Skip to main content

Modeling Future Conditions

The modeling process, which considers the existing transportation, socio-economic, land use, and environmental conditions of the region, helps determines what planning actions are most supportive of the goals and objectives identified for the planning horizon.

Modeling

The combined input about the current and future transportation system conveyed strong public support for a set of overlapping ideas about the future of transportation: a more environmentally sustainable transportation system, additional pedestrian and bicycle infrastructure, shorter off-campus transit times, equitable access to transportation services, and a compact urban area that supports active transportation and limits sprawl development. Because of the complementary nature of these ideas, CUUATS staff, with the help of the LRTP 2045 Steering Committee, developed one preferred scenario for 2045 that incorporated these overarching ideas summarized by the LRTP 2045 goals, as well as emerging transportation technology such as electric and automated vehicles (AVs). The LRTP 2045 goals correlate with the pillars identified in the previous LRTP 2040, also based on public input, bolstering the foundation of the community’s vision for the future.

  • Bicycle and pedestrian infrastructure: 100 percent of recommended pedestrian and bicycle projects from local plans were incorporated into the preferred scenario for 2045. Although no local agency could feasibly commit to implementing more than 80 percent of all currently-recommended pedestrian and bicycle projects in their respective jurisdictions over the next 25 years, the small scale and large volume of these projects made it time-prohibitive to select which projects should be excluded.

  • Shorter off-campus transit times: MTD staff helped develop a hypothetical future transportation network that would shorten some off-campus transit times though route changes, frequency changes, and the incorporation off off-campus hubs.

  • A more environmentally-sustainable transportation system: Staff used increased rates of active transportation as well as industry projections for the increase of electric vehicles to reduce vehicle emissions in the region. In addition, staff incorporated growing rates of solar production potential to decrease the region’s future reliance on fossil fuels for residential, business, and electric car energy needs.

  • Limited sprawl: To discourage additional sprawl development and maintain short commute distances to services, restrictions were placed on peripheral development and incentives were assumed for infill development. In addition, the selection of projects modeled for the future emphasize existing transportation network maintenance over new roadway construction.

  • Automated Vehicles (AVs): Automated vehicles, fully autonomous, or self-driving vehicles are vehicles that can guide themselves without humans behind the wheel. CUUATS staff used industry projections as well as local research to estimate the timing and framework of future AV integration. As a smaller urban area, staff used conservative estimates for when AV technology might start to infiltrate the region. In conjunction with other transportation goals, staff also assumed that the vast majority of AVs would be both electric (in order to maintain a decrease in vehicle emissions) and shared (in order to discourage additional congestion and potential inequality in terms of access to new AV technology and infrastructure).

  • High Speed Rail: Being able to travel to Chicago via high speed rail in 45 minutes has been a goal of residents for many years, as documented in the LRTP 2045 public input as well as the two previous LRTPs for 2040 and 2035. The technology is not new, but the expense is well beyond the scope of local agencies and it’s hard to predict how and when it would be possible to obtain that kind of funding. Due to this uncertainty, it was determined by the LRTP Steering Committee and CUUATS staff that the preferred scenario would be modeled twice: once without high speed rail and once with high speed rail.

The three separate future scenarios modeled for the Metropolitan Planning Area (MPA) and compared with a 2015 baseline are briefly defined in the table below.

Scenario Name Scenario Description
2015 Baseline Based on 2015 data, intended to reflect current conditions.
2045 Business-as-Usual Scenario Forecasts how and where development will occur between 2015 and 2045 based on historic development trends, relatively certain future development projects and transportation system changes, as well as conservative integration of connected and autonomous vehicles starting around 2030.
2045 Preferred Scenario Designed to incorporate relatively certain future developments and transportation system changes as well as Federal, State, and local goals as summarized in the Goals, Future Conditions, and Future Projects sections. This scenario reflects ambitious implementation of bike and pedestrian recommendations in current plans, projected transit system changes, future environmental considerations and actions, and an emphasis on infill (over peripheral or sprawl) development.
2045 Preferred Scenario + HSR Incorporates the construction of a high speed rail (HSR) line to Chicago by the year 2040 into the 2045 Preferred Scenario.

The CUUATS modeling suite is designed to provide a holistic approach to planning analysis through the integration of localized transportation, land use, emission, social costs, accessibility, and mobility data. Each model addresses a specific area of concern at the necessary level of detail to make it appropriate for Champaign County or the metropolitan planning area. The synergy of the different models allows CUUATS planners to assess how different population changes and development patterns will impact the transportation system in the future. By quantifying the various impacts of potential transportation system changes, planners are able to compare different future development scenarios as well as develop individual performance measures. While all transportation improvements require some combination of labor, materials, and expertise to carry out, some desired improvements are more straightforward to measure and model than others. For instance, completing gaps in the bicycle network or improving curb ramps to improve multimodal connectivity can be counted, measured, and tracked as improvements are completed over time. Other desired improvements, like limiting sprawl development or reducing greenhouse gas emissions to improve environmental health, rely on a number of additional factors and future unknowns that are not as clearly measurable and are therefore more difficult to model for the future.

Two models serve as the foundation of the CUUATS modeling suite: a land use model and a travel demand model (TDM). The land use model projects population, employment, and land use change into the future while the TDM estimates the number and location of auto and transit in the future. The TDM is integrated with the land use model, running in a 5-year iterative process over the 25-year planning period, to identify the relationships between land use changes and travel patterns in the region. Three additional models, SCALDS, MOVES, and Access Score, use the outputs from the first two models to project different costs of development, vehicle emissions, and transportation network accessibility for all modes. Models can’t predict the future, but they can help us imagine the future and try to understand how our actions today could impact our transportation system down the road. Countless variables will determine the health and relative success of the region over the next 25 years. While the LRTP 2045 models and projections were carefully designed and validated whenever possible, they are not perfect reflections of the social and physical processes in play, and the input data used is imperfect.

Overall, the models’ results show that the increased density of the two 2045 Preferred Scenarios reduce vehicle miles traveled, new infrastructure costs, resource usage, and per capita emissions compared with the 2045 Business-as-Usual Scenario. The inclusion of high speed rail increases commuters between Champaign County and Chicago and produces additional but modest increases in population, employment, and density in the metropolitan planning area over the 2045 Preferred Scenario without high speed rail. For more information about the different models and related inputs utilized in the development of the LRTP 2045 scenarios, see the Data and Models section.

CUUATS LRTP 2045 Modeling Suite

Diagram showing the types of software used to develop the future vision. CUUATS staff started with the Travel Demand Model and Urban Sim. These two programs analyze the vehicle miles traveled, mode choice, traffic volume by road link, congestion speed, population projections, employment projections, and areas of future growth. From there, the data collect from the Travel Demand Model and Urban Sim are used in three other programs: Social Cost of Alternative Land Development Scenarios (SCALDS), Motor Vehicle Emissions Simulator (MOVES), and Neighborhood Level Accessibility Analysis (Access Score). SCALDS measures transportation cost by mode, energy cost, infrastructure cost, and water and sewer cost. MOVES calculates GHG emissions, urban/rural emissions, and PM 25 and other emissions. Access Score measures the level of traffic stress, by transportation mode, accessibility score by transportation mode, and the accessibility scores by destination type.
Diagram of the modeling process used by CUUATS staff to develop the Long-Range Transportation Plan 2045 vision. Note: To expand the image to full-size, right-click the graphic and choose 'open image in new tab'. Image: CUUATS

UrbanCanvas - Land Use Modeling

UrbanSim’s UrbanCanvas modeler is a microsimulation land use model designed to support the regional transportation planning process by forecasting future scenarios based on the interactions of individual persons and households. UrbanCanvas consists of a series of interconnected models that allow planning agencies to analyze anticipated and possible future changes in land use and development patterns, regulations, and growth rates, over a designated planning horizon. The outputs forecasted by the model can vary based on the wide array of inputs.

CUUATS employed a Census block-level cloud platform of the model to analyze each of the future scenarios considered from the base year, 2015, to the horizon year, 2045. To create these predictions, UrbanCanvas relies on a set of core tables, including the Census block designated boundaries, residential units, households, persons, and jobs. In the cloud platform, the households and persons tables are generated by UrbanSim, using a population synthesizer, which uses block-group level American Community Survey (ACS) estimates to distribute households and individual people to the blocks. As part of defining the 2015 base year for the model, CUUATS opted to replace UrbanSim’s standard jobs data from Census Longitudinal Employer-Household Dynamics (LEHD), with region specific data from Database USA, accessed through EMSI. This data was further processed and refined based on research and local knowledge. CUUATS staff also used an iterative proportional fitting (IPF) process to distribute ACS tract-level housing unit data to the block-level to create the 2015 residential units table.

In addition to the core tables, the model requires user-uploaded datasets, including household and employment control tables, which establish the total regional households and employment by North American Industry Classification System (NAICS) defined sectors. The user uploads also include zoning constraints, additional development constraints, known future developments, and travel model skims, which provide measures of the ability to travel between each traffic analysis zone (TAZ).

UrbanCanvas was used to analyze and predict population, households, employment by industry, residential units, and areas of future growth for each of the scenarios. Population and employment projections play an especially critical role in visualizing future regional characteristics.

Based on the process used to create the population and employment projections, the 2045 Business-as-Usual Scenario and the 2045 Preferred Scenario have the same projected population and employment. This is because the same control tables were used to generate these values, as it was assumed that no other variables controlled by the model would indicate increases in these values. Greater population and employment growth are projected for the 2045 Preferred Scenario + High Speed Rail. This is because a network benefits study from the Midwest High Speed Rail Association indicated that in simulation years 2042 to 2045, following the completion of the rail line, new households and jobs would be attracted to the area. Staff reflected these values through a manual edit to the household and employment controls tables. These increases can be seen in the population and employment projections table above.

While the overall population and employment values are projected to be the same for the 2045 Business-as-Usual Scenario and the 2045 Preferred Scenario, the distribution of growth is different in these two potential futures. While Business-as-Usual represents on-going development at the edges of the municipal boundaries in the county, the two Preferred Scenarios include additional development controls that result in increased redevelopment within current municipal boundaries post-2020, when most known development projects will be complete. The following two maps highlight Census blocks that are projected to see 20 percent or more population and/or employment growth by 2045.

Map highlights Census blocks projected to see 20 percent or more population and/or employment growth by 2045 according to the Business-as-Usual scenario for 2045.
Note: To expand the image to full-size, right-click the graphic and choose 'open image in new tab'. Image: CUUATS
Map highlights Census blocks projected to see 20 percent or more population and/or employment growth by 2045 according to the Preferred scenario for 2045 and the Preferred scenario with high speed rail for 2045.
Note: To expand the image to full-size, right-click the graphic and choose 'open image in new tab'. Image: CUUATS

Through the outputs it produces, UrbanCanvas is linked with the other tools in the CUUATS modeling suite. It has the strongest connection with the TDM. By allowing user defined geographic boundaries to be uploaded to the cloud platform, UrbanCanvas is able to distribute the population, household, and employment projections it creates to each of the TAZs used in the TDM. Based on local knowledge, staff made some adjustments to these values to better reflect what currently exists in the county, as well as certain known futures. The distributed population, employment, and household values create the socioeconomic inputs needed to run the TDM. The UrbanCanvas model and TDM then run in an iterative process that allows any congestion or other transportation patterns that may impact regional development in the next 25 years to be considered in UrbanCanvas’s prediction of future development. Outputs from UrbanCanvas are also used in the Social Costs of Alternative Land Development Scenarios (SCALDS) model.

Travel Demand Model (TDM)

The Champaign County travel demand model (TDM) is a person trip model built using the Cube Voyager software platform. The TDM employs the four-step travel forecasting process to evaluate auto and transit trips for both daily and peak hour scenarios. The countywide travel demand model is integrated with UrbanSim’s UrbanCanvas modeler to identify the relationships between land use changes and travel patterns in the region. Due to the unique relationship between the TDM and UrbanCanvas, multiple horizon year forecasts can be evaluated. The base year for the model is 2015.

To forecast trips and roadway volumes for each of the future scenarios, the TDM relies on a set of core inputs, including population and employment projections generated from UrbanCanvas, future roadway network and transit service projections identified with state and local roadway agencies, and other variables including overall vehicle stock and fuel economy. The CUUATS TDM also considers the advent of connected and/or autonomous vehicles (C/AV) and their impact on regional travel patterns and transportation infrastructure, detailed in the Data and Models section.

The TDM models trips for all purposes, including work, school, shopping, and other trips. The following two figures report the TDM’s vehicle miles traveled (VMT) and mode share projections for the 2015 Baseline and the three 2045 scenarios.

The total VMT is projected to increase approximately 39 percent from between 2015 and 2045 under the Business-As-Usual Scenario. This is mainly due to the projected increase in population and employment, as well as the projected additional induced-trips from connected/autonomous vehicles. The total VMT under the Preferred Scenario is projected to be less than that of the Business-As-Usual Scenario as a result of an increased share of transit, bicycling, and walking trips due to improved and expanded infrastructure for those modes. Under the Preferred Scenario + HSR, the higher population and employment in the MPA results in a higher VMT, but still lower than that of the Business-As-Usual Scenario, as approximately 204,750 annual car trips between the Champaign-Urbana region and Chicago are projected to be replaced with high speed rail trips.

The CUUATS TDM was utilized to measure levels of congestion during peak hours within the urbanized area roadway network. Levels of congestion were determined based on Volume to Capacity (V/C) ratio values of different roadway segments during the peak hours. A Volume to Capacity ratio compares demand (roadway vehicle volumes) with supply (roadway carrying capacity). The following maps show modeled congested links in baseline year 2015 and projected congested links in 2045 where the volume of traffic on the roadway threatens to exceed (a value of 0.9-1) or exceeds (a value greater than 1) the capacity of the roadway during peak travel times in the MPA.

To view traffic congestion for different scenarios, use the menu button to expand the map options.

Social Cost of Alternative Land Development Scenarios (SCALDS)

The Social Cost of Alternative Land Development Scenarios (SCALDS) model was created in 1998 by Parsons Brinckerhoff and ECONorthwest for the Federal Highway Administration. The model tests the impacts of various scenarios of land development that are often unforeseen by policymakers. It is a comprehensive model of the initial costs of development (such as building out utilities), ongoing maintenance costs, and externalities like travel time and natural resource use. CUUATS staff have updated and reviewed the model using local data for Champaign County.

When comparing the Business-as-Usual with both preferred scenarios, the key differences in land development are the location and density of new growth. While Business-as-Usual assumes ongoing development at the peripheries of municipal areas in the region, the preferred scenarios implemented building constraints, limiting new development to the existing municipal boundaries. These growth patterns can be seen in the future growth area maps shown above. This shift in development patterns maximizes the utility of existing infrastructure and helps to preserve land for important agricultural uses.

Limiting the area for new development not only changes the location of future growth, but also the type of residential units constructed. While Business-as-Usual predicts an increase of 2,021 single family homes, the preferred scenarios show much larger increases in multi-family housing units. Due to the assumed increase in population, the preferred scenario with high speed rail still predicts the construction of new single-family homes, but only about 10% of what is expected from Business-as-Usual. The preferred scenario without rail predicts a negative value for the construction of single-family homes, which represents the demolition and redevelopment of low-density residential parcels to higher density multi-family uses. The following two figures show these trends in residential development.

Shifting a large portion of the regional growth to the core of the community and prioritizing urban redevelopment over new peripheral development, decreases the need to extend infrastructure to meet the needs of new development far away from already developed areas. Both preferred scenarios predict significantly lower new infrastructure costs between 2015 and 2045, at about 30% lower than that predicted by Business-as-Usual. These cost estimates can be seen in the following chart. The infrastructure cost estimate for the Preferred Scenario + HSR is higher than the Preferred Scenario without HSR due to the need to provide for the additional population and employment projected with the installation of high speed rail.

While both preferred scenarios predict a lower demand for new infrastructure to be built, this does not necessarily represent reduced maintenance costs. All three scenarios predict very similar per capita annual operating costs in 2045, with the Preferred Scenario + HSR predicted to be the highest. This estimate includes maintenance of transportation, sanitary and storm sewer, and all energy infrastructure. While there may be less new infrastructure constructed in the preferred scenarios, an increase in biking and walking was assumed, which would increase demand and maintenance for new and existing infrastructure associated with active transportation. The larger population and employment assumed with high speed rail would also increase the use and therefore the maintenance of all infrastructure represented in this model.

In addition to reducing the cost of new infrastructure, increased density of development in both of the preferred scenarios also results in lower annual transportation and non-transportation energy consumption per capita (see figure below). Increasing density reduces the energy consumed by buildings, as well as transportation energy needs. Annual transportation and non-transportation energy consumption per capita is also lower in both of the preferred scenarios because it was assumed that solar production would continue to increase at the present rate of adoption, so the estimated solar production was removed from the predicted consumption. The preferred scenarios also anticipated a greater adoption of electric vehicles, but the increase of solar production offsets this increased demand for electricity.

Mobile Source Emissions (MOVES)

The MOtor Vehicle Emission Simulator (MOVES) model was developed by the Environmental Protection Agency’s (EPA) Office of Transportation and Air Quality. MOVES is required for use by states and metropolitan planning organizations (MPO) where measurements of one or more pollutants exceed the maximum allowable levels under the National Ambient Air Quality Standards (NAAQS) and must be run at the county scale for all non-attainment areas. Although the Champaign-Urbana urbanized area is currently an attainment area for all emissions quality standards, CUUATS staff proactively includes MOVES in the modeling suite to estimate the environmental impact of alternative planning scenarios. This data also allows the region to continually track and better understand how ongoing development affects emissions in order to remain an attainment area. Several calculated assumptions impact the MOVES 2045 outputs including increased temperature, a significant increase in the share of electric vehicles, and transportation network recommendations outlined in the Data and Models section.

The figure below shows the modeled annual mobile vehicle emissions in the MPA. Since a 39 percent electric vehicle fleet share is projected for the 2045 Business-as-Usual scenario, the percentage increase in emissions from the 2015 baseline is less than the percentage increase in VMT shown from the TDM. The projected increase in active transportation mode share and an even higher share of electric vehicles under both 2045 Preferred Scenarios result in a decrease in the amount of emissions generated compared with 2015, despite projected population and employment gains.

Access Score

Many of the long-range transportation planning assessment processes are concerned with data and trends that occur at the regional level. While this is beneficial for understanding the overall future direction of the community, it is not localized enough to help identify specific limitations in the transportation network. To help address this spatial mismatch and make the CUUATS transportation planning and modeling processes more complete, staff developed a geography neutral, multimodal accessibility assessment, known as Access Score. This tool utilizes level of traffic stress (LTS) assessments for each mode and travel time to calculated accessibility scores to several destination types. These accessibility scores help staff to assess the current and potential future status of accessibility in the Champaign-Urbana region, to identify areas in need of improvement, and to observe potential benefits from the construction of new infrastructure.

To develop Access Score staff utilized an existing bicycle level of traffic stress (BLTS) assessment from the Mineta Transportation Institute, and an existing pedestrian level of traffic stress (PLTS) assessment from the Oregon Department of Transportation. Automobile level of traffic stress (ALTS) is assessed using an in-house analysis, created by CUUATS staff to emulate the assessments for BLTS and PLTS, by considering elements of the automobile transportation network and its interactions with other modes. In each of these LTS assessments, network infrastructure characteristics were assessed and assigned a level of stress, one (1) being the lowest stress and four (4) being the highest. Each segment in the network was assigned an overall level of stress based on the highest, or most stressful score it received for any one of the characteristics considered. Transit level of traffic stress (TLTS) is assessed using the Pandana accessibility tool, which uses general transit feed specification (GTFS) and transit headway and schedule data to assess transit trips based on the time required to reach a destination, which is then combined with the pedestrian score required to get from the point of origin to the nearest bus stop.

Once the modal LTS scores were complete, accessibility was calculated by multiplying the LTS scores by the travel time. The assessment includes accessibility to the following ten destination types: grocery stores, health facilities, jobs, parks, public facilities, retail stores, restaurants, schools, arts and entertainment, and services.

Access Score allows not only for the assessment of current accessibility, but also for the analysis of future scenarios. To evaluate future accessibility for both preferred scenarios, staff geocoded all bicycle, pedestrian, and automobile projects recommended in current plans that would impact infrastructure considered in one or more of the LTS scores. Other LTS characteristics were also updated using the relevant TDM scenario outputs. Due to the significant overlaps in the elements of the two preferred scenarios, a comparison of the two different assessments showed no discernable differences. Overall, the addition of new infrastructure from 2015 to 2045 had positive impact on bicycle and pedestrian access scores, with slight decreases to automobile scores. Transit accessibility increased slightly, despite no change in the routes and schedule in the analysis. This increase can be attributed to improved scores for the pedestrian portions of those trips. The scores for the 2015 and 2045 scenarios can be seen in the Access Score application embedded below.