Statewide Travel Demand Model
A statewide model was developed to provide projected 2035 traffic volumes, miles of travel and
hours of travel data, as well as to identify anticipated locations of congestion and need for
improvements. The model reflects base year 2005 and future year 2035 growth, population and
employment based on readily available data from sources including the U.S. Census, Census
Transportation Planning Package (CTPP) 2000, University of Alabama (UA) data center, regional
RPOs and MPOs, state agencies, and private industry/business organizations. The model platform
is Tranplan with Cube interface.
The model process is initiated with assessment of current conditions. To begin, TAZ boundaries
are defined through review of Census data, MPO models, and previous ALDOT evaluations of
transportation conditions. The number of TAZs was limited to between three and ten per county
in rural areas. In MPO study areas, the model retained the structure and geography of the MPO
regional model, but aggregated MPO zonal geography into some TAZ as appropriate. The model
consists of a total of 1,081 TAZ, including 617 in MPO areas and 464 in rural areas.
The base year SE (socioeconomic) data is then developed and distributed by TAZ. The SE data
development methodology and data was provided to all RPOs for comment and then revised as
necessary. MPO SE data from the MPO models was aggregated for new TAZ as appropriate. Annual
growth rates from 2000 to 2030 were applied to project 2005 and 2035 SE data.
The source of population figures was the U.S. Census 2005 population estimates by county and
the year 2000 population figures by census tract. As 2005 population estimates were not
available by census tract, 2005 county level estimates were manually distributed proportionately
based on 2000 census tract figures to most accurately derive 2005 TAZ population figures. The 2035
future year population for each TAZ was derived by applying countywide growth rates based on annual
growth projections provided by UA. The adjustments were then distributed proportionately to each
TAZ assuming similar proportions as year 2005.
Year 2000 employment by census tract was derived by industrial category from the CTPP 2000 and
was deemed the best available source. Unfortunately, 2005 census tract data was not produced.
To develop credible estimates of 2005 employment by census tract/TAZ, 2005 employment by county
from UA was proportioned to the TAZ level based on 2000 data. To derive 2035 employment figures,
an average annual growth rate was applied to each of the 2005 base year TAZ figures, resulting in
2035 employment figures by TAZ. The projected annual growth rate from 2004 to 2014 by workplace
investment areas (WIA) from the Alabama Department of Industrial Relations was used to project
2035 employment.
The base year network was developed based on updated roadway characteristics reflecting
completed improvements, changes to functional classification, and the addition of any key
non-State system facilities determined to be key due to their critical function. Development
of base year network utilized ALDOT roadway functional classifications and GIS data. The model
network includes 11,988 miles of roadway, including 1,043 miles of interstate/freeway,
8,579 miles of arterials, and 2,250 miles of collectors. The network also includes 1,081
intra zones (TAZ within Alabama’s state boundary), as well as 77 external stations that connect
with Florida, Georgia, Mississippi, and Tennessee. The network includes interstates, state
roads, and any key facilities which are not on the State system. All interchanges were coded
into the network.
The trip generation module reflects current trip making activity based on a review of historic
and current trip patterns on State system roads, as well as a relationship between trips
generated and socioeconomic variables (population, households, employment, household median
income, and automobile ownership). The tool, created using Excel and based on the ALDOT trip
generation methodology, is easy to use and its parameters easy to modify if needed. The trip
generation module identifies special generators statewide with unique characteristics depending
on the type and location, such as seaports and other economic hubs.
Following development of the trip generation module, the trip distribution and traffic
assignment modules can be developed. The statewide model utilizes Cube catalog and applications
with Tranplan to perform trip distribution and traffic assignment. The loaded network includes
traffic volumes, volume to capacity (v/c) ratios, and other measures. By reflecting the number
of trips between TAZs and trips assigned to the network, the relationship between variables
(socioeconomic development, land use, roadway capacity, and speed) affecting travel patterns
can be measured. By running the base year model and reviewing current operating conditions of
all trips assigned to the network, an estimate of the total number of vehicles using each
roadway in the network is developed. When compared to the roadway capacity, the model identifies
congested areas and shows the level of congestion projected to occur. This in turn becomes the
basis for assessing the performance of the transportation system and documenting MOCs.
The base year model is calibrated and validated to FHWA specifications provided in the Model
Calibration and Reasonableness Checking Manual. Calibration and validation of the base year
model ensures it accurately reflects current travel patterns and traffic volumes, thereby
providing confidence in the model and its ability to reflect current as well as future
conditions.
After approval of the base year model, the future year model is developed to project future
conditions and determine the transportation network’s future MOC performance. Initial efforts
include review of RPO and MPO future year SE data and distribution to TAZs. Future year SE data
distribution in the model maintains a relationship to the base year SE data distribution as
well as consistency with the factors used in the base year trip generation module. The next
step entails applying any E+C (Existing and Committed) projects to the future year network.
The E+C network reflects current transportation program activities and includes any projects
that are at an advanced stage of programming – the project is either under construction or
right-of-way has been purchased prior to construction. The future year model is then run using
forecasted SE data and the E+C network. The results identify the sufficiency of the E+C network
to address future demands on the transportation network.
To analyze proposed improvements and document the anticipated benefits, the combinations of
projects selected for alternative SWTP Update scenarios are programmed into the future (E+C)
travel demand model to produce the “build” model. The MOC benefits documented include criteria
such as miles of travel, hours of travel, travel speed, ratio of volume to capacity, and other
metrics available from the model.
As a final effort, the model application report is prepared for inclusion as part of the SWTP
Update documentation. The report is comprehensive, showing the process and methodologies followed
in development of the model, the calibration and validation of the base year model and the
assumptions made.