Comparing Regional Sustainability and Transportation Sustainability at the Metropolitan Level in the U.S. using Artificial Neural Network Clustering Techniques

Abstract

Regional sustainability and transportation sustainability have been intensely discussed and analyzed in recent decades. Though the use of indicators has been adopted in those models, debates continue on what indicators should be used and how to optimize the number of indicators. This results in the lack of a comprehensive and efficient method to assess and compare the sustainability of a sub-system, such as transportation system, and overall regional sustainability. A thorough literature review is conducted to identify indicators used to assess regional sustainability and transportation sustainability. Then, based on the available data, two sets of indicators for regional sustainability and transportation sustainability are identified and calculated respectively for the 382 metropolitan statistical areas (MSAs) in the U.S. A self-organizing map, which is a type of artificial neural network, is used to cluster the MSAs and compare their regional sustainability and transportation sustainability as well as to investigate the relationships among indicators. The results show that MSAs with a higher score on regional sustainability do not necessarily have a higher score on transportation sustainability. Some MSAs that are geographically close to each other have similar scores in regional sustainability and transportation sustainability. These findings provide insights to decision makers that the assessment of sustainability should consider both correlation and heterogeneity of different indicators within a region. Therefore, it is important to develop a comprehensive and efficient method to evaluate the role of sustainability in one urban sub-system, such as transportation, in the overall regional sustainability.