ADVANCING TRANSPORTATION ANALYSIS WITH MOBILE DEVICE LOCATION DATA: INSIGHTS INTO BEHAVIOR, CONGESTION, AND ENVIRONMENTAL IMPACTS

Abstract

Emerging technologies, such as Mobile Device Location Data (MDLD), are reshaping the landscape of transportation analysis by offering continuous, large-scale observations of travel behavior. This dissertation explores the potential of MDLD in addressing critical challenges in transportation planning, focusing on data integration, travel behavior analysis, road congestion, and environmental impacts.The research begins by addressing the limitations of traditional travel surveys, which often suffer from small sample sizes and short observation periods. To overcome these challenges, this study integrates MDLD with regional travel survey data through probabilistic record linkage algorithms, creating comprehensive datasets that combine detailed demographic information with extensive, multi-day travel records. These fused datasets capture underreported travel behaviors and provide a more holistic representation of population mobility, laying a strong foundation for advanced transportation research. The second project validates MDLD’s ability to capture personal travel behaviors by examining the relationship between vehicle miles traveled (VMT) and gasoline prices in the Washington, DC, metro area. Using longitudinal MDLD datasets, we uncover reasonable price elasticities of VMT and demonstrate how aggregate VMT informs local gasoline demand. This analysis establishes the reliability of MDLD for studying travel behavior and energy consumption trends. The study then investigates changes in road congestion in the post-pandemic era, with a particular focus on the role of teleworking. Using MDLD-based measures of travel demand, trip distance, and congestion patterns in the Washington, D.C. region, the analysis reveals that while the frequency of commute trips has not fully recovered, average trip distances and non-work travel have increased, contributing to heightened congestion, especially during weekends. These findings highlight a fundamental shift in travel behavior rather than a simple rebound in traffic volume. Finally, the research explores the environmental implications of transportation using MDLD to estimate greenhouse gas emissions. By integrating MDLD with vehicle fleet characteristics, fuel economy, and contextual factors, the study employs engineering models and Monte Carlo simulations to quantify emissions. This work highlights the impacts of shifting travel behaviors and fuel prices in the post-pandemic era, offering actionable insights for emissions reduction strategies. Collectively, this dissertation showcases the transformative potential of MDLD in understanding mobility patterns and addressing energy and environmental challenges, providing a pathway for innovative transportation planning and policy development.