Understanding Origin-Destination (O-D) travel demand is crucial for transportation management. However, traditional spatial-temporal deep learning models grapple with addressing the sparse and long-tail characteristics in high-resolution O-D matrices and quantifying prediction uncertainty. This dilemma arises from the numerous zeros and over-dispersed demand patterns within these matrices, which challenge the Gaussian assumption inherent to deterministic deep learning models. To address these challenges, we propose a novel approach: the Spatial-Temporal Tweedie Graph Neural Network (STTD). The STTD introduces the Tweedie distribution as a compelling alternative to the traditional 'zero-inflated' model and leverages spatial and temporal embeddings to parameterize travel demand distributions. Our evaluations using real-world datasets highlight STTD's superiority in providing accurate predictions and precise confidence intervals, particularly in high-resolution scenarios. GitHub code is available online(https://github.com/STTDAnonymous/STTD).

Decarbonizing transport is one of the core tasks for achieving Net Zero targets, but the COVID-19 pandemic disrupts human mobility and the established transport development strategies. Although existing research has explored the relationship between virus transmission, human mobility, and restrictions policies, few have studied the responses of multimodal human mobility to the pandemic and their impacts on the achievement of decarbonizing transport. This paper employs 32 consecutive biweekly observations of mobile phone application data to understand the influences of the pandemics on multimodal human mobility from February 2020 to April 2021 in London. We here illustrate that multimodal travel behavior and traffic flows significant changed after the pandemic and related lockdowns, but the decline or recovery varies across different travel modes and lockdowns. The car mode has shown the most resilience throughout the pandemic, but the travel modes in the public transit sector were hit hard. Cycle and walk modes remained high at the beginning of the pandemic, but the trend did not continue as the pandemic developed and the season changed. Our findings suggest that the COVID-19 pandemic brought more challenges to travel mode shifting and the achievement of decarbonizing transport rather than opportunities. This analysis will assist transport authorities to optimize the established transport policies and to redistribute limited resources for accelerating the achievement of decarbonizing transport.

Gauging viral transmission through human mobility in order to contain the COVID-19 pandemic has been a hot topic in academic studies and evidence-based policy-making. Although it is widely accepted that there is a strong positive correlation between the transmission of the coronavirus and the mobility of the general public, there are limitations to existing studies on this topic. For example, using digital proxies of mobile devices/apps may only partially reflect the movement of individuals; using the mobility of the general public and not COVID-19 patients in particular, or only using places where patients were diagnosed to study the spread of the virus may not be accurate; existing studies have focused on either the regional or national spread of COVID-19, and not the spread at the city level; and there are no systematic approaches for understanding the stages of transmission to facilitate the policy-making to contain the spread.To address these issues, we have developed a new methodological framework for COVID-19 transmission analysis based upon individual patients’ trajectory data. By using innovative space–time analytics, this framework reveals the spatiotemporal patterns of patients’ mobility and the transmission stages of COVID-19 from Wuhan to the rest of China at finer spatial and temporal scales. It can improve our understanding of the interaction of mobility and transmission, identifying the risk of spreading in small and medium-sized cities that have been neglected in existing studies. This demonstrates the effectiveness of the proposed framework and its policy implications to contain the COVID-19 pandemic.

Background: With the global spreading of Coronavirus disease (COVID-19), many primary care medical workers have been infected, particularly in the early stages of this pandemic. Although extensive studies have explored the COVID-19 transmission patterns and (non-) pharmaceutical intervention to protect the general public, limited research has analysed the measures to prevent nosocomial transmission based upon detailed interpersonal contacts between medical staff and patients. Aim: This paper aims to develop and evaluate proactive prevention measures to contain the nosocomial transmission of COVID-19. The specific objectives are (1) to understand the virus transmission via interpersonal contacts among medical staff and patients; (2) to define proactive measures to reduce the risk of infection of medical staff and (3) evaluate the effectiveness of these measures to control the COVID-19 epidemic in hospitals. Methods: We observed the operation of a typical primary hospital in China to understand the interpersonal contacts among medical staff and patients. We defined effective distance as the indicator for risk of transmission. Then three proactive measures were proposed based upon the observations, including a medical staff rotation system, the establishment of a separate fever clinic and medical staff working alone. Finally, the impacts of these measures are evaluated with a modified Susceptible-Exposure-Infected-Removed model accommodating the situation of hospitals and asymptomatic and latent infection of COVID-19. The case study was conducted with the hospital observed in December 2019 and February 2020. Findings: The implementation of the medical staff rotation system has the most significant impact on containing the epidemic. The establishment of a separate fever clinic and medical staff working alone also benefits from inhibiting the epidemic outbreak. The simulation finds that if effective prevention and control measures are not taken in time, it will lead to a surge of infection cases in all asymptomatic probabilities and incubation periods.

Structural disruptions in road networks, such as bridge closures or road outages, can severely impact traffic flow, leading to significant connectivity losses and unpredictable shifts in traffic patterns. Traditional traffic prediction models, designed for stable network conditions, often fail to adapt to these sudden changes in road capacity and connectivity. To address this challenge, we formalize flow redistribution caused by structural changes as a dynamic network prediction task. We then propose a novel feature-aware subgraph augmentation framework that enables Spatio-Temporal Graph Neural Networks (STGNNs) to learn robust redistribution patterns—even with limited historical data. Our framework simulates disruptions via subgraph perturbations to generate realistic training samples, effectively enriching the dataset and enhancing model generalizability to structural changes. Evaluated on the Hammersmith Bridge closure in London, the proposed augmentation strategy significantly improves model performance and outperforms data-hungry baselines, accurately capturing the disruption and its network-wide effects. This study demonstrates that targeted data augmentation can make STGNNs more effective in disruption scenarios with scarce historical data—offering a new, data-efficient paradigm for daily traffic prediction under both planned and unplanned network changes.

Green space exposure is associated with better mental health, but the independent contributions of accessibility and utilisation, and whether utilisation mediates the effects of accessibility and mental health, remain unclear. The objectives of this study were to (i) quantify the independent associations of green space accessibility and utilisation with common mental health difficulties (CMD) prevalence; (ii) test whether utilisation mediates the effect of accessibility; and (iii) investigate whether accessibility and utilisation moderate one another's associations with CMD. We integrated novel small-area anonymised mobile phone data, high-resolution environmental metrics, and primary care records for 2.6 million people across Cheshire and Merseyside, England to quantify these pathways in 2024. Results showed that accessibility and utilisation were independently associated with lower CMD prevalence. Residential tree exposure showed the most consistent association among accessibility measures (RR = 0.936), while walking time in green space showed the strongest association among utilisation measures (RR = 0.900). Mediation analyses provided no evidence that utilisation mediated the association between accessibility and CMD. However, moderation models identified a compensatory interaction, whereby associations between active green space engagement and lower CMD prevalence were strongest in areas with poorer environmental access. Our findings demonstrate that accessibility and utilisation have distinct but interacting associations with population mental health, supporting urban policies that promote both locally accessible green environments and opportunities for their active use.