3.6 Daysim Inputs
The following sections provide a brief overview of the inputs to the DaySim components of the model system.
The NERPM activity-based model system uses microzones as the fundamental spatial unit for generating travel demand. Use of microzones improves the sensitivity of the model system to land use, fine-grained urban form and accessibility attributes. The microzone data input file contains fields that describe the quantities of households, school enrollment by type and employment by industrial sector within quarter mile and half-mile buffers. Note that these buffers are based on “all streets” based network accessibilities and employ decay functions that weight closer opportunities more than distant opportunities.
The synthetic population is comprised of lists of households and persons that are based on observed or forecasted distributions of socioeconomic attributes and are typically created by sampling detailed Census microdata. These lists function as the basis for all subsequent choice-making simulated in the activity-based model. The PopulationSim synthetic population tool was used to generate the synthetic population input to the NERPM-AB Daysim activity-based model system. Details pertaining to the design of synthetic population were described in the previous section.
Worker IXXI Fractions
Although the modeling area is defined in such a way as to capture as much “internal” travel by regional residents as possible (that is, travel with both origins and destinations with the modeling area), a certain portion of observed regional travel involves either regional residents travelling to destinations outside the modeling area or people who are not regional residents travelling to destinations within the modeling area. As in a traditional trip-based travel demand model system, these travel markets are typically incorporated into the model through the use of internal-external trip tables, which may be either fixed or dynamic.
A distinguishing feature of the DaySim activity-based model system is that, due to the spatial and behavioral detail embedded in the model, it is sensitive to how this internal-external travel affects the choices made by regional residents. A particular focus of this detail is on ensuring that the right numbers of workers are “out-commuting” to employment locations outside the modeling area, and that the right number of regional jobs are being consumed by non-residents “in-commuting” to the region. At present, this is accomplished by using a file (worker IXXI fractions) that contains TAZ-based shares of workers who are in-commuting and out-commuting, which is provided as an external input to the DaySim model system. The shares either can be held fixed, or may be updated by deriving updates shares from the trip-based model outputs.
The “out-commuting’ proportion of workers in each TAZ are assumed to go to work locations outside the model region. Since ABM involves simulating choices of each individual in the synthetic population, the workers that “out-commute” are determined as a probabilistic choice based on the fraction specified in the IXXI file. DaySim does not simulate the day patterns of such workers. Similarly, the “in-commuting” proportion results in a portion of job in each TAZ not being available for workers in the model region. These are assumed to be taken up workers from outside the model region.
The TAZ index file enables users to flexibly define non-continuous zones numbering systems, and to identify the availability of external and other zones as destination choices, without impacting DaySim performance.
The PNR file provides park-and-ride (PNR) and kiss-and-ride (KNR) locations with corresponding capacity and parking cost. The locations of these nodes are used by DaySim inform the choice of PNR station for a PNR transit trip.
A coefficient file provides a list of variables used in the model and corresponding coefficient values and t-statistics. Each Daysim model component is associated with a coefficient file. For the NERPM-AB model, the model coefficients were borrowed from SACOG model and later calibrated to match Jacksonville survey data.
A key set of inputs to any travel demand forecasting model system are the files that contain the scenario, mode, user-class, and time period-specific measures of network impedance, often referred to as network “skims.” The roster provides users with the ability to flexibly specify the skims that are associated with the different mode, time period and user classes used in the NERPM activity-based models system, without necessitating changes to the core DaySim model code. For example, a user may want to increase the number of time periods used in the model system to better reflect changes in network impedance by detailed time-of-day. In order to implement such an enhancement, a user would first revise the Cube-based network-processing scripts in order to generate the desired skims and would only need to revise the DaySim impedance roster to make DaySim sensitive to this additional detail.
The "Roster Combinations" file gives the possible mode/path type combinations used in DaySim. The file has columns that enumerate the 9 modes used in the current model system (walk, bike, SOV, HOV2, HOV3, transit, park-and-ride, school-bus, other) and 8 rows that enumerate the path types currently used (full-network, no-tolls, bus, project mode, pnr-bus, pnr-project, knr-bus, and knr-project). The path type with “project” in it is used to new modes that may be introduced in alternate scenarios. The cells are TRUE for valid combinations within DaySim and FALSE otherwise.