Abstract Background

Alternative and advanced fuel vehicles, including electric, hydrogen, biodiesel, and ethanol-powered options, offer promising solutions to reduce greenhouse gas emissions and reliance on conventional fuels. Nevertheless, high upfront costs, sparse charging/refueling infrastructure, and vehicle performance uncertainties continue to hinder widespread adoption. While several policy initiatives, such as financial incentives, regulatory measures, and infrastructure investments, have been rolled out to address those challenges, the efficacy of such interventions varies significantly across states due to diverse local conditions. This highlights a critical need for thorough assessment of the policies to help stakeholders create more effective sustainable transportation strategies.

Description of Abstract

This research conducts a nationwide analysis of clean vehicle policies across the U.S., focusing on their influence on the clean vehicle adoption between 2016 and 2023. The approach begins with compiling an extensive policy dataset encompassing policy details and the corresponding statewide annual vehicle registrations by fuel type. Leveraging large language models (LLMs), specifically GPT-4, the study clusters policies based on their type (e.g., financial incentives, regulatory measures, and infrastructure support). Such clustering addresses the challenge of overlapping policies, which complicates isolating the impact of individual policies. Additionally, the use of LLMs reduces potential human error when handling extensive textual data, improving efficiency and consistency. Clusters were validated by manually reviewing a random 10% sample from each category.

Subsequently, the analysis adopts a Difference-in-Differences (DID) statistical framework, augmented by a two-way fixed effects model, to estimate the effectiveness of each policy cluster in increasing the share of targeted clean vehicles within a state, while accounting for temporal trends and state-specific factors like geography, culture, and climate.

Preliminary findings (focused on electric vehicle policies only) suggest that direct financial incentives often yield robust increases in vehicle adoption, whereas infrastructure and regulatory initiatives have more variable and generally moderate effects across different states. Therefore, it is recommended that in states where infrastructure incentives have limited impact—typically those with existing robust infrastructure—resources should be redirected towards enhancing financial incentives, such as tax credits, rebates, and purchase grants.

Statement on Why Abstract is Noteworthy

This work combines state-of-the-art text clustering via LLMs with rigorous policy impact evaluation, offering an unprecedented nationwide perspective on clean vehicle promotion strategies. The automated policy categorization expands the analysis scope, offering planners and policymakers insights from a wealth of unstructured policy documents. The dual-methodology framework can guide tailored recommendations for different states, accelerating the transition toward more sustainable transportation across the nation’s diverse landscapes.

Next Steps

Future work should incorporate additional control variables, including sociodemographic profiles, fuel prices, and charging/refueling infrastructure density, into the DID model. In addition, the analysis can be expanded to evaluate policy impacts on other clean vehicle types beyond electric vehicles.

Background

Travel demand modeling is highly refined for evaluating a few scenarios, but most models cannot quickly analyze the hundreds of scenarios needed to estimate project performance across a range of uncertain conditions, including those created by natural hazards such as flooding, earthquakes and wildfires. Therefore, when comparing projects for prioritization, it is difficult to place resilience projects (which show their benefit when uncertain events occur) on the same footing as capacity, accessibility, and safety projects.

Description

The open-source Resilience and Disaster Recovery (RDR) Tool Suite, available to download at https://volpeusdot.github.io/RDR-Public, is a metamodeling tool for evaluating the return-on-investment (ROI) of network infrastructure resilience projects across a range of hazard severities. The RDR Tool Suite has been publicly available since 2023 and continues to be updated in support of the U.S. Department of Transportation (DOT) Office of the Assistant Secretary for Research.  The RDR Tool Suite includes the Exposure Analysis Tool that translates hazard severity to network link availability, the RDR Metamodel, which uses a fast network routing model from the open-source AequilibraE project and concepts from the TMIP Exploratory Modeling and Analysis Tool to complete the scenario space, an ROI Analysis Tool that can perform Benefit Cost Analysis, Regret Analysis, or Breakeven Analysis for the user, depending on the information available, and a Benefits Analysis Tool to attribute project benefits among TAZs with differing attributes (categorical or continuous).  The economic analyses from the RDR Tool Suite are aligned with the U.S. DOT guidance on Benefit Cost Analysis, providing consistency with other methodologies and programs.  Planning agencies can use their own road and transit networks, hazard, and travel demand modeling data to populate the initial analyses to kick off the expansion of scenarios and metamodeling approach. Agencies can also use publicly available data (from OpenStreetMap, the Census, and other sources) to populate the RDR Tool Suite as a standalone model for estimating network performance across a range of conditions and projects. Both approaches are demonstrated in the documentation and sample reference scenarios distributed with the tool. The U.S. DOT Volpe Center provides technical assistance to users.

For several years, the Hampton Roads Transportation Planning Organization (HRTPO) has been using the fast scenario analyses of the RDR Tool Suite to aid in project prioritization, including using the RDR Tool Suite to analyze fiscally constrained resilience and other infrastructure investment projects and rural project prioritization. We will share an overview of the RDR Tool Suite capabilities and functionality, the data requirements to get started, and examples of analyses from the HRTPO to demonstrate the utility of the Tool Suite for project planning and prioritization.

Statement on Why Abstract is Noteworthy

The RDR Tool Suite can address a variety of hazard conditions, leveraging existing data available to state DOTs and TPO/MPOs. It can quickly assess hundreds of scenarios, to include a variety of hazard severities and possible mitigation projects. It then organizes the results to facilitate comparison among projects.

Project is Complete. 

The RDR Tool Suite was initially released in 2023 and continues to be enhanced in response to user needs, with semiannual releases.

Enhancing climate resilience is a key strategy in mitigating the flooding risks facing Houston's transportation system. The TxDOT Houston District identified a 1.8-mile segment of the IH-10 freeway, located west of downtown and carrying over 200,000 AADT, as vulnerable to flooding from the nearby White Oak Bayou. This stretch of the IH-10 White Oak corridor has become impassable multiple times during past major rainfall events. In response, TXDOT proposed rebuilding this section of the freeway as an elevated structure above the floodplain.

A regional mesoscopic Dynamic Traffic Assignment (DTA) model was used to simulate traffic’s rerouting behaviors, assess congestion propagation across IH-10 and neighboring routes, and quantify the traffic’s impacts using selected measures of effectiveness. This analysis was conducted for both flooding events and the subsequent repair and reconstruction period. The flooding scenarios considered both full and partial impassable freeways (i.e., capacity reductions to the traffic network).

The DTA model simulate different driver behaviors on the flooding day and the subsequent repair. As flooding event is not always predictable, many drivers could not anticipate its severeness before they left home and have to reroute their path on the roads. On subsequent scenarios, drivers  are aware of construction closure and accordingly plan their routes before starting their trips.  

The DTA model identified areas of new or exacerbated congestion in the transportation system resulting from flooding, confirming the overall negative impact on regional mobility. Here are some major takeaways from the model results

·         A full closure construction reduces more than half of the VMT on I-10 corridor and  significantly increase congestions on paralle facilities and almost all adjacent local arterials.

·         Over 90% of VMT stay on I-10 for partial closure construction scenario. Therefore, it impact is more localized. It increase relatively less congestion on parallel facilites. There are some local crossing streets with less traffic on upstream of the flooded I-10 segment.

·         Constrution closure on freeway and associate ramps functions as metering and ramp consolidation which improve the travel times on the rest of corridor by reducing turbulence of excessive travel demands.

·         On the flooding day, drivers scramble to leave IH-10 and hence increase demands on local cross streets, and parallel freeway and arterial routes. Demand drops on the major freeway connecting the IH-10 congestion as well. 

·         The fully impassible flooding-day scenario has highest VHT of all scenarios.

·         Parallel freeways were heavility relied on for alternative routes for all scenarios.

 The results showed that a partial capacity reduction during construction caused less regional delay and congestion propagation compared to scenarios where the freeway was completely impassable and damaged. Potential delay caused by severe flooding and subsequent full-closure repair construction is higher than partial consturction closure of[BB1]  elevating the segment above flood plain.

The modeling analysis has been completed, and reconstruction commenced in January 2025. Model results will be compared with observed data before and during the reconstruction process.

Abstract Background

Modern transportation planning and policymaking have focused on connecting communities through equitable transit-oriented development. LinkUS Columbus initiative, which includes five Bus Rapid Transit (BRT) corridors, is a comprehensive mobility and growth plan aimed at transforming transportation in Central Ohio. This study applied activity-based model to quantify potential reductions in both air quality pollutants and greenhouse gases resulting from the implementation of the BRT corridors.

Description of Abstract

Central Ohio, one of the fastest growing regions in the United States, is anticipated to gain nearly 1 million residents over the next 25 years, resulting in a total population of at least 3.15 million. To accommodate this growth, the Central Ohio Transit Authority (COTA) is modernizing the transportation system through the LinkUS initiative, which includes the development of five BRT lines. Mid-Ohio Regional Planning Commission (MORPC), the MPO for Columbus urbanized area, is one of the partner agencies for LinkUS collaborative initiative. MORPC conducted a study to quantify the emissions reduction resulting from the five BRT corridors.

MORPC's state-of-the-art regional activity-based travel demand model was used to simulate two future 2050 (all 5 BRT corridors are planned to be built by then) scenarios: one with the BRT corridors build and one no build. For network coding, some assumptions were made regarding the location of stops and roadway lane configurations due to the BRT corridors being in either planning stage or partial design phase during the study. The VMT projections in combination with local and default data sources were used for US EPA MOVES4 modeling to estimate county-scale emissions. In this analysis, emissions are computed and compared using two distinct methodologies. The first method followed air quality conformity analysis in accordance with the Ozone standards, focusing on Volatile Organic Compounds (VOC) and Nitrogen Oxides (NOx). The second method was to compute greenhouse gas emissions in terms of Carbon dioxide (CO2) equivalent.

The findings, firstly computed for a typical summer day and converted to annual emissions, indicate significant reduction with the implementation of the BRT corridors. This study underscores the potential of BRT systems to significantly mitigate the environmental impact of urban growth and offers valuable insights for sustainable transportation planning.

Statement on Why Abstract is Noteworthy

Policymakers from Ohio Environmental Council (OEC) and Columbus City council's priority was to understand the emissions impact of LinkUS BRT corridors to support this initiative. This study, applying regional activity-based travel demand model and US EPA’s MOVES4 model, quantified emissions reduction from implementing BRT. Other planning agencies can replicate the study for justifying more equitable and accessible transportation systems.

Project is Complete

Travel demand models are essential tools for understanding the interplay between land use and transportation, allowing us to assess the impact of future demand on the transportation supply. These models are typically validated against traffic counts and transit boardings observed on typical non-summer weekdays to simulate typical "worst-case" conditions. But are non-summer weekdays truly the worst-case scenarios?

Natural disasters—such as hurricanes, flooding, wildfires, landslides, and rising sea levels—can force large, vulnerable populations to evacuate, placing tremendous pressure on the transportation network. This highlights the need for resilience studies that identify connectivity vulnerabilities and assess improvement projects aimed at ensuring timely evacuations and emergency responses.

For San Bernardino and Western Riverside Counties we are utilizing an expanded version of the San Bernardino Transportation Analysis Model (SBTAM) to support an emergency evacuation and network resiliency (EENR) study. This study uses the travel model in an unconventional manner, as evacuation scenarios differ greatly from typical conditions. During an evacuation event, both the transportation supply (available network) and demand (people needing to travel) can vary significantly based on the type and location of an event. The travel model is supporting several aspects of this study.

·        Identifying an evacuation population – the number of people who need to evacuate an area at a given time of day. This evacuation population may be primarily residents during an overnight evacuation, but is a more complex mix of residents and other people in an area during the day.

·       Pinpointing areas with limited egress options, such as areas with only one way out, or areas where a secondary exit requires a significant and time consuming detour.

·       Locating areas with existing and forecast recurring congestion, and identifying cases where recurring congestion overlaps with critical evacuation routes.

·       Evaluating emergency response time, helping identify areas where emergency responders such as firefighters, EMTs, and police response times may be lowest – especially in cases of network disruption.

Results from these specialized modeling case studies, along with data obtained from location-based services (LBS) and GPS data, is informing discussions with stakeholders and local governmental agencies on improvements to network connectivity and resilience to emergency events. Modeling work is expected to be completed by Fall 2025.