1.3 Principios estructurales del derecho ambiental
1.3.4 Desarrollo sustentable
Before getting into a more detailed methodological discussion and suggesting suitable improvement strategies, the current version (as of 2010) of the Maryland Statewide Transportation Model (MSTM) is summarized here. A more complete presentation of the MSTM can be found in the MSTM Users Guide, which is available from SHA upon request.
Geographically, the Statewide Model incorporates the entire states of Maryland and Delaware, District of Columbia, and portions of New Jersey, Pennsylvania, Virginia, and West Virginia. These areas comprise a total of 1607 zones. Zones outside Maryland are included in the model based on the work commutes going in and out of Maryland and are extracted from the 2000 Census Transportation Planning Package (CTPP). A Regional Model including the entire United States, Canada, and Mexico is divided into 189 zones, and includes information on long distance passenger travel.
Socioeconomic and demographic data from the 2000 Census survey and the 2003 Quarterly Census of Employment and Wages (ES-202) are used to create travel demand data. Data from the Census include population, occupation status, household incomes divided into quintiles, household sizes and the number of workers within those households, average household income, median household income, and the total number of workers.
The number of trips produced is estimated based on household information and trip generation patterns from recent household travel surveys conducted within the two MPOs (i.e. Washington D.C. and Baltimore). Long distance trips are micro-simulated based on the National Household Travel Survey in 2001. Several trip-purpose categories are considered in the MSTM and separately modeled: Home Based Work (HBW), Journey to Work (JTW), Journey at Work (JAW), School (SCH), Home Based Shop (HBS), and Home Based Other (HBO). Trip attractions are derived from regression analyses based on a household travel survey and socioeconomic data. For freight travel analysis, the statewide truck model is an adaptation of the Baltimore and Washington DC MPO truck and commercial vehicle models. The regional truck model generates long-distance truck trips based on given commodity flow data and the Freight Analysis Framework Version 2 (FAF2) developed by the Federal Highway Administration (FHWA).
MD-11-SP009B4S Project Final Report UMD Transportation Systems Research Lab 46 Figure 27. Flowchart of the Maryland Statewide Transportation Model
(Source: MSTM User Guide)
The first step of the MSTM’s sequential modeling modules, i.e. the person trip generation, generally follows a similar approach as the Baltimore Metropolitan Council (BMC) model, in terms of methodology and trip purpose identification. Trip production rates by household category have been taken directly from the BMC model and adjusted to the statewide income quintiles. Trip attractions are calculated based on regression-type equations in the BMC model. Adjustments to BMC model coefficients have been made to yield statewide trip production and attraction results. ES202, CTPP, and other socioeconomic data are used to customize trip attraction rates with alternative measures of area type.
Gravity model techniques are used for trip distribution. Trips with home base are organized in order of income categories. An exponential form is used for the impedance function in the gravity model. The impedance function is based on generalized travel time that considers various travel disutilities (time, cost, etc.) and composite travel time that considers travel costs on multiple modes.
MD-11-SP009B4S Project Final Report UMD Transportation Systems Research Lab 47 Figure 28. Structure of the MSTM Mode Choice Model
The mode choice step employs a nested logit model, as shown in Figure 28. In the driving nest, 2 passengers and 3+ passengers are specified separately in the ride sharing sub-nest. In the transit nest, transit riders with walking and driving access modes are considered separately. Transit options include regular bus, express bus, rail, and commuter rail.
Variables specified in the mode choice model include in-vehicle time, terminal time (time at the production and attraction zones), auto operating cost (the cost of driving per mile based on fuel prices and maintenance costs), auto toll cost, auto parking cost, transit walk time (the time to transfer by walking to or from a transit station), waiting time for mass transit to arrive or depart, transit fare, area type bias (suburban vs. central business district), and the drive-access time (the time it takes to drive to a transit or park-and-ride station).
The MSTM considers time-of-day with fixed trip percentage in each of the four time periods (AM peak, PM peak, mid-day and other off-peak) for each trip purpose. The temporal distributions of trips with different purposes were taken directly from the BMC model.
The aforementioned demand modeling steps deal with passenger trips with both ends in Maryland. Internal-external trips and external-internal trips are analyzed by the Person Long Distance Module (PeLoDiT) and the Visitor Travel Module (ViMo) respectively in MSTM. Each individual long-distance trip crossing the state boundary is micro-simulated with the trip origin-destination information derived from the 2001 National Household Travel Survey Long Distance Trip Sample. The flowchart in Figure 29 explains the procedure of the PeLoDiT simulation (ViMo has a similar structure.)
Person Drive Transit Share Ride Drive Alone 2 Passengers 3+ Passengers Walk Drive Bus Express Bus Rail Commuter Rail Bus Express Bus Rail Commuter Rail
MD-11-SP009B4S Project Final Report UMD Transportation Systems Research Lab 48 Figure 29. The Flowchart of the PeLoDiT Simulation
(Source: MSTM User Guide)
MSTM employs Multi-Class Assignment (MCA) to assign auto/truck and transit trips. This MCA is integrated in the CUBE Voyager package. In the first iteration, the shortest path is loaded with all trips between each OD pair (i.e. all-or-nothing assignment). During iterations, trips with the same OD are then averaged out between the current- and previous-iteration shortest paths until certain stopping criteria have been satisfied. During the assignment step, travelers are segmented into five different income groups and assigned onto the highway network system together with other four user classes (i.e. Commercial Vehicles, Regional Autos, Regional Trucks, and Medium/Heavy Statewide Trucks). The transit assignment adopts an “All-or-Nothing” approach, which assigns transit trips on the shortest paths between all OD pairs.
In its present form, MSTM can be employed for a number of transportation planning and policy applications. Model outputs can be utilized to supplement MPO travel analysis, conduct sub-area planning studies, forecast future-year auto/truck traffic volumes and transit ridership, and develop statewide and regional transportation plans.
MD-11-SP009B4S Project Final Report UMD Transportation Systems Research Lab 49