To what extent does an ABM improve the evaluation of transport policy beyond what can be done with a simpler 4-step model? This work offers new evidence on the differences between 4-step models and ABMs using a systematic approach.
The state of travel modeling practice is characterized by a co-existence of aggregate trip-based 4-step models and a newer generation of activity-based models (ABMs) which follow a disaggregate, microsimulated approach. While the conceptual advantages of ABMs are well known and generally accepted by the modeling community, they are frequently formulated in such terms as “behavioral realism” or “internal consistency” that are less tangible for a wider community of transportation planners. A legitimate question is to what extent an ABM affects transport project evaluations as compared with a simpler 4-step model. Direct comparisons are limited and rarely reported in a convincing way.
This research provides new evidence on the differences between 4-step models and ABMs using a recently developed travel demand modeling platform (AGENT for EMME) that allows for the flexible assembly and calibration of different types of travel models. In this work we assemble and calibrate both a 4-step and activity-based model which may be run with the same socio-economic, land-use, and network scenario inputs to facilitate a side-by-side comparison of the results. A key aspect of this work is that the model segmentation, such as trip purposes and modes, have been pruposefully aligned to the extent possible. However, in certain cases, for example temporal resolution, the gap between the models is unavoidable and should be recognized as a key distinguishing feature.
Both the 4-step and activity-based model were calibrated to the same data for a base scenario and then applied to study model sensitivity on illustrative planning scenarios including introduction of i) a new rail service with differential frequency by time of day, ii) road pricing differentiated by time of day, and iii) a mixed land-use scenario. For each test, the 4-step model response is compared to the ABM response with analysis and explanation of the differences. The authors believe that results from these sensitivity tests are relevant to inform planning expectations and current practice regarding systematic differences in the application of 4-step vs activity-based models.
|
