Using a spatial autoregressive model with spatial autoregressive disturbances to investigate origin-destination trip...

Area of Science:

• Regional Science
• Spatial Analysis
• Transportation Geography

Background:

• Spatial interaction models with origin-destination (OD) filters are crucial for analyzing trip flows.
• Existing models often oversimplify spatial dependence, primarily using autoregressive processes.
• The full extent of spatial effects in trip distribution remains incompletely understood.

Purpose of the Study:

• To investigate the presence and extent of spatial dependence in both autoregressive and error terms within spatial interaction models.
• To introduce and evaluate a spatial autoregressive model with spatial autoregressive disturbances (SARAR) model incorporating OD filters.
• To determine if SARAR models offer superior statistical performance and more insightful marginal effects compared to traditional SAR and SEM models for trip distribution analysis.

Main Methods:

• Specification and estimation of SARAR models with OD filters.
• Comparative analysis of SARAR models against spatial autoregressive (SAR) and spatial error models (SEM).
• Application of models to empirical trip distribution data from Hangzhou, China.

Main Results:

• The SARAR model with OD filters successfully disentangles the location and magnitude of spatial dependence in trip flows.
• Statistical comparisons indicate that SARAR models can outperform SAR and SEM models in specific contexts.
• Marginal effects derived from the SARAR model provide a more nuanced understanding of trip distribution drivers.

Conclusions:

• The developed SARAR model with OD filters represents a novel approach to trip distribution analysis.
• This methodology offers a more comprehensive characterization of spatial dependencies in trip flows.
• The findings support the utility of SARAR models for regional science and transportation studies.