Spatial measures of accessibility are widely used in urban and transportation planning to understand the impact of the transportation system on influencing people’s access to places, but publicly available large-scale datasets are rare and limited. This paper presents a highly parametric dataset containing values of spatial accessibility measured by combinations of multiple metrics, travel modes, types of opportunity (including jobs and amenities like schools, hospitals, and electric vehicle charging stations), and travel time thresholds. This includes both cumulative opportunities types of measures as well as competition metrics. A total of 600 accessibility values are computed for each zone at three administrative levels for the 50 most populous urban areas of the United States. Additionally, the dataset also includes the travel time matrix files for each of these urban areas by three travel modes – driving, walking, and bicycling – to facilitate self-validation. Further, comparisons with similar travel time and accessibility datasets show a high degree of similarity with our dataset. ... Read more

The concept of ‘distance decay’ curves is used in spatial interaction and accessibility analysis to represent the diminishing likelihood of visiting places with increasing travel impedance, mainly distance and travel time. The shape of the resulting impedance decay curves varies by several factors, but these influential factors are often dismissed in favor of just travel mode. In this study, we examine which factors should be used to distinguish impedance functions for use. Using data from a large national travel survey, we first show that the impedance curves are well approximated by functions of the exponential family and the related Tanner function. We use two methods – variable-wise function fit and Shapley additive explanations – to conclude the importance of four factors for developing impedance functions. These are travel mode, trip purpose, urbanity class of trip origin and destination, and the socioeconomic status grouping of the travelers. We then show that the use of a generalized impedance function can significantly over- or underestimate spatial accessibility compared to factor-specific impedance function, with up to 80 % overestimation on average in the case of public transit and 16 % overestimation for low socio-economic status travelers. These findings highlight the importance of the choice of impedance function which has applications in spatial economics, transportation planning, and human mobility analyses. ... Read more

Estimation of people’s home locations using location-based services data from smartphones is a common task in human mobility assessment. However, commonly used home detection algorithms (HDAs) are often arbitrary and unexamined. In this study, we review existing HDAs and examine five HDAs using eight high-quality mobile phone geolocation datasets. These include four commonly used HDAs as well as an HDA proposed in this work. To make quantitative comparisons, we propose three novel metrics to assess the quality of detected home locations and test them on eight datasets across four U.S. cities. We find that all three metrics show a consistent rank of HDAs’ performances, with the proposed HDA outperforming the others. We infer that the temporal and spatial continuity of the geolocation data points matters more than the overall size of the data for accurate home detection. We also find that HDAs with high (and similar) performance metrics tend to create results with better consistency and closer to common expectations. Further, the performance deteriorates with decreasing data quality of the devices, though the patterns of relative performance persist. Finally, we show how the differences in home detection can lead to substantial differences in subsequent inferences using two case studies—(i) hurricane evacuation estimation, and (ii) correlation of mobility patterns with socioeconomic status. Our work contributes to improving the transparency of large-scale human mobility assessment applications. ... Read more

Disasters faced by human society are becoming more frequent and complex, raising a need to model the combinations of different types of disasters, such as hurricanes and pandemics. In this paper, we explore various modeling options for predicting aggregated individual evacuation metrics under the compound risks drawn by COVID-19 and Hurricane Ida (2021) using large-scale location-based services data. For each model, we compare its performance with other options and analyze the SHapley Additive exPlanation (SHAP) values to understand the impact of different explanatory variables on the model outcome. The results suggest that the COVID-19 factors marginally enhance the modeling of evacuation rates and distance, holding similar importance to traditionally recognized factors such as the percentage of senior people and vehicle ownership. Further analysis also suggests the impact of COVID-19 factors diminishes with distance from the coast. Moreover, we observed that the contributions of COVID-19 factors increase significantly when their values reach extreme levels, both very low and very high, suggesting that evacuation patterns were notably impacted under these conditions. Our findings contribute to understanding the impacts of various factors on evacuation patterns during Hurricane Ida, inform model selection for predicting critical evacuation/return metrics, and enrich the knowledge base of evacuation modeling in scenarios involving compound risks. ... Read more

Protective action decisions are complex and require households to consider a variety of factors including hazard risks, household characteristics, location, and experience. Previous research has examined these decisions in the context of a single hazard; however, it is also necessary to consider scenarios whereby two coupled hazards pose simultaneous risks. This paper examines the decision-making of households when making evacuation and shelter-in-place decisions during two simultaneous events: the 2021 Hurricane Ida and the COVID-19 pandemic. Survey data gathered from six parishes eight months after the storm were used to analyze risk perception tradeoffs and protective action decisions of households before Hurricane Ida made landfall. The results indicate that higher perceived risks of the hurricane causing injury or death, being vaccinated, having a higher income, having children residing in the home, and having previous experience of evacuating increased the odds of evacuating. Likewise, variables such as higher perceived risks related to being hospitalized or killed by COVID-19, being elderly, and being located further away from the storm track all decreased the likelihood of a household undertaking evacuation. The findings of this study improve understanding of how households consider competing risks during simultaneous hazard events, which in turn can help inform strategies for managing future disaster events involving multiple hazards and risks. ... Read more

Measures of walkability generally do not provide a detailed quantitative assessment of pedestrian infrastructure development prioritization. In this study, a link-based composite measure of walkability and walking is introduced to overcome this limitation. This measure, called ‘pednet score’, is based on a weighted pedestrian network (‘pednet’) made of sidewalks and crosswalks whose edge weights are descriptive of their popularity. Edge popularity is derived from home-based walk trip assignments derived from simulated pedestrian demand. Properties of the pednet score are studied using three hypothetical variants of the pednet in three North American cities, each involving the addition of candidate sidewalk and/or crosswalk segments. It is shown that a strategic selection of these segments based on pednet score can substantially increase walking trips, in some cases up to 236%, and reduce current mean pedestrian trip distances by up to 340 m. A mixed development approach involving both sidewalks and crosswalks also shows considerably higher improvement than those segments considered alone. Results from marginal benefit curves strongly indicate the utility of the pednet score as a measure of link criticality for segment prioritization in pedestrian network design. ... Read more

Pedestrian network creation and completion for various tasks in transportation planning engineering generally requires laborious manual labeling of sidewalks and crosswalks. In this study, we employed the deep learning-based object detection model YOLO v5 for detecting crosswalks using Google Maps satellite imagery data of four North American cities. We also tested for two additional scenarios involving two image-processing techniques to exploit the special properties of crosswalks. Observations showed that the original images performed better than with binary thresholding and nearly the same even when noisy regions outside the right-of-way were removed. We proposed an algorithm to assign the detected crosswalks to intersection legs. We demonstrated the effectiveness of this technique for Washington, D.C. and Los Angeles, CA, showing classification accuracy rates of 71% and 89%, respectively. We also showed the influence of increasing distance threshold, a tolerance radius of prediction accuracy, in degrading the classification performance. This method, when combined with existing methods for sidewalk detection from aerial and street view images, could be reasonably used to help complete urban pedestrian networks in cities where high-resolution satellite images are available. ... Read more

Conventional evacuation studies typically do not gauge the development of participants’ certainty about evacuation-related decisions with the updates in the information provided to them. This study uses an online survey that provides three kinds of progressively varied information about the current status of a hypothetical hurricane for five days leading to its landfall and collects respondents’ certainty of their situational comprehension and evacuation-related decisions each day. Most participants (84%) made a final decision (60% evacuate) after seeing information of just one day (four days before the landfall), indicating a tendency of swift decision-making. Modeling shows that the time spent looking at information, especially uncertainty cone forecast maps, positively influences the understanding of the hurricane’s status, which in turn helps in increasing the certainty of making evacuation-related decisions, with an increasing temporal effect. This study contributes to the understanding of the public perception of information and its association with evacuation-related decision-making. ... Read more

We investigate the effects of the amount and kind of information received by hurricane evacuees on the level of urban evacuation-induced traffic congestion. With the help of agent-based simulation driven by survey data for evacuees of Hurricane Matthew in Jacksonville, FL, we find that sending evacuation notices to households stands out as the most dominant factor impacting evacuation congestion. We use travel time metrics and introduce a percolation congestion index to show that congestion increases marginally by providing more mandatory than voluntary notices, which compensates for the benefits that are obtained by higher evacuation. We also observe that segments of commonly used evacuation routes in the flood-prone areas are more likely to be congested during the evacuation period than the other road segments. This study affirms the importance of evacuation notices in evacuation planning and suggests that planning agencies might benefit by strategically sending these notices to people to control peak congestion. ... Read more

The rapid early spread of COVID-19 in the US was experienced very differently by different socioeconomic groups and business industries. In this study, we study aggregate mobility patterns of New York City and Chicago to identify the relationship between the amount of interpersonal contact between people in urban neighborhoods and the disparity in the growth of positive cases among these groups. We introduce an aggregate spatiotemporal contact density index (CDI) to measure the strength of this interpersonal contact using mobility data collected from mobile phones, and combine it with social distancing metrics to show its effect on positive case growth. With the help of structural equations modeling, we find that the effect of CDI on case growth was consistently positive and that it remained consistently higher in lower-income neighborhoods, suggesting a causal path of income on case growth via CDI. Using the CDI, schools and restaurants are identified as high contact density industries, and the estimation suggests that implementing specific mobility restrictions on these point-of-interest categories is most effective. This analysis can be useful in providing insights for government officials targeting specific population groups and businesses to reduce infection spread as reopening efforts continue to expand across the nation. ... Read more

Estimation of vehicular emissions at network level is a prominent issue in transportation planning and management of urban areas. For large networks, macroscopic emission models are preferred because of their simplicity. However, these models do not consider traffic flow dynamics that significantly affect emissions production. This study proposes a network-level emission modeling framework based on the network-wide fundamental diagram (NFD), via integrating the NFD properties with an existing microscopic emission model. The NFD and microscopic emission models are estimated using microscopic and mesoscopic traffic simulation tools at different scales for various traffic compositions. The major contribution is to consider heterogeneous vehicle types with different emission generation rates in a network-level model. This framework is applied to the large-scale network of Chicago as well as its central business district. Non-linear and support vector regression models are developed using simulated trajectory data of 13 simulated scenarios. The results show a satisfactory calibration and successful validation with acceptable deviations from the underlying microscopic emissions model regardless of the simulation tool that is used to calibrate the network-level emissions model. The microscopic traffic simulation is appropriate for smaller networks, while mesoscopic traffic simulation is a proper means to calibrate models for larger networks. The proposed model is also used to demonstrate the relationship between macroscopic emissions and flow characteristics in the form of a network emissions diagram. The results of this study provide a tool for planners to analyze vehicular emissions in real time and find optimal policies to control the level of emissions in large cities. ... Read more

Winter maintenance trucks (WMTs) often operate at lower speeds during inclement weather and roadway conditions, creating potential safety issues for motorists following close behind. In this study, a new prototype radar-based rear-end collision avoidance and mitigation system (CAMS) was tested to assess its impact on the behavior of drivers following WMTs. The system is designed to flash an auxiliary rear-facing warning light upon detection of a vehicle encroaching within an unsafe relative headway with the rear of the WMT. A series of field evaluations was performed during actual winter maintenance operations to assess the effectiveness of the system compared with normal operating conditions (i.e., without the CAMS warning light) toward improving driver behavior related to rear-end crash risk. Specifically, two measures were assessed: (a) rate of vehicles encroaching beyond a safe time headway threshold to the rear of the WMT, and (b) the reaction–response time of drivers. Classification and regression tree models were created for identifying the relevant factors influential in determining the change in driver response. The results indicate that this warning light was effective in reducing the likelihood of the subject drivers crossing beyond a relative headway of 4.5 s. It was also effective in reducing the reaction and response times of the drivers by 0.83 and 0.55 s (36% and 20% reduction), respectively. Although the results were encouraging, additional field testing is recommended before conclusions are drawn regarding the traffic safety impacts of the system. ... Read more

A microsimulation software, PTV-VISSIM is gaining popularity to model heterogeneous traffic conditions that exists on Indian roads. However, for such traffic condition, a detailed calibration of VISSIM model in terms identification of model parameters considering different vehicle class is still lacking. The present study proposed an efficient technique to calibrate VISSIM model using data collected from signalized intersections located in Delhi, India. As traffic flow varies in urban networks during different time of day, the calibration of VISSIM model is done for both peak and off-peak hour traffic. Model calibration is done for one site whereas another one is used to check models’ transferability. Initially, an automated sensitivity analysis is carried out for different traffic hour models with respect to each vehicle class by considering saturation flow (SF) as a performance measure. Results shows that all the vehicle classes have a distinct influence on the driving behaviour parameters as well as model’s sensitivity differ with respect to peak and off-peak hour volume. Subsequently, simulated SF and the optimum values of calibrated model parameters are obtained for each vehicle class using an optimization technique, Genetic Algorithm (GA). The error between simulated SFs and field ones are identified in terms of GEH statistics for each case of calibration, validation and model transferability. For all cases, GEH values are found within the recommended value for acceptable fit thus indicating the models are applicable for locations of similar as well as varying intersection width and different traffic volume ranging from 2057 to 4012 veh/hr. ... Read more

Estimation of vehicular emissions is one of the main challenges that planning agencies and municipalities encounter. This is particularly the case for large cities that typically struggle with congestion problems. Microscopic models are widely used to estimate the vehicular emissions. These models assist system planners in proposing strategies to control the level of emission. However, applications of these microscopic models are limited to estimation at the facility-level or small networks. To estimate emissions for large-scale networks, which is a major challenge for city planners, macroscopic models are typically proposed. However, these models fail to consider traffic flow dynamics. This study proposes a mesoscopic approach that is capable of estimating emission for large networks, as well as capturing the effects of traffic flow dynamics. ... Read more