The PECAS Model of Atlanta

Dr. John E. Abraham

March 7, 2013

Where:  Instructional Center, Room 109

Video

Abstract: The PECAS Model of Atlanta is a spatial economic forecasting model for the 20 county region.  It forecasts development patterns, home and business locations, economic interactions and economic performance into the future.  It is policy sensitive to transportation infrastructure, transportation services, land use regulations, taxes and environmental policy.  The developer model micro simulates individual construction events on 1.9 million parcels records representing the land in the region using a joint discrete-continuous logit formulation.  The spatial economic module simulates the location choices of businesses and households within the developed land, and their interactions that lead to travel on the road network, using an additive logit formulation.  The simulation is founded on a synergy between spatial economic theory and random utility theory.  The integrated modeling system includes both of these PECAS components, as well as the Atlanta Regional Commission's travel demand model and a spatial economic forecasting model.

The presentation will cover the theory behind the PECAS modeling framework, the development of the PECAS model for Atlanta, the data used in the Atlanta region, the software implementation of the model, and the application of the model for the regional transportation plan.  Examples from other PECAS models (such as San Diego, Baltimore, Sacramento, Oregon etc.) will provide additional context.

Bio: Dr. John E. Abraham is the principal software developer of the PECAS system for integrated land use transportation modeling, and co-author (with Dr. John Douglas Hunt) of the theoretical formulation.  Dr. Abraham has 20 years of experience in developing and calibrating spatial economic and travel demand models. Dr. Abraham has expertise in developing and calibrating models to provide computer simulations that are both accurate and practical for analyzing policy and scenarios. He is an expert on survey techniques for understanding preferences, measuring trade off rates, and predicting behavior. Dr. Abraham can quickly program advanced modeling techniques into software due to his strong mathematical and statistical background and his knowledge of algorithms and data structures. He is an expert software architect with a focus on object-oriented design. John is a principal at HBA Specto Incorporated in Calgary, AB, Canada.