published journal article
working paper
The Alameda corridor provides a crucial rail link for moving freight in and out of the Ports of Los Angeles and Long Beach, also known as the San Pedro Bay Ports (SPBP). While the benefits of this trade are enjoyed by the whole nation, the associated air pollution costs are born mostly by the people who live in the vicinity of the Alameda corridor and the two freeways (the I-710 and the I-110) that serve the Ports. Although they are more energy efficient than trucks, trains contribute heavily to regional air pollution; in addition, rail traffic in the South Coast Air Basin is projected to almost double in the next twenty years. This paper presents an analysis of the emissions and the dispersion of PM and NOx emitted by train operations in and around the Alameda corridor. We find spatial and temporal variations in the dispersion of these pollutants, which justifies our approach. Moreover, the railyards in our study area are responsible for the bulk of PM and NOx emissions (compared to line haul operations). While PM emissions from train operations contribute only a fraction of the recommended maximum concentration, NOx emissions go over recommended guidelines in different areas. The affected population is mostly Latino or African American. Our approach is also useful for better understanding trade-offs between truck and rail freight transport.
other
presentation |
Mar 2023
conference paper |
Jan 2023
published journal article |
Aug 2022
conference paper
research report
conference paper
Despite the growing presence of on-demand, app-based ride services at Los Angeles International Airport (LAX), not much is known about their characteristics and usage. This paper estimates and compares travel times and costs of rides by Uber and Lyft (also known as shared rides) and other forms of ground transportation to LAX. Using the estimated travel times and costs and the 2015 Los Angeles International Airport Passenger Survey, it models the airport access mode choice decisions to analyze how different circumstances including an increase in shared ride costs or a decrease in cost for a ten minute increase in travel time would impact shared ride market shares. The analysis has implications on airport management policies and predicting usersâ?? responses to carpool version of share ride services such as Uber Pool and Lyft Line. Shared ride fares are currently only about 55% of the cost of taxis. The estimated nested logit model implies that if fares were to increase to match the cost of taxis, the demand for shared rides would fall from 9% to 7%. Meanwhile, if fares were cut by 50% and travel time increased by 10 minutes, the demand would rise by about 1.5%.
working paper
A dynamic (panel data) structural equations model is developed that links four dependent travel behavior variables at two points in time, one year apart. The four dependent variables are: car ownership, travel time per week by car, travel time by public transit, and travel time by nonmotorized modes. Exogenous variables include 13 household characteristics and variables accounting for period effects over the 1985 to 1987 time frame in the Netherlands. The model treats car ownership as ordered-response probit variables and all travel times as censored (tobit) continuous variables. The model accounts for serially-correlated errors and panel conditioning biases. Results are interpreted in terms of recommendations for forecasting procedures.
working paper |
May 1990
working paper |
Nov 1988
working paper |
Jun 1991
working paper |
Jun 1989
working paper |
Nov 1991
Preprint Journal Article
In the Lighthill-Whitham-Richards (LWR) model for single-lane traffic, vehicle trajectories follow the first-in-first-out (FIFO) principle and can be represented by a monotone three-dimensional surface of cumulative vehicle count. In contrast, the generalized bathtub model, which describes congestion dynamics in transportation networks using relative space, typically violates the FIFO principle, making its representation more challenging. Building on the characteristic distance ordering concept, we observe that trips in the generalized bathtub model can be ordered by their characteristic distances (remaining trip distance plus network travel distance). We define a new cumulative number of trips ahead of a trip with a given remaining distance at a time instant, showing it forms a monotone three-dimensional surface despite FIFO violations. Using the inverse function theorem, we derive equivalent formulations with different coordinates and dependent variables, including special cases for Vickrey’s bathtub model and the basic bathtub model. We demonstrate numerical methods based on these formulations and discuss trip-based approaches for discrete demand patterns. This study enhances understanding of the generalized bathtub model’s properties, facilitating its application in network traffic flow modeling, congestion pricing, and transportation planning.
research report
Phd Dissertation
This dissertation explores changes in the channels used for consuming prepared food (restaurant meals) and proposes optimization approaches for better managing a fleet of delivery vehicles. In the context of the COVID-19 pandemic, Chapter 1 examines how the consumption of prepared meals has evolved in California, with meal delivery gaining in popularity, dine-in experiences shrinking, and takeout witnessing marginal growth. I estimated heterogeneous ordered logit models to explain the frequency of consumption of restaurant meals before, during, and possibly after the pandemic for dine-in, takeout, and online orders with delivery using a broad range of explanatory variables, including components of the Social Vulnerability Index (SVI). My results show disparities in dine-in, takeout, and delivery frequencies, which have implications for equitable access to prepared meals.Chapter 2 extends my investigation to meal delivery in California and contributes to the traditional Food-Away-From-Home (FAFH) literature. I estimate spatial Durbin models to explain the demand for monthly meal delivery at the census tract level in three major MSAs (Metropolitan statistical areas) in California before and during the pandemic. Unique dynamics in meal delivery behavior emerge across regions and time, with accessibility proving pivotal in driving demand. In particular, I find that meal deliveries benefitted marginalized communities, which underscores the role of meal deliveries in enhancing food access. This chapter presents a holistic perspective, which encompasses business strategies and discusses policy implications. Chapter 3 explores a fleet management framework for meal delivery platforms based on graph theory optimization algorithms. I identified critical parameters for meal delivery operations and measured platform performance metrics such as Vehicle Hours Traveled (VHT), Vehicle Miles Traveled (VMT), and fleet size by adjusting the parameters. The comparative analysis of the Hopcroft-Karp and Karp algorithms reveals trade-offs between cost minimization and computational complex based on the algorithmic objects. My evaluation of Proposition 22’s impact on platform costs underscores the importance of modeling legal constraints. This chapter provides practical insights for platform operators to optimize service efficiency. It also provides directions for future research for more realistic simulations, including a dynamic approach, vehicle repositioning strategy, and consideration of different modes. Overall, this dissertation helps understand dynamic shifts in prepared meal consumption and delivery, and shows the importance of modeling legal constraints when optimizing the size of a delivery fleet. Findings could guide equitable policy interventions by highlighting the influence of demographic, regional, and economic factors on the frequency of restaurant meal consumption. My research bridges academia and practices through its interdisciplinary approach, which helps promote informed decision-making for platform managers, restaurant owners, and equity-conscious urban planners.
