Measuring density and diversity to model travel behavior in Indian context

Abstract

We propose Total Activity Density (TAD) and Weighted Index of Dissimilarity (WID) as indicators of density and diversity against the conventional indicators of population density and non-residential to residential ratio, respectively. The proposed indicators account for complementary land uses i.e. residential and non-residential activities that serve as origins and destinations in the home-based trip, respectively.Both the proposed and conventional indicators are measured for the city of Vishakhapatnam at NMT-TAZ level. We have validated the application of the proposed indicators by estimating linear regression models for trip length using both the proposed and conventional indicators. The analysis shows that the trip length models with TAD and WID resulted in modest or negligible improvement over the model with conventional indicators. However, the models estimated using the proposed indicators have resulted in a significant reduction in the sum of squared errors as compared to the models estimated using the conventional indicators. We, therefore, interpret that the proposed indicators of density and diversity provide a better understanding of travel behavior in the Indian context. Our study shows the inverse and significant impact of TAD and WID on trip lengths. Therefore, more dense and diverse zones are likely to result in short distance trips and encourage non-motorized modes of transport.