Home
This wiki is intended to serve as a user guide to describe the overall structure of the modeling system, how to set up and run the model system, and the model system inputs and outputs. Additionally, this wiki discusses the reporting system developed for the model.
The SANDAG resident travel model is based on the CT-RAMP (Coordinated Travel Regional Activity-Based Modeling Platform) family of activity-based models. The model has been developed to ensure that the regional transportation planning process can rely on forecasting tools that will be adequate for new socioeconomic environments and emerging planning challenges. It is also equally suitable for conventional highway projects, transit projects, and various policy studies such as highway pricing and HOV analysis.
In addition to the CT-RAMP resident travel model, a number of other model components have been developed and integrated into an overall modeling system. These other model components include:
- A heavy truck model covering heavy (8,500 pounds or more) trucks into, out of, and through San Diego
- An interim commercial vehicle model covering other goods and services movements within San Diego
- An internal-external travel model covering travel into and out of San Diego made by San Diego residents
- An external-internal travel model covering non-resident travel into and out of San Diego made by non-Mexican residents
- An external-external travel model covering travel through the San Diego region.
- A Mexican resident (cross border) travel model covering travel into, out of, and within San Diego County made by Mexican residents
- An airport model covering trips made to and from San Diego airport
- A visitor model covering trips made within San Diego County by overnight visitors
- A special event model, covering trips made to and from special events
Some of these models are also implemented in Java using the CMF, while other, aggregate model components are implemented in Emme, along with network skimming and assignment procedures. Furthermore, the model developed for SANDAG relies on the Production, Exchange, and Consumption (PECAS) land-use model for land use, employment, and population inputs. The figure below shows the overall model system.
Steps in green are implemented in Emme using Emme Modeller tools implemented as Python scripts. These include the initial construction of transport networks from files created by the SANDAG Geographic Information System (GIS), assignment of trip tables to networks, and network skimming. In addition, aggregate special market models, such as the commercial vehicle and heavy truck model, the external-internal travel model, and the external-external travel model, are implemented in Emme. Emme is also used to construct trip tables from all model components and control the overall model flow. Java is used to implement the main CT-RAMP resident travel model, as well as all disaggregate simulation-based special market models, such as the internal-external travel model, the visitor travel model, the airport model, and the Mexican resident travel model. Outputs from each model are loaded into a database and the SQL scripts are used to report results from the simulation. EMFAC input files are also created from the simulation using a Python procedure so that air quality from mobile sources can be measured.
There are four other documents that complement the User’s Guide, which can be referred to for further information:
- A model specification report, which describes the structure of the resident and special market models (with the exception of the heavy truck model), the mathematical form of each model component, and the variables considered;
- The SANDAG Travel Model in Emme User Guide describing the model as implemented using the Emme modeling framework and components (assignment and demand modeling tools). Integrated with disaggregate demand model components (CT-RAMP family of AB models). The report also provides “how-to” instructions for master run model operation and the operation of the components implemented within the Emme framework;
- A model estimation report, which describes the estimation results for each model component, with more details on the formulation of each model and the interpretation of the parameter values; and,
- The Activity-Based Model and Commercial Vehicle Model Validation Report, which describes the goodness-of-fit of each model component to observed data for the base year (2008) and the process of adjusting parameters to better match observed data. Since then, the base year has been changed to 2010, to 2012, and to 2016 to reflect changes in 2010 Census and San Diego Forward, the new SANDAG RTP. Additional calibration and validation reports for base year 2010,2012, and 2016 are also available as additions to the 2008 calibration and validation report.