equilibria.” Transportation Research Part B: Methodological, Vol. 40, No. 6, pp. 437–459. Fu, Q., Liu, R., and S. Hess. 2012. “A Review on Transit Assignment Modelling Approaches to Congested Networks: A New Perspective.” Procedia–Social and Behavioral Sciences, Vol. 54, No. 4, pp. 1145–1155. Nökel, K., and S. Wekeck. 2009. “Boarding and Alighting in Frequency-Based Transit Assignment.” Transportation Research Record: Journal of the Transportation Research Board, No. 2111, pp. 60-67. Schmöcker, J-D., Bell, M.G.H., and F. Kurauchi. 2008. “A Quasi-Dynamic Capacity- Constrained Frequency-Based Transit Assignment Model.” Transportation Research Part B: Methodological, Vol. 42, No. 11, pp. 925–945. Schmöcker, J-D., Fonsone, A., Shimamota, H., Kurauchi, F., and M.G.H. Bell. 2011. “Frequency-Based Transit Assignment Considering Seat Capacities.” Transportation Research Part B: Methodological, Vol. 45, No. 2, pp. 392–408.

Methods Reference A-81 Integrated Multiresolution Model Short Description Integrated multiresolution models are “supermodels” linking macroscopic demand models and mesoscopic or microscopic traffic simulation models. Long Description Integrated multiresolution models are “supermodels” linking macroscopic demand models and mesoscopic or microscopic traffic simulation models. Each domain scale (i.e., macro-, meso-, and micro-) of demand modeling and traffic operations simulation has level-of-resolution and spatial-scope limitations. Macroscopic models can analyze large networks and estimate mode shift, but they cannot provide detailed information about vehicles or interactions between vehicles. Microscopic models can estimate individual vehicle movements, but only on relatively small networks. Mesoscopic simulations attempt to strike a balance between the two by modeling route choices of individual drivers with a limited level of detail pertaining to driver behavior. To incorporate the benefits of each mono-resolution simulation type, many practitioners have utilized an integrated approach to transportation modeling. This integrated (or multiresolution) modeling utilizes each modeling domain in the following manner: Macroscopic. Trip table manipulation for the discernment of overall trip patterns. Mesoscopic. Analysis of the impact of driver behavior in reaction to mitigation strategies. Microscopic. Analysis of the impact of traffic control strategies at roadway junctions. Some integrated multiresolution models may only include two of the three possible levels. FHWA’s report, The Effective Integration of Analysis, Modeling and Simulation, evaluates each domain’s practicality and applicability and establishes means by which several domains are correlated. For example, many MPOs and transit agencies currently utilize macroscopic travel demand models for transit network improvements; incorporating DTA or microsimulation into an integrated modeling approach may better serve these agencies’ objectives. However, the tools to connect these areas may still be in development or only available through proprietary software. Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Not proven/not validated

A-82 Method Selection for Travel Forecasting: User Guide Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Arterial Operations and Management Study Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Freeway Operations and Management Study Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Major Highway Corridor Study 15-minutes or less Age of Head of Household City City/County Complex Transit Networks County Daily Detailed Planning Dynamic Tolls Employment Categories Establishment Characteristics Establishment Size Groups Fares Fixed Tolls General Purpose Lanes Highway Network Supply Highway Operations Hour Household Size Household Socioeconomic Characteristics Income Groups Investment Grade Managed Lanes Microzone Number of Workers in Household Operating Costs Operational Planning Project Design Region Region Road Geometry and Controls Roadway Segment or Parcel Route Simple Transit Networks Accessibility Air Quality Auto Congestion/Traffic Measures Economic Development Energy Consumption Noise Pollution Revenues Safety Schools Truck Congestion/Traffic Measures User Benefits

Methods Reference A-83 Subarea or Corridor Subregion Time Period Transit Network Supply Transit Operations Travel Analysis Zone Travel Demand Management Policies Travel Distance Traveler Demographics Travel Purpose Travel Time Truck Lanes Turn Lanes/Auxiliary Lanes Value-of-Time Segments Vehicle Occupancy Groups Resources Lower-class Methods Method Relationships Advanced Travel Forecasting Background Census Custom Programming Employment Data High-performance Hardware Highway Networks Household Travel Survey Intersection Configurations and Signal Timings Observed Origin- Destination Trip Table Sociodemographic Estimates Standard Travel Forecasting Background Statistics Background Traffic Counts Traffic Microsimulation Background Traffic Microsimulation Software Origin-Destination Matrix Estimation Mode Choice (Disaggregate) Time of Day (Disaggregate) Traffic Microsimulation Dynamic Capacity- Constrained Transit Assignment Dynamic Traffic Assignment Destination Choice Mode Choice (Disaggregate) Time of Day (Disaggregate) Route-Path Choice Models Pedestrian/Bicycle Simulation

A-84 Method Selection for Travel Forecasting: User Guide Traffic Operations— Highway Capacity Manual Traffic Speeds Transportation Network Planning Package Truck Counts Budget $500,000–$1,500,000 Schedule 12–36 months References Nevers, B. L., Nguyen, K. M., Quayle, S. M., Zhou, X., and J. Taylor. 2013. “The Effective Integration of Analysis, Modeling and Simulation.” Research, Development, and Technology. http://www.fhwa.dot.gov/publications/research/operations/13036/13036.pdf (As of Dec. 9, 2016). Sloboden, J., Lewis, J., Alexiadis, V., Chiu, Y., and E. Nava. 2012. “Guidebook on the Utilization of Dynamic Traffic Assignment in Modeling.” Traffic Analysis Toolbox, Vol. XIV, http://ops.fhwa.dot.gov/publications/fhwahop13015/fhwahop13015.pdf (As of Dec. 9, 2016). Intersection Delay Short Description Intersection delay accounts for impacts of intersection turning delays on travel times and route choice. Long Description Intersection delay accounts for impacts of intersection turning delays on travel times and route choice. An intersection delay function can improve the sensitivity of a travel demand model to intersection improvements by explicitly incorporating dynamic turn delays at an intersection into the equilibrium assignment process. Dynamic link penalty functions can be applied in combination with a two-level equilibrium search process to obtain a solution. The proposed intersection delay functions are not directly sensitive to actuated signal controls, employing a fixed capacity for each link and a computed average intersection delay that applies to all turn movements on a link and to all links feeding the intersection. An integrated multiresolution model can be used for better sensitivity to turn movement and signal-timing-specific delay.

Methods Reference A-85 Method Type and Performance Consistency Method type: Aggregate Performance consistency: Somewhat proven Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Arterial Operations and Management Study Congestion Management Plan Environmental Clearance and Preliminary Design for Transportation Projects Environmental Justice Plan Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Major Highway Corridor Study Traffic Impact Study 15-minutes or less City Highway Operations Hour Project Design Road Geometry and Controls Roadway Segment or Parcel Time Period Travel Time Auto Congestion/Traffic Measures Truck Congestion/Traffic Measures Resources Lower-class Methods Method Relationships Custom Programming Dynamic Traffic Assignment High-performance Hardware Highway Networks Traffic Counts Traffic Speeds Dynamic Traffic Assignment Traffic Microsimulation Route-Path Choice Models Multiclass Equilibrium Traffic Assignment Volume-Delay Functions Stochastic User Equilibrium Assignment Single-class Equilibrium Traffic Assignment Budget $50,000–$100,000

A-86 Method Selection for Travel Forecasting: User Guide Schedule 6–12 months References Horowitz, A.J. 1997. “Intersection Delay in Regionwide Traffic Assignment: Implications of 1994 Update of the Highway Capacity Manual.” Transportation Research Record 1572, pp. 1–8. Jeihani, M., Lawe, S., and J.P. Connolly. 2006. “Improving Traffic Assignment Model Using Intersection Delay Function.” Presented at the 47th Annual Transportation Research Forum, New York, NY. Mazloumi, E., Moridpour, S., and H. Mohsenian. 2010. “Delay Function for Signalized Intersections in Traffic Assignment Models.” Journal of Urban Planning and Development, Vol. 136, No. 67, pp. 67–74. Ran, B., and D. Boyce. 1994. Dynamic Urban Transportation Network Models: Theory and Implications for Intelligent Vehicle-Highway Systems. Springer-Verlag Berlin Heidelberg. Multiclass Equilibrium Traffic Assignment Short Description Multiclass equilibrium traffic assignment assigns different classes of vehicles to different routes between same origin-destination pairs. Long Description Multiclass equilibrium traffic assignment assigns different classes of vehicles to different routes between same origin-destination (O-D) pairs. Both single and multiclass equilibrium assignment find the set of link traffic volumes where no single traveler (user) going between a specified O-D pair can shift route and find a faster path. Multiclass assignment makes different subsets of the available routes available to each class of traveler. This feature is useful when modeling the demand for facilities with high-occupancy vehicle lanes or truck lanes. Like all traffic assignment models, single or multiclass user equilibrium methods are used to assign one or more vehicular O-D tables to the highway network. These models employ some measure of impedance, combining travel times, cost, and perhaps reliability to estimate which routes will be used by travelers. These assignment (or route-choice) models may use a relatively simplistic model of traffic congestion to simulate travel times (as a function of volume-capacity ratios). Alternatively, the assignment models may use a more elaborate multiresolution process, combining dynamic traffic assignment with mesoscopic or microscopic simulation models to estimate traveler route preferences and traffic congestion. Method Type and Performance Consistency Method type: Aggregate Performance consistency: Proven or validated

Methods Reference A-87 Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Arterial Operations and Management Study Bicycle and Pedestrian Capital Investments Comprehensive Plans Congestion Management Plan Economic Development Plan Economic Impact Analysis Emergency Evacuation Plan Energy Use Study Environmental Clearance and Preliminary Design for Transportation Projects Environmental Justice Plan Freeway Operations and Management Study Freight Plan Greenhouse Gas Mitigation Study Health and Physical Activity Plan Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Long-Range Transportation Plan Major Highway Corridor Study Major Transit Corridor Study Annual City City/County County Daily Detailed Planning Federal Standards Fixed Tolls General Purpose Lanes High-Level Planning Highway Network Supply Highway Operations Hour Income Groups Investment Grade Managed Lanes Megaregion Operating Costs Operational Planning Region Region Resident Passenger Travel Road Geometry and Controls Roadway Segment or Parcel Route Season/Month State State Standards Subarea or Corridor Subregion Time Period Travel Analysis Zone Travel Distance Travel Time Truck Lanes Turn Lanes/Auxiliary Accessibility Air Quality Auto Congestion/Traffic Measures Economic Development Energy Consumption Noise Pollution Revenues Safety Schools Truck Congestion/Traffic Measures User Benefits

A-88 Method Selection for Travel Forecasting: User Guide Pricing Study Project Prioritization Safety Program Sustainable Community Strategies Traffic Impact Study Transit-Oriented Development Study Transportation Improvement Program Lanes Value-of-Time Segments Vehicle Occupancy Groups Resources Lower-class Methods Method Relationships Highway Networks Standard Hardware Standard Travel Forecasting Background Traffic Counts Transportation Network Planning Package Truck Counts Single-class Equilibrium Traffic Assignment Stochastic User Equilibrium Assignment Mode Choice (Disaggregate) Trip Distribution (Gravity) Mode Choice (Fixed Factors) Time of Day (Fixed Factors) Time of Day (Disaggregate) Pivot-Point and Incremental Models Time of Day (Fixed Factors) Origin-Destination Matrix Estimation Destination Choice Intersection Delay Mode Choice (Fixed Factors) Volume-Delay Functions Schedule-based Transit Assignment Route-Path Choice Models Volume-Delay Functions Budget $25,000–$50,000 Schedule 1–3 months

Methods Reference A-89 References Fundamentals of Transportation/Route Choice, Wikibooks, available at: https://en.wikibooks. org/wiki/Fundamentals_of_Transportation/Route_Choice (As of Dec. 12, 2016). Leurent, F. 1994. “Cost Versus Time Equilibrium over a Network.” Transportation Research Record 1443, pp. 84–91. Federal Highway Administration. 2009. TMIP Email List Technical Synthesis Series 2007-2010, Speed Adjustments Using Volume-Delay Functions. http://www.fhwa.dot.gov/planning/tmip/ publications/other_reports/technical_synthesis_report/page13.cfm (As of Dec. 12, 2016). Pedestrian/Bicycle Simulation Short Description Pedestrian/bicycle simulation models simulate the motion of pedestrians and bicycles in facility. Long Description Pedestrian simulation is used to identify capacity bottlenecks and queue locations in the pedestrian network either inside a structure, at its entrances, or outside on the street or pedestrian walkway. It is often used in the design of major facilities with heavy pedestrian flows, such as airport terminals, transit stations, sports facilities, and for designing major structures to quickly evacuate large crowds in emergencies. Several commercially sponsored websites describe software products that specialize in pedestrian simulation. Bicycle simulation models are used to identify capacity bottlenecks in the bicycle network. Vehicle driving simulators may be used for bicycle collision analysis. Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Somewhat proven Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Bicycle and Pedestrian Capital Investments Bicycle and Pedestrian Plan Environmental Justice Plan Health and Physical Activity Plan 15-minutes or less Hour Microzone Operational Planning Pedestrian/Bike Facilities Pedestrian/Bike Network Supply Accessibility Active Transportation Schools

A-90 Method Selection for Travel Forecasting: User Guide Project Design Resident Passenger Travel Roadway Segment or Parcel Special Markets Subarea or Corridor Time Period Resources Lower-class Methods Method Relationships Built Environment Data Census Employment Data Global Positioning System Surveys High-performance Hardware Pedestrian/Bicycle Simulation Software Pedestrian/Bike Counts Pedestrian/Bike Networks Traffic Microsimulation Background Direct-Demand Model Integrated Multiresolution Model Origin-Destination Matrix Estimation Mode Choice (Disaggregate) Time of Day (Disaggregate) Route-Path Choice Models Traffic Microsimulation Budget $100,000–$200,000 Schedule 6–12 months References Abdelghany, A., Abdelghany, K., Mahmassani, H., and A. Al-Zahrani. 2012. “Dynamic Simulation Assignment Model for Pedestrian Movements in Crowded Networks.” Transportation Research Record: Journal of the Transportation Research Board, No. 2316, pp. 95–105. About the Bicycle Transportation Committee, http://www.pedbikeinfo.com/trbbike/ (As of Dec. 21, 2016). Bak, R., and M. Kiec. 2012. “Influence of Midblock Pedestrian Crossings on Urban Street Capacity.” Transportation Research Record: Journal of the Transportation Research Board, No. 2316, pp. 76–83. Faghri, A., and E. Egyhaziova. 1999. “Development of a Computer Simulation Model of Mixed Motor Vehicle and Bicycle Traffic on an Urban Road Network.” Transportation Research Record: Journal of the Transportation Research Board, No. 1674, 86–93.

Methods Reference A-91 Ottomanelli, M., Iannucci, G., and D. Sassanelli. 2012. “Simplified Model for Pedestrian- Vehicle Interactions at Road Crossings Based on Discrete Events System.” Transportation Research Record: Journal of the Transportation Research Board, No. 2316, pp. 58–68. TRB Pedestrian Committee Official Web Page, http://www.pedbikeinfo.org/trbped/?/trbped/ (As of Dec. 21, 2016). Route-path Choice Models Short Description Route-path choice models are used when choosing motor vehicle paths or transit routes through network. Long Description Route/path choice models identify a set of candidate routes for travelers between each pair of origin and destination points. Traffic assignment models and transit assignment models (also part of route choice modeling) then estimate how many travelers will use each of the candidate routes. The simulations of travelers’ route choice may be based on travel time, service frequency, distance, cost, or on “impedance,” which is a weighted combination of those and other factors. Route choice may be modeled using deterministic or probabilistic models. Route-path choice models are usually built into the selected traffic or transit assignment method. Extra costs would only be incurred if a planner wanted to investigate different path choice models. Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Proven or validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Arterial Operations and Management Study Comprehensive Plans Congestion Management Plan Environmental Clearance and Preliminary Design for Transportation Projects Environmental Justice Plan Freeway Operations and Management Study Freight Plan Annual City City/County Complex Transit Networks County Daily Detailed Planning Dynamic Tolls Fares Federal Standards Fixed Tolls General Purpose Lanes Accessibility Air Quality Auto Congestion/Traffic Measures Economic Development Energy Consumption Noise Pollution Revenues Safety Schools Truck Congestion/Traffic Measures

A-92 Method Selection for Travel Forecasting: User Guide Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Long-Range Transportation Plan Major Highway Corridor Study Major Transit Corridor Study Pricing Study Project Prioritization Traffic Impact Study Transit Operations Study Transit-Oriented Development Study Transportation Improvement Program Highway Network Supply Highway Operations Hour Managed Lanes Megaregion Microzone Operating Costs Region Region Road Geometry and Controls Roadway Segment or Parcel Route Season/Month Simple Transit Networks State Subarea or Corridor Subregion Time Period Transit Network Supply Transit Operations Travel Analysis Zone Travel Distance Travel Time Truck Lanes Turn Lanes/Auxiliary Lanes User Benefits Resources Lower-class Methods Method Relationships Advanced Travel Forecasting Background Global Positioning System Surveys Highway Networks Standard Hardware Statistics Background Traffic Counts Traffic Operations— Highway Capacity Manual Traffic Speeds Transit Counts Transit Networks Volume-Delay Functions Intersection Delay Dynamic Traffic Assignment Integrated Multiresolution Model Pedestrian/Bicycle Simulation Single-class Equilibrium Traffic Assignment Multiclass Equilibrium Traffic Assignment Stochastic User

Methods Reference A-93 Transit On-Board Survey Transportation Network Planning Package Truck Counts Truck Speeds Truck Surveys Equilibrium Assignment Budget $15,000–$30,000 Schedule 1–6 months References Ben-Akiva, M.E., Ramming, M.S., and S. Bekhor. 2004. “Route Choice Models.” Human Behaviour and Traffic Networks, pp. 23–45. Schreckenberg, M., and R. Selten (Eds.). 2004. Human Behaviour and Traffic Networks. Springer-Verlag Berlin Heidelberg. Wikipedia. “Route Assignment.” https://en.wikipedia.org/wiki/Route_assignment (Accessed on Dec. 21, 2016). Schedule-based Transit Assignment Short Description Schedule-based transit assignment allows travelers to choose transit routes based on detailed departure or arrival times of each transit vehicle. Long Description Schedule based transit assignment considers transit schedules, transfer wait times (as appropriate), and individual vehicle capacities in assigning transit passengers to buses and routes. Wait times are determined by the time between scheduled departures of a transit route instead of headways or frequency of service. Method Type and Performance Consistency Method type: Aggregate Performance consistency: Somewhat proven

A-94 Method Selection for Travel Forecasting: User Guide Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Comprehensive Plans Congestion Management Plan Economic Impact Analysis Energy Use Study Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Long-Range Transportation Plan Major Transit Corridor Study Pricing Study Project Prioritization Transit Operations Study Transit-Oriented Development Study Transportation Improvement Program Travel Demand Management Program 15-minutes or less City City/County Complex Transit Networks County Detailed Planning Fares Federal Standards Hour Megaregion Microzone Operational Planning Region Region Roadway Segment or Parcel Route Simple Transit Networks State Subarea or Corridor Subregion Time Period Transit Network Supply Transit Operations Travel Analysis Zone Value-of-Time Segments Accessibility Economic Development Revenues Schools Transit Measures User Benefits Resources Lower-class Methods Method Relationships Highway Networks Standard Hardware Standard Travel Forecasting Background Transit Counts Transit Networks Transportation Network Planning Package Frequency-based Transit Assignment Pivot-Point and Incremental Models Direct-Demand Model Mode Choice (Disaggregate) Mode Choice (Fixed Factors) Time of Day (Fixed Factors) Multiclass Equilibrium

Methods Reference A-95 Traffic Assignment Single-class Equilibrium Traffic Assignment Stochastic User Equilibrium Assignment Time of Day (Disaggregate) Budget $50,000–$100,000 Schedule 3–6 months References Hamdouch, Y., and S. Lawphongpanich. 2008. “Schedule-Based Transit Assignment Model with Travel Strategies and Capacity Constraints.” Transportation Research Part B: Methodological, Vol. 42, No. 7–8, pp. 663–684. Hamdouch, Y., Ho, H.W., Sumalee, A., and G. Wang. 2011. “Schedule-Based Transit Assignment Model with Vehicle Capacity and Seat Availability.” Transportation Research Part B: Methodological, Vol. 45, No. 10, pp. 1805–1830. Single-class Equilibrium Traffic Assignment Short Description Single-class equilibrium traffic assignment is a method for balancing link traffic flows on congested networks. Long Description Single-class equilibrium traffic assignment finds the set of link traffic volumes where no single traveler (user) going between a specified origin and destination can shift route and find a faster path. Like all traffic assignment models, single-class user equilibrium methods assign one vehicular origin-destination table to the highway network. These models employ some measure of impedance, combining travel times, cost, and reliability to estimate which routes will be used by travelers. These assignment (or route-choice) models may use a relatively simple model of traffic congestion to simulate travel times (e.g., based on volume-capacity ratios). Alternatively, the assignment models may use a more elaborate multiresolution process, combining dynamic traffic assignment with mesoscopic or microscopic simulation models to estimate traveler route preferences and traffic congestion.

A-96 Method Selection for Travel Forecasting: User Guide Method Type and Performance Consistency Method type: Aggregate Performance consistency: Proven or validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Arterial Operations and Management Study Bicycle and Pedestrian Capital Investments Comprehensive Plans Congestion Management Plan Economic Development Plan Economic Impact Analysis Emergency Evacuation Plan Energy Use Study Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Freeway Operations and Management Study Freight Plan Greenhouse Gas Mitigation Study Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Long-Range Transportation Plan Annual City City/County County Daily Detailed Planning Federal Standards Fixed Tolls General Purpose Lanes High-Level Planning Highway Network Supply Highway Operations Hour Income Groups Investment Grade Managed Lanes Megaregion Operating Costs Operational Planning Region Region Resident Passenger Travel Road Geometry and Controls Roadway Segment or Parcel Route Season/Month State State Standards Subarea or Corridor Subregion Time Period Travel Analysis Zone Accessibility Air Quality Auto Congestion/Traffic Measures Economic Development Energy Consumption Noise Pollution Safety Schools

Methods Reference A-97 Major Highway Corridor Study Major Transit Corridor Study Pricing Study Project Prioritization Safety Program Sustainable Community Strategies Traffic Impact Study Transit-Oriented Development Study Transportation Improvement Program Travel Demand Management Program Travel Distance Travel Time Truck Lanes Turn Lanes/Auxiliary Lanes Value-of-Time Segments Vehicle Occupancy Groups Resources Lower-class Methods Method Relationships Highway Networks Standard Hardware Standard Travel Forecasting Background Traffic Counts Transportation Network Planning Package Origin-Destination Matrix Estimation Direct-Demand Model Pivot-Point and Incremental Models Time of Day (Fixed Factors) Mode Choice (Fixed Factors) Volume-Delay Functions Schedule-based Transit Assignment Trip Distribution (Gravity) Mode Choice (Fixed Factors) Time of Day (Fixed Factors) Intersection Delay Route-Path Choice Models Volume-Delay Functions Budget $25,000–$50,000

A-98 Method Selection for Travel Forecasting: User Guide Schedule 1–3 months References Dial, R. 1995. “Multicriterion Equilibrium Traffic Assignment: Basic Theory and Elementary Algorithms.” Federal Highway Administration Travel Model Improvement Program. Federal Highway Administration. 2009. TMIP Email List Technical Synthesis Series 2007-2010, Speed Adjustments Using Volume-Delay Functions. http://www.fhwa.dot.gov/planning/tmip/publications/other_reports/technical_synthesis_report/pa ge13.cfm (As of Dec. 12, 2016). Fundamentals of Transportation/Route Choice, Wikibooks, available at: https://en.wikibooks.org/wiki/Fundamentals_of_Transportation/Route_Choice (As of Dec. 12, 2016). Stochastic User Equilibrium Assignment Short Description Stochastic User Equilibrium is a capacity-constrained traffic assignment method. Long Description Stochastic User Equilibrium is a capacity-constrained traffic assignment method that identifies traffic flows where no driver can lower his or her perceived travel time by unilaterally changing his or her current route. Unlike single or multiclass user equilibrium, stochastic user equilibrium assumes there is an unobserved random utility component in the choice of path. Method Type and Performance Consistency Method type: Aggregate Performance consistency: Somewhat proven Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Arterial Operations and Management Study Bicycle and Pedestrian Capital Investments Comprehensive Plans Annual City City/County County Daily Detailed Planning Federal Standards Fixed Tolls Accessibility Air Quality Auto Congestion/Traffic Measures Economic Development Energy Consumption Noise Pollution Revenues

Methods Reference A-99 Congestion Management Plan Economic Development Plan Economic Impact Analysis Emergency Evacuation Plan Energy Use Study Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Freeway Operations and Management Study Freight Plan Greenhouse Gas Mitigation Study Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Long-Range Transportation Plan Major Highway Corridor Study Major Transit Corridor Study Pricing Study Project Prioritization Safety Program Sustainable Community Strategies Traffic Impact Study Transit-Oriented Development Study Transportation Improvement Program Travel Demand Management Program General Purpose Lanes High-Level Planning Highway Network Supply Highway Operations Hour Income Groups Investment Grade Managed Lanes Megaregion Operating Costs Operational Planning Region Region Resident Passenger Travel Road Geometry and Controls Roadway Segment or Parcel Route Season/Month State State Standards Subarea or Corridor Subregion Time Period Travel Analysis Zone Travel Distance Travel Time Truck Lanes Turn Lanes/Auxiliary Lanes Value-of-Time Segments Vehicle Occupancy Groups Safety Schools

A-100 Method Selection for Travel Forecasting: User Guide Resources Lower-class Methods Method Relationships Highway Networks Standard Hardware Standard Travel Forecasting Background Traffic Counts Transportation Network Planning Package Time of Day (Fixed Factors) Volume-Delay Functions Mode Choice (Fixed Factors) Schedule-based Transit Assignment Origin-Destination Matrix Estimation Mode Choice (Disaggregate) Direct-Demand Model Intersection Delay Trip Distribution (Gravity) Mode Choice (Fixed Factors) Time of Day (Fixed Factors) Destination Choice Pivot-Point and Incremental Models Route-Path Choice Models Volume-Delay Functions Budget $25,000–$50,000 Schedule 1–4 months References Fundamentals of Transportation/Route Choice, Wikibooks, available at: https://en.wikibooks.org/wiki/Fundamentals_of_Transportation/Route_Choice (As of Dec. 12, 2016). Maher, M. 1998. “Algorithms for Logit-Based Stochastic User Equilibrium Assignment.” Transportation Research Part B: Methodological, Vol. 32, No. 8, pp. 539–549.

Methods Reference A-101 Traffic Microsimulation Short Description Traffic microsimulation simulates the movements and interactions of individual vehicles on facilities. Long Description Traffic microsimulation models have become the prevailing methodology for modeling the response of individual drivers to changes within the network. Macroscopic and mesoscopic dynamic traffic assignment models allow practitioners to simulate large networks, but these models do not provide a detailed analysis into individual motorists’ lane choices. A microscopic model allows practitioners to evaluate the behavior of a vehicle within the traffic stream, whether done with a car-following, lane-changing, or route-choice model. The use of microscopic simulation requires detailed input; the calibration and validation of model inputs can be arduous. When selecting a simulation type, practitioners must be cognizant of the level of detail that the project requires. Traffic microsimulation software is used to model the traffic performance of highways, streets, transit, and pedestrian facilities where second-by-second analysis is useful. However, it is important to consider the scope of such projects, where careful consideration must be paid to securing the proper expertise, allotting sufficient time and financial resources, and fostering a detailed database for the development of a base model. Traffic microsimulation can be time consuming and data intensive, but there are instances where microsimulations are warranted and cost effective: Detailed interactions of vehicle movements are the principal motivation for the study (these include interactions of vehicles with pedestrians/bicyclists, HOVs and buses, and congested traffic). Alternative design considerations are needed related to lane changing. The visual animation of traffic conditions will improve the credibility of potential solutions. Traffic microsimulation has also been used at a regional scale. Several factors could hinder the successful application of the 4-step travel model toward simulating individual activities and travel (i.e., extensive data and computational requirements). However, microsimulation showed potential for significantly improving existing travel forecasting procedures used by MPOs and state DOTs. Method Type and Performance Consistency Method type: Disaggregate Performance consistency: Proven or validated

A-102 Method Selection for Travel Forecasting: User Guide Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Arterial Operations and Management Study Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Freeway Operations and Management Study Highway Detailed Design Highway Preliminary Engineering Intelligent Transportation Systems Plan Major Highway Corridor Study Safety Program Traffic Impact Study 15-minutes or less General Purpose Lanes Highway Network Supply Highway Operations Hour Managed Lanes Operational Planning Project Design Road Geometry and Controls Roadway Segment or Parcel Subarea or Corridor Time Period Travel Distance Travel Time Truck Lanes Turn Lanes/Auxiliary Lanes Accessibility Air Quality Auto Congestion/Traffic Measures Energy Consumption Noise Pollution Safety Schools Truck Congestion/Traffic Measures Resources Lower-class Methods Method Relationships High-performance Hardware Highway Networks Traffic Counts Traffic Microsimulation Background Traffic Microsimulation Software Traffic Operations— Highway Capacity Manual Traffic Speeds Integrated Multiresolution Model Time of Day (Disaggregate) Mode Choice (Disaggregate) Intersection Delay Destination Choice Origin-Destination Matrix Estimation Parking Location Choice Pedestrian/Bicycle Simulation Budget $100,000–$250,000

Methods Reference A-103 Schedule 6–18 months References Cervenka, K. 1997. “Large-Scale Traffic Microsimulation from an MPO Perspective.” Presented at 6th TRB Conference on the Application of Transportation Planning Methods, Dearborn, MI. Dowling, R., Skabardonis, A., and V. Alexiadis. 2004. Traffic Analysis Toolbox Volume III: Guidelines for Applying Traffic Microsimulation Modeling Software. http://ops.fhwa.dot.gov/trafficanalysistools/tat_vol3/volguidelines.pdf (As of Dec. 12, 2016). Sbayti, H., and D. Roden. 2010. “Best Practices in the Use of Micro Simulation Models.” http://statewideplanning.org/wp-content/uploads/259_NCHRP-08-36-90.pdf (As of Dec. 12, 2016). Volume-delay Functions Short Description Volume-delay functions simulate changes in traffic speed as a function of changes in volume and the characteristics of the link. Long Description Volume-Delay functions can take many forms, if it continuously increases (in terms of travel time) with increasing volumes. A common volume-delay function is the Bureau of Public Roads curve shown below: T=T0 (1+Ax^B) Equation 1 where T = link travel time (h); T0 = link travel time at low, near-zero volumes (h); A = ratio of speed at capacity to free-flow speed, minus one (standard value = 0.15); B = parameter that affects the rate at which speed drops (standard value = 4.0); and x = the link demand-to-capacity ratio (unitless). The calibration parameter A is selected so that the travel time equation will simulate the mean speed of traffic when demand is equal to capacity. Substituting x = 1.00 in the travel time equation and solving for A yields: A=Sf/Sc–1 Equation 2 where

A-104 Method Selection for Travel Forecasting: User Guide A = BPR speed-at-capacity calibration parameter, Sc = mean speed at capacity (mph), and Sf = mean free-flow speed (mph). The calibration parameter B is selected to simulate the approximate delay when demand exceeds capacity for a target range of demand-to-capacity ratios (generally for v/c ratios in the range of 1.7 to 1.9). Method Type and Performance Consistency Method type: Aggregate Performance consistency: Proven or validated Programs, Requirements, Metrics, Resources, Lower-class Methods, and Method Relationships Programs Requirements Metrics Air Quality Emissions Inventory for Conformity Analysis Arterial Operations and Management Study Comprehensive Plans Congestion Management Plan Economic Development Plan Economic Impact Analysis Energy Use Study Environmental Clearance and Preliminary Design for Transportation Projects Environmental Impact Study Environmental Justice Plan Freeway Operations and Management Study Freight Plan Greenhouse Gas Mitigation Study Highway Detailed Design Highway Preliminary Engineering City City/County County Daily Detailed Planning Dynamic Tolls Fares Fixed Tolls General Purpose Lanes High-Level Planning Highway Network Supply Highway Operations Hour Investment Grade Managed Lanes Megaregion Operating Costs Operational Planning Region Region Road Geometry and Controls Roadway Segment or Parcel Route State Auto Congestion/Traffic Measures Truck Congestion/Traffic Measures

Methods Reference A-105 Intelligent Transportation Systems Plan Long-Range Transportation Plan Major Highway Corridor Study Major Transit Corridor Study Pricing Study Project Prioritization Safety Program Sustainable Community Strategies Traffic Impact Study Transit-Oriented Development Study Transportation Improvement Program Travel Demand Management Program Subarea or Corridor Subregion Time Period Travel Analysis Zone Travel Time Truck Lanes Turn Lanes/Auxiliary Lanes Value-of-Time Segments Resources Lower-class Methods Method Relationships Highway Networks Standard Hardware Standard Travel Forecasting Background Traffic Counts Traffic Operations— Highway Capacity Manual Traffic Speeds Transportation Network Planning Package Route-Path Choice Models Pivot-Point and Incremental Models Stochastic User Equilibrium Assignment Multiclass Equilibrium Traffic Assignment Single-class Equilibrium Traffic Assignment Intersection Delay Multiclass Equilibrium Traffic Assignment Single-class Equilibrium Traffic Assignment Stochastic User Equilibrium Assignment Budget $25,000–$50,000

A-106 Method Selection for Travel Forecasting: User Guide Schedule 1–3 months References Cambridge Systematics, Inc., Global Insight, Cohen H., Horowitz A., and R. Pendyala. 2008. NCHRP Report 606: Forecasting Statewide Freight Toolkit. Transportation Research Board, Washington, DC. Dowling, R., Ryus, P., Schroeder, B., Kyte, M., Creasey, F.T., Rouphail, N., Hajbabaie, A., and D. Rhoades. 2016. NCHRP Report 825: Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual, Transportation Research Board, Washington, DC. Moses, R., Mtoi, E., McBean, H., and S. Ruegg. 2013. “Development of Speed Models for Improving Travel Forecasting and Highway Performance Evaluation.” Final Report for Florida Department of Transportation.

Methods Reference A-107 REFERENCES Abdelghany, A., Abdelghany, K., Mahmassani, H., and A. Al-Zahrani. 2012. “Dynamic Simulation Assignment Model for Pedestrian Movements in Crowded Networks.” Transportation Research Record: Journal of the Transportation Research Board, No. 2316, pp. 95–105. About the Bicycle Transportation Committee, http://www.pedbikeinfo.com/trbbike/ (As of Dec. 21, 2016). Abraham, J.E., Stefan, K.J., and J.D. Hunt. 2012. “Population Synthesis Using Combinatorial Optimization at Multiple Levels.” Presented at the Transportation Research Board 91st Annual Meeting, Washington, DC. Adler, T., and M. Ben-Akiva. 1977. “A Theoretical and Empirical Model of Trip Chaining Behavior.” Transportation Research B, Vol. 13B, pp. 243–257. Allen Jr., W.G. 1983. “Trip Distribution Using Composite Impedance.” Transportation Research Record 944, pp. 118–127. Allen Jr., W.G., and G.W. Schultz. 1996. “Congestion-Based Peak Spreading Model.” Transportation Research Record 1556, pp. 8–15. Alsup, R., Freedman, J., Bettinardi, A., and S. Payne. 2014. “A University Student Tour-based Travel Demand Model.” Presented at the Innovations in Travel Demand Forecasting Conference, Baltimore, MD. Arentze, T., and H. Timmermans. 2004. “Albatross: A Learning-based Transportation Oriented Simulation System.” Transportation Research Part B Methodological, Vol. 38, No. 7, pp. 613– 633. Bak, R., and M. Kiec. 2012. “Influence of Midblock Pedestrian Crossings on Urban Street Capacity.” Transportation Research Record: Journal of the Transportation Research Board, No. 2316, pp. 76–83. Beckman, R.J., K.A. Baggerly, and M.D. McKay. 1996. “Creating Synthetic Baseline Populations.” Transportation Research Part A: Policy and Practice, Vol. 30, No. 6, pp. 415–429. Ben-Akiva, M., and S.R. Lerman. 1985. Discrete Choice Analysis: Theory and Application to Travel Demand. The MIT Press. Ben-Akiva, M.E., Ramming, M.S., and S. Bekhor. 2004. “Route Choice Models.” Human Behaviour and Traffic Networks, pp. 23–45. Bernardin, V.L., Koppelman, F., and D. Boyce. 2009. “Enhanced Destination Choice Models Incorporating Agglomeration Related to Trip Chaining While Controlling for Spatial Competition.” Transportation Research Record: Journal of the Transportation Research Board, No. 2132, pp. 143–151. Bhat, C., and J. Steed. 2002. “A Continuous-Time Model of Departure Time Choice for Urban Shopping Trips.” Transportation Research Part B: Methodological, Vol. 36, No. 3, pp. 207–224.

A-108 Method Selection for Travel Forecasting: User Guide Bhat, C., Govindarajan, A., and V. Pulugata. 1998. “Disaggregate Attraction-End Choice Modeling: Formulation and Empirical Analysis.” Transportation Research Record: Journal of the Transportation Research Board 1645, pp. 60–68. Bhat, C.R. 2000. “Incorporating Observed and Unobserved Heterogeneity in Urban Work Mode Choice Modeling.” Transportation Science, Vol. 34, No. 2, pp. 228–238. Bhat, C.R., Guo, J.Y., Srinivasan, S., and A. Sivakumar. 2004. “Comprehensive Econometric Microsimulator for Daily Activity-Travel Patterns. ” Transportation Research Record: Journal of the Transportation Research Board, No. 1894, pp. 57–66. Borgers, A., and H. Timmermans. 1987. “Choice Model Specification, Substitution and Spatial Structure Effects: A Simulation Experiment.” Regional Science and Urban Economics, Vol. 17, No. 1, pp. 29–47. Bowman, J., and M. Ben-Akiva. 1997. “Activity-based Travel Forecasting.” Presented at the Activity-based Travel Forecasting Conference, Washington, DC. Bradley, M., and P. Vovsha. 2005. “A Model for Joint Choice of Daily Activity Pattern Types of Household Members.” Transportation, Vol. 32, No. 5, pp. 545–571. Cambridge Systematics, Inc. 1999. “Time-of-Day Modeling Procedures: State-of-the-Art, State- of-the-Practice,” DOT-T-99-01, US Department of Transportation, Washington, DC. https://www.fhwa.dot.gov/planning/tmip/publications/other_reports/tod_modeling_procedures/c h01.cfm (Accessed on Dec. 21, 2016). Cambridge Systematics, Inc. 2010. “Travel Model Validation and Reasonableness Checking Manual - Second Edition.” Prepared for the Federal Highway Administration Travel Model Improvement Program. FHWA-HEP-10-042. https://www.fhwa.dot.gov/planning/tmip/publications/other_reports/validation_and_reasonablen ess_2010/ (Accessed on Dec. 21, 2016). Cambridge Systematics, Inc. 2011. A Working Demonstration of a Mesoscale Freight Model for the Chicago Region Final Report and User’s Guide. Prepared for the Chicago Metropolitan Agency for Planning. Cambridge Systematics, Inc., Global Insight, Cohen H., Horowitz A., and R. Pendyala. 2008. NCHRP Report 606: Forecasting Statewide Freight Toolkit. Transportation Research Board, Washington, DC. Cambridge Systematics, Inc., Vanasse Hangen Brustlin, Inc., Gallop Corporation, Bhat, C., Shapiro Transportation Consulting, LLC, Martin/Alexiou/Bryson, PLLC. 2012. NCHRP Report 716: Travel Demand Forecasting: Parameters and Techniques. Transportation Research Board, Washington, DC. Cavalcante, R., and M.J. Roorda. 2010. “A Disaggregate Urban Shipment Size/Vehicle-Type Choice Model.” Presented at the Transportation Research Board 89th Annual Meeting, Washington, DC.

Methods Reference A-109 Center for Urban Transportation Studies, University of Wisconsin – Milwaukee and Wisconsin Department of Transportation. 1999. Guidebook on Statewide Travel Forecasting. https://www.fhwa.dot.gov/planning/processes/statewide/forecasting/swtravel.pdf (As of Dec. 12, 2016). Cepeda, M., Cominetti, R., and M. Florian. 2006. “A Frequency-Based Assignment Model for Congested Transit Networks with Strict Capacity Constraints: Characterization and Computation of Equilibria.” Transportation Research Part B: Methodological, Vol. 40, No. 6, pp. 437–459. Cervenka, K. 1997. “Large-Scale Traffic Microsimulation from an MPO Perspective.” Presented at 6th TRB Conference on the Application of Transportation Planning Methods, Dearborn, MI. Chiu, Y-C., Bottom, J., Mahut, M., Paz, A., Balakrishna, R., Waller, T., and J. Hicks. 2011. Transportation Research Circular E-C153: Dynamic Traffic Assignment: A Primer. http://onlinepubs.trb.org/onlinepubs/circulars/ec153.pdf (As of Dec. 9, 2016). Chow, L.-F., Zhao, F., Li, M.-T., and S.-C. Li. 2005. “Development and Evaluation of Aggregate Destination Choice Models for Trip Distribution in Florida.” Transportation Research Record: Journal of the Transportation Research Board, No. 1931, pp. 18–27. Daly, A. 1982. “Estimating Choice Models Containing Attraction Variables.” Transportation Research Part B: Methodological, Vol. 16, No. 1, pp. 5–15. de Dios Ortúzar, J. and L.G. Willumsen. 1990. Modelling Transport. Wiley. de Jong, G., and M. Ben-Akiva. 2007. A Micro-Simulation Model of Shipment Size and Transport Chain Choice. Transportation Research Part B: Methodological, Vol. 41, No. 9, pp.950–965. Dial, R. 1995. “Multicriterion Equilibrium Traffic Assignment: Basic Theory and Elementary Algorithms.” Federal Highway Administration Travel Model Improvement Program. Dowling, R., Ryus, P., Schroeder, B., Kyte, M., Creasey, F.T., Rouphail, N., Hajbabaie, A., and D. Rhoades. 2016. NCHRP Report 825: Planning and Preliminary Engineering Applications Guide to the Highway Capacity Manual, Transportation Research Board, Washington, DC. Dowling, R., Skabardonis, A., and V. Alexiadis. 2004. Traffic Analysis Toolbox Volume III: Guidelines for Applying Traffic Microsimulation Modeling Software. http://ops.fhwa.dot.gov/trafficanalysistools/tat_vol3/volguidelines.pdf (As of Dec. 12, 2016). Faghri, A., and E. Egyhaziova. 1999. “Development of a Computer Simulation Model of Mixed Motor Vehicle and Bicycle Traffic on an Urban Road Network.” Transportation Research Record: Journal of the Transportation Research Board, No. 1674, 86–93. Fagnant, D.J., and K. Kockelman. 2016. A Direct-Demand Model for Bicycle Counts: The Impacts of Level of Service and Other Factors. Environment and Planning B: Planning and Design. Vol. 43, pp. 93–107. Federal Highway Administration. 2009. TMIP Email List Technical Synthesis Series 2007-2010, Speed Adjustments Using Volume-Delay Functions.

A-110 Method Selection for Travel Forecasting: User Guide http://www.fhwa.dot.gov/planning/tmip/publications/other_reports/technical_synthesis_report/pa ge13.cfm (As of Dec. 12, 2016). Florian, M., He, S., and S. Velan. 2010. “A Practical Method for Adjusting Temporal Origin- Destination Matrices for Dynamic Traffic Assignment.” Submitted to TRB Innovations in Travel Demand Forecasting. Fotheringham, A.S. 1983. “Some Theoretical Aspects of Destination Choice and Their Relevance to Production-Constrained Gravity Models.” Environment and Planning, Vol. 15, No. 8, pp. 1121–1132. Fox, J., Daly, A., and H. Gunn. 2005. “Review of RAND Europe’s Transport Demand Model Systems.” Prepared for TRL Limited by RAND Europe. https://www.rand.org/content/dam/rand/pubs/monograph_reports/2005/MR1694.pdf (As of Dec. 12, 2016). Fu, Q., Liu, R., and S. Hess. 2012. “A Review on Transit Assignment Modelling Approaches to Congested Networks: A New Perspective.” Procedia—Social and Behavioral Sciences, Vol. 54, No. 4, pp. 1145–1155. Fundamentals of Transportation/Route Choice, Wikibooks, available at: https://en.wikibooks.org/wiki/Fundamentals_of_Transportation/Route_Choice (As of Dec. 12, 2016). Gentile, G., and K. Noekel (Eds.). 2016. Modeling Public Transport Passenger Flows in the Era of Intelligent Transport Systems. Springer. Gosling, G. 2008. ACRP Synthesis of Airport Practice 5: Airport Ground Access Mode Choice Models. Transportation Research Board, Washington, DC. Gunn, H. 1994. “The Netherlands National Model: A Review of Seven Years of Application.” International Association of Operational Research, Vol.1, No. 2, pp. 125–133. Hamdouch, Y., and S. Lawphongpanich. 2008. “Schedule-Based Transit Assignment Model with Travel Strategies and Capacity Constraints.” Transportation Research Part B: Methodological, Vol. 42, No. 7-8, pp. 663–684. Hamdouch, Y., Ho, H.W., Sumalee, A., and G. Wang. 2011. “Schedule-Based Transit Assignment Model with Vehicle Capacity and Seat Availability.” Transportation Research Part B: Methodological, Vol. 45, No. 10, pp. 1805–1830. Hensher, D., Rose, J.M., and W.H. Greene. 2005. Applied Choice Analysis: A Primer. Cambridge University Press. Holguin-Veras, J., Jaller, M., Sanchez-Diaz, I., Wojtowicz, J., Campbell, S., Levinson, H., Lawson, C., Powers, E. L., and L. Tavasszy. 2012. NCHRP Report 739/NCFRP Report 19: Freight Trip Generation and Land Use. Transportation Research Board, Washington, DC. Horowitz, A.J. 1997. “Intersection Delay in Regionwide Traffic Assignment: Implications of 1994 Update of the Highway Capacity Manual.” Transportation Research Record 1572, pp. 1–8.

Methods Reference A-111 Hunt, J.D., and S. Teply. 1993. “A Nested Logit Model of Parking Location Choice.” Transportation Research Part B: Methodological, Vol. 27, No. 4, pp. 253–265. Jeihani, M., Lawe, S., and J.P. Connolly. 2006. “Improving Traffic Assignment Model Using Intersection Delay Function.” Presented at the 47th Annual Transportation Research Forum, New York, NY. Jonnalagadda, N., Freedman, J., Davidson, W.A., and J.D. Hunt. 2001. “Development of Microsimulation Activity-Based Model for San Francisco: Destination and Mode Choice Models.” Transportation Research Record: Journal of the Transportation Research Board, No. 1777, pp. 25–35. Koppelman, F.S., and C. Bhat. 2006. A Self-Instructing Course in Mode Choice Modeling: Multinomial and Nested Logit Models. US Department of Transportation, Federal Transit Administration. Kuzmyak, J.R. 2008. NCHRP Synthesis of Highway Practice 384: Forecasting Metropolitan Commercial and Freight Travel. Transportation Research Board, Washington, D.C. Leachman, R., Prince, T., Brown T., and G. Fetty. 2005. Final Report: Port and Modal Elasticity Study. Southern California Association of Governments. Leurent, F. 1994. “Cost Versus Time Equilibrium over a Network.” Transportation Research Record 1443, pp. 84–91. Litman, T. 2004. “Transit Price Elasticities and Cross-Elasticities.” Journal of Public Transportation, Vol. 7, No. 2., pp. 37–58. Maher, M. 1998. “Algorithms for Logit-Based Stochastic User Equilibrium Assignment.” Transportation Research Part B: Methodological, Vol. 32, No. 8, pp. 539–549. Mazloumi, E., Moridpour, S., and H. Mohsenian. 2010. “Delay Function for Signalized Intersections in Traffic Assignment Models.” Journal of Urban Planning and Development, Vol. 136, No. 67, pp. 67–74. McFadden, D., Talvitie, A., Cosslett, S., Hasan, I., Johnson, M., Reid, A., and K. Train. 1977. “Demand Model Estimation and Validation.” The Urban Travel Demand Forecasting Project, Volume V. Meyer, M., and E.J. Miller. 2000. Urban Transportation Planning. McGraw-Hill. Mishra, S., Wang, Y., Zhu, X., Moeckel, R., and S. Mahaparta. 2013. “Comparison between Gravity and Destination Choice Models for Trip Distribution in Maryland.” Presented at Transportation Research Board 92nd Annual Meeting, Washington, DC. Morgan, D., and R. Mayberry. 2010. “Application of a Combined Travel Demand and Microsimulation Model for a Small City.” http://www.caliper.com/PDFs/Morgan_Paper.pdf (As of Dec. 9, 2016).

A-112 Method Selection for Travel Forecasting: User Guide Moses, R., Mtoi, E., McBean, H., and S. Ruegg. 2013. “Development of Speed Models for Improving Travel Forecasting and Highway Performance Evaluation.” Final Report for Florida Department of Transportation. Müller, K., and K.W. Axhausen. 2010. “Population Synthesis for Microsimulation: State of the Art.” ETH Zürich, Institut für Verkehrsplanung, Transporttechnik, Strassen-und Eisenbahnbau (IVT). Nevers, B.L., Nguyen, K.M., Quayle, S.M., Zhou, X., and J. Taylor. 2013. “The Effective Integration of Analysis, Modeling and Simulation.” Research, Development, and Technology, http://www.fhwa.dot.gov/publications/research/operations/13036/13036.pdf (As of Dec. 9, 2016). Nökel, K., and S. Wekeck. 2009. “Boarding and Alighting in Frequency-Based Transit Assignment.” Transportation Research Record: Journal of the Transportation Research Board, No. 2111, pp. 60-67. Ottomanelli, M., Iannucci, G., and D. Sassanelli. 2012. “Simplified Model for Pedestrian- Vehicle Interactions at Road Crossings Based on Discrete Events System.” Transportation Research Record: Journal of the Transportation Research Board, No. 2316, pp. 58–68. Papola, N., Francesco, F., Gentile, G., and L. Meschini. 2008. “Schedule-Based Transit Assignment: New Dynamic Equilibrium Model with Vehicle Capacity Constraints.” Schedule- Based Modeling of Transportation Networks: Operations Research/Computer Science Interfaces Series, Vol. 46, pp. 1–26. Pourabdollahi, Z., Karimi, B., Mohammadian, A., and K. Kawamura. 2014. “Shipping Chain Choices in Long-Distance Supply Chains: Descriptive Analysis and a Decision Tree Model.” Submitted to the Transportation Research Board 93rd Annual Meeting, Washington, DC. Ran, B., and D. Boyce. 1994. Dynamic Urban Transportation Network Models: Theory and Implications for Intelligent Vehicle-Highway Systems. Springer-Verlag Berlin Heidelberg. RDC, Inc. 1995. “Activity-Based Modeling System for Travel Demand Forecasting.” US Department of Transportation, Washington, DC, Report DOT-T-96-02. RSG, University of Illinois at Chicago, and J. Bowman. 2012. Tour-Based and Supply Chain Freight Forecasting Framework: Final Report (BAA DTFH61-10-R-00013). Federal Highway Administration. Ruan, M., Lin, J., and K. Kawamura. 2011. “Modeling Commercial Vehicle Daily Tour Chaining.” Presented at the Transportation Research Board 90th Annual Meeting, Washington, DC. Samimi, A., Mohammadian, A., and K. Kawamura. 2010. A Behavioral Freight Movement Microsimulation Model: Method and Data. Transportation Letters: The International Journal of Transportation Research, Vol. 2, pp. 53–62.

Methods Reference A-113 Sbayti, H., and D. Roden. 2010. “Best Practices in the Use of Micro Simulation Models.” http://statewideplanning.org/wp-content/uploads/259_NCHRP-08-36-90.pdf (As of Dec. 12, 2016). Schmöcker, J-D. 2006. “Dynamic Capacity-Constrained Transit Assignment.” PhD dissertation, Centre for Transport Studies, Department for Civil and Environmental Engineering, Imperial College London. Schmöcker, J-D., Bell, M.G.H., and F. Kurauchi. 2008. “A Quasi-Dynamic Capacity Constrained Frequency-Based Transit Assignment Model.” Transportation Research Part B: Methodological, Vol. 42, No. 11, pp. 925–945. Schmöcker, J-D., Fonsone, A., Shimamota, H., Kurauchi, F., and M.G.H. Bell. 2011. “Frequency-Based Transit Assignment Considering Seat Capacities.” Transportation Research Part B: Methodological, Vol. 45, No. 2, pp. 392–408. Schreckenberg, M., and R. Selten (Eds.). 2004. Human Behaviour and Traffic Networks. Springer-Verlag Berlin Heidelberg. Scott, D. M., and K. W. Axhausen. 2006. “Household Mobility Tool Ownership: Modeling Interactions Between Cars and Season Tickets.” Transportation, Vol. 33, No. 4, pp. 311–328. Sloboden, J., Lewis, J., Alexiadis, V., Chiu, Y., and E. Nava. 2012. “Guidebook on the Utilization of Dynamic Traffic Assignment in Modeling.” Traffic Analysis Toolbox, Vol. XIV, http://ops.fhwa.dot.gov/publications/fhwahop13015/fhwahop13015.pdf (As of Dec. 9, 2016). Smith, C., Chen, J., Sana, B., and M. Outwater. 2013. “A Disaggregate Tour-Based Truck Model with Simulation of Shipment Allocation to Trucks.” Presented at the Transportation Research Board 92nd Annual Meeting, Washington, DC. Train, K. 2003. Discrete Choice Methods with Simulation. Cambridge University Press. TRB Pedestrian Committee Official Web Page, http://www.pedbikeinfo.org/trbped/?/trbped/ (As of Dec. 21, 2016). Vovsha, P., and M. Bradley. 2004. “Hybrid Discrete Choice Departure Time and Duration Model for Scheduling Travel Tours.” Transportation Research Record: Journal of the Transportation Research Board, No. 1894, pp. 46–56. Vovsha, P., and K.A. Chiao. 2008. “Development of New York Metropolitan Transportation Council Tour-based Model.” TRB Conference Proceedings 42: Innovations in Travel Demand Modeling, Vol. 2. Transportation Research Board, Washington, DC, pp. 21–23. Vovsha, P., Hicks, J.E., Paul, B.M., Livshits, V., Maneva, P., and K. Jeon. 2014. “New Features of Population Synthesis.” Presented at the Transportation Research Board 94th Annual Meeting, Washington, DC. Vovsha, P., and E. Petersen. 2009. “Model for Person and Household Mobility Attributes.” Transportation Research Record: Journal of the Transportation Research Board, Vol. 2132, pp. 95–105.

A-114 Method Selection for Travel Forecasting: User Guide Vovsha, P., Petersen, E., and R. Donnelly. 2003. “Explicit Modeling of Joint Travel by Household Members: Statistical Evidence and Applied Approach.” Transportation Research Record: Journal of the Transportation Research Board, No. 1831, pp. 1–10. Wikipedia. “Route Assignment.” https://en.wikipedia.org/wiki/Route_assignment (Accessed on Dec. 21, 2016). Williamson, P., Birkin, M., and P.H. Rees. 1998. “The Estimation of Population Microdata by Using Data from Small Area Statistics and Samples of Anonymised Records.” Environment and Planning A, Vol. 30, No. 5, pp. 785–816. Willumsen, L.G. 1978. “Estimation of O-D Matrix from Traffic Counts: A Review.” Working Paper 99, Institute for Transport Studies, University of Leeds. Wilson, A.G. 1967. “A Statistical Theory of Spatial Distribution Models.” Transportation Research, Vol. 1, pp. 253–269. Zenina, N., and A. Borisov. 2013. “Regression Analysis for Transport Trip Generation Evaluation.” Information Technology and Management Science, Vol. 16, No. 1, pp. 89–94.