Urban Currents

2026-08-02

7 arXiv categories· 96 journals· 273 candidates — 12 worth your time· 6 without an open abstract

No headline

a quiet day in urban data science

Today, together

tag shift
no tag ran above its 30-day average
canon
no foundational work is cited twice today
coupling
From individual trajectories to system flows: A multimodal mobility analysis framework using mobile phone application data shares 6 references with DIT: an end-to-end model for pointwise transportation mode identification (2026-07-30)
3 of them
  • Inferring hybrid transportation modes from sparse GPS data using a moving window SVM classification
  • Review of GPS Travel Survey and GPS Data-Processing Methods
  • Transportation mode detection – an in-depth review of applicability and reliability
; Affective dimensions of perceived safety for older adults’ mobility in urban Indonesia shares 6 references with Determinants of perceived accessibility of public transport and implications for equity: Evidence, gaps, and a research agenda (2026-06-12)
3 of them
  • Accessibility evaluation of land-use and transport strategies: review and research directions
  • Transport and social exclusion: Where are we now?
  • The importance of understanding perceptions of accessibility when addressing transport equity: A case study in Greater Nottingham, UK
; Affective dimensions of perceived safety for older adults’ mobility in urban Indonesia shares 5 references with Transport disadvantage and suppressed travel among low-income older adults: Lived experiences of social housing tenants in Ghent (2026-07-31)
3 of them
  • Transport and social exclusion: Where are we now?
  • The social consequences of transport decision-making: clarifying concepts, synthesising knowledge and assessing implications
  • The spatial context of transport disadvantage, social exclusion and well-being
institutions
University of Liverpool on 2 papers today; Tongji University on 30 papers in 30 days; Peking University on 22 papers in 30 days; Hong Kong Polytechnic University on 21 papers in 30 days; University of Hong Kong on 18 papers in 30 days

published: Journal of Transport Geography

Yuan Liao, Rafael H. M. Pereira, Jorge Gil, et al.
Journal of Transport Geography · was arXiv

Lund University

Digital Accessibility for Disabilities · Recreation, Leisure, Wilderness Management · Assistive Technology in Communication and Mobility · Paris · Lund University · Technical University of Denmark

Space–time accessibility supports participation in after-work leisure activities

This study evaluated how space-time accessibility (STA) influences weekday leisure participation using high-resolution GPS data from 2,415 working residents in the Paris region. Using structural equation modeling, the researchers measured how STA, calculated based on home-to-work routines and transport modes, affects the diversity and duration of visited leisure locations, finding a significant positive total effect (beta = 0.14) on participation.

Why it matters — It demonstrates that accessibility metrics based on daily routines and individual constraints, rather than static home locations, accurately predict actual leisure choices. This provides a validated, person-centered framework for planners to measure and address inequalities in urban social participation.

Caveat: The study relies on data from a single metropolitan area, the Paris region, which may limit the generalizability of the specific behavioral coefficients to cities with different urban forms or transit systems.

Journal of Transport Geography · pdf · doi

published: Journal of Transport Geography

Xianghui Zhang, Tao Cheng
Journal of Transport Geography

University of Liverpool

Human Mobility and Location-Based Analysis · Opportunistic and Delay-Tolerant Networks · Data Management and Algorithms · London · University of Liverpool · University College London

From individual trajectories to system flows: A multimodal mobility analysis framework using mobile phone application data

This study develops a data-driven framework that processes passive, crowdsourced mobile phone application data to reconstruct street-level multimodal traffic flows. The system uses a hierarchical travel mode detection algorithm combining Support Vector Machines and GIS map matching to classify individual trips with 89.56% accuracy despite irregular sampling and positioning noise. When aggregated and validated against physical traffic sensors in London, the resulting hourly and daily flow estimates show strong spatial and temporal correlations (r > 0.70).

Why it matters — It demonstrates that sparse, passive mobile phone data can serve as a highly scalable and cost-effective proxy for continuous, network-wide traffic monitoring, bypassing the need for dense physical sensor networks or highly labeled deep learning datasets.

Caveat: The framework's validation and performance metrics are based on a single case study in London, which may limit immediate generalizability to cities with different urban layouts or mobile data penetration rates.

Journal of Transport Geography · doi

published: Sustainable Cities and Society

Warnakulasooriya Umesh Ashen Lowe, Páraic Carroll, Leonhard Lades, et al.
Sustainable Cities and Society

University College Dublin

Urban Transport and Accessibility · Transportation Planning and Optimization · Older Adults Driving Studies · agent-based model · travel survey · University College Dublin

Using agent-based modelling for investigating the impact of travel behaviour interventions on mode shift: A travel behaviour simulation in Dublin

An agent-based model was developed using the Irish National Household Travel Survey to simulate how Dublin commuters shift from private cars to sustainable transport under different policy scenarios. The simulation evaluated soft policies, such as cycling campaigns and travel feedback programs, alongside hard policies like increased motor taxes and fuel prices, measuring changes in travel mode share, behavioral stage progression, and carbon dioxide emissions.

Why it matters — The model demonstrates that combining soft promotional campaigns with financial disincentives yields the greatest reduction in car use and emissions, while showing that tracking psychological stages of behavior change provides a more detailed assessment of policy effectiveness than mode share alone.

Caveat: The study relies on simulated agent behaviors derived from survey data rather than observed real-world responses to the implemented policies.

Sustainable Cities and Society · doi

published: Transport Policy

Dong‐Ping Song
Transport Policy

University of Liverpool

Merger and Competition Analysis · Climate Change Policy and Economics · Maritime Ports and Logistics · University of Liverpool

Incentive distortions in FuelEU Maritime: Reforming the GHG intensity penalty structure

This study analyzes the mathematical formulation of the FuelEU Maritime Regulation's greenhouse gas intensity standard and demonstrates that its current penalty structure creates a declining penalty per unit of non-compliance as emissions rise. Through analytical modeling and numerical examples, the research shows that the total penalty for a fleet is minimized when non-compliance is concentrated in a single vessel rather than distributed evenly, which incentivizes strategic manipulation of compliance balances.

Why it matters — The paper exposes a structural flaw in maritime decarbonization policy where marginal abatement incentives weaken for the highest-emitting vessels. By proposing a reformed penalty structure with constant marginal incentives, it provides a blueprint for policy adjustments that prevent strategic market distortions and ensure equitable enforcement.

Caveat: The findings and proposed reforms are based on analytical proofs and numerical simulations rather than empirical compliance data from active shipping operations.

Transport Policy · pdf · doi

published: Journal of Urban Mobility

Octaviani Ariyanti, Dora Sampaio, Ajay Bailey
Journal of Urban Mobility

Utrecht University

Older Adults Driving Studies · Urban Transport and Accessibility · Traffic and Road Safety · Utrecht University

Affective dimensions of perceived safety for older adults’ mobility in urban Indonesia

This study conceptualizes perceived safety as an embodied, affective experience for older adults navigating urban Indonesia. Using in-depth interviews, go-along interviews, and direct observations across two cities, the research documents how older residents experience fear, uncertainty, and emotional strain alongside practices of adaptation and endurance when encountering uneven infrastructure, mixed traffic, and limited transport services.

Why it matters — It shifts the analytical focus of transport safety away from technical metrics like accident rates or crime statistics toward the continuous, emotional negotiation of the built environment. This provides a framework for understanding how older adults' mobility decisions are shaped by the intersection of physical infrastructure, bodily health, and financial constraints.

Caveat: The findings are based on qualitative interviews and observations in two Indonesian cities, which may limit direct generalizability to regions with different cultural norms of aging or distinct transport infrastructures.

Journal of Urban Mobility · pdf · doi

published: Journal of Transportation Engineering Part A Systems

Omar Almasry, Ernest-John Ignacio, Jesus Osorio, et al.
Journal of Transportation Engineering Part A Systems

University of Illinois Urbana-Champaign

BIM and Construction Integration · Transport and Logistics Innovations · Transportation Systems and Logistics · University of Illinois Urbana-Champaign

Planning the Deployment and Layout of Smart Work Zone Systems on Roadway Projects

A novel decision-support model was developed to determine when to deploy smart work zone (SWZ) systems and to generate their physical layout designs. The model integrates analytical tools to evaluate quantitative and qualitative factors, such as predicted crash rates and queue lengths, and was demonstrated using a case study of a roadway construction project in Illinois.

Why it matters — It automates and standardizes the planning process for transportation departments, translating complex safety and mobility risks directly into actionable, site-specific equipment layouts that can help achieve reported safety benefits like reducing rear-end collisions by up to 70%.

Caveat: The model's practical capabilities and validation are demonstrated using only a single case study in Illinois.

Journal of Transportation Engineering Part A Systems · doi

published: Journal of Transportation Engineering Part A Systems

Luka Trček, Irena Strnad
Journal of Transportation Engineering Part A Systems

University of Ljubljana

Traffic control and management · Traffic and Road Safety · Transportation Planning and Optimization · University of Ljubljana

Stochastic Modeling of Motorway Work Zone Capacity from Field Data in Slovenia

This study applies a stochastic modeling framework combining survival-based capacity sampling with Weibull distribution fitting to estimate motorway work zone capacity. The model was calibrated using detailed field traffic data from several work zone configurations on the Slovenian motorway network, focusing on areas with prebreakdown flow observations. The analysis quantifies how different geometric constraints, such as lane narrowing and crossovers, reduce capacity and alter breakdown likelihood.

Why it matters — It transitions motorway work zone capacity planning from rigid, deterministic estimates to a probabilistic framework. This allows traffic managers to measure and predict breakdown risks under temporary lane closures and geometric restrictions rather than relying on static capacity values.

Caveat: The calibration of the Weibull distributions is limited to work zone configurations where sufficient prebreakdown flow observations were available in the Slovenian field data.

Journal of Transportation Engineering Part A Systems · doi

published: Urban Geography

Liliana De Simone
Urban Geography

Pontificia Universidad Católica de Chile

Urban Planning and Governance · Consumer Retail Behavior Studies · Water Governance and Infrastructure · Santiago · Pontificia Universidad Católica de Chile

Consumption infrastructure, retail urbanism and the production of space in Santiago

This study analyzes how shopping malls shape urban space in Santiago, Chile, using a combination of georeferenced mapping of retail infrastructure and semiotic discourse analysis of corporate and media documents. The research conceptualizes these malls as 'retail urbanism' hubs that align real estate development, urban planning, and corporate discourse to position private consumption spaces as pseudo-public civic anchors.

Why it matters — It demonstrates that while shopping malls are declining in the Global North, they remain highly resilient, expanding, and socially central in South American neoliberal urban contexts, where they actively drive gentrification and socio-spatial exclusion.

Caveat: The empirical findings are based on a qualitative and spatial analysis of a single city, Santiago, which limits direct statistical generalization to other regions.

Urban Geography · doi

published: Urban Science

José Carlos Rojas Quiroz, Carlos Ramiro Marmolejo Duarte
Urban Science

Universitat Politècnica de Catalunya

Housing Market and Economics · 3D Modeling in Geospatial Applications · Urban Planning and Valuation · machine learning · gradient boosting · clustering

More than Black Boxes: Machine Learning Models’ Capacity to Capture Housing Submarket Patterns

The study developed a sequential framework using OLS, Spatial Durbin Models, and an XGBoost model trained on 6,112 housing listings in Barcelona to extract SHAP values for individual properties. Clustering these SHAP values identified three distinct, latent housing submarkets characterized by systematic differences in valuation patterns. The stability and generalizability of these submarkets were validated using five methods, including spatial block bootstrap resampling and comparison against 2017 cadastral values.

Why it matters — It demonstrates that tree-based machine learning models can be used to discover latent, spatially coherent housing submarkets without relying on predefined administrative or socioeconomic boundaries, bridging the gap between predictive machine learning and econometric interpretability.

Caveat: The methodology was developed and validated using a relatively small dataset of housing listings within a single city.

Urban Science · pdf · doi

published: Urban Science

Pawita Boonrat, Voravika Wattanasoontorn
Urban Science

Prince of Songkla University

Municipal Solid Waste Management · Food Waste Reduction and Sustainability · Healthcare and Environmental Waste Management · random forest · Prince of Songkla University

Baseline Estimation and System Performance Assessment of Municipal Solid Waste in Tourism Areas: A Case Study of Phuket, Thailand

This study developed a data-driven framework to forecast municipal solid waste generation and detect system anomalies on Phuket Island, Thailand, using monthly tourism data as an exogenous factor. A Support Vector Regression model established the waste baseline with a Mean Absolute Percentage Error of 3.3%, outperforming five other models including SARIMAX and Random Forest. System performance and operational anomalies are identified using a deviation analysis that flags observed waste levels falling outside three standard deviations of the baseline residuals.

Why it matters — It provides island and tourism-dependent municipalities with a method to distinguish normal, tourism-driven fluctuations in waste from actual operational failures or systemic anomalies across the entire waste management chain.

Caveat: The predictive framework currently relies on aggregate monthly data and lacks spatially explicit information or waste composition details to guide localized collection strategies.

Urban Science · doi

preprint

Clarissa Rui Min Ong, Elizabeth Tee Inn Loo, Kenneth Woon Hao Soh, et al.
arXiv cs.CV

satellite imagery · Jakarta

From Forest to Future Capital: Tracking Land Cover Change in Ibu Kota Nusantara (IKN) from 2021 to 2026 with PlanetScope Imagery

This study tracked land use and vegetation cover changes in the Core Government Area of Indonesia's new capital, Ibu Kota Nusantara, from 2021 to 2026 using high-resolution PlanetScope SuperDove satellite imagery. Using spectral indices and a Support Vector Machine classifier, the analysis revealed a 672% expansion of developed land, an 18.1% decline in total vegetation, a 17.1% drop in mean NDVI, and a 0.28% decrease in total carbon stock, with the most intense vegetation loss occurring between 2023 and 2024.

Why it matters — It provides empirical, high-resolution measurements of the environmental trade-offs occurring during the construction of Indonesia's new capital, establishing a baseline that challenges the project's 'Forest City' branding with documented vegetation and carbon losses.

Caveat: The study area is restricted to the Core Government Area of the new capital, and the optical imagery used faced limitations in distinguishing specific construction phases and plantation-related land cover types.

arXiv · pdf

preprint

Dan Zhu, Chi Sin Ng, Litian Xie, et al.
arXiv cs.LG

recurrent neural network

Interpretable Machine Learning for Traffic Congestion Prediction: Unveiling the Impact of Different COVID-19 Periods

This study models traffic congestion in Alameda County, California, across pre-lockdown, lockdown, and post-lockdown periods using support vector regression, multiple linear regression, recurrent neural networks, and long short-term memory (LSTM) networks. Incorporating weather, seasonality, and pandemic-specific variables, a Bidirectional LSTM optimized with an adaptive parameter selection approach achieved the lowest Normalized Root Mean Square Error. Model interpretations via Integrated Gradients and SHAP revealed that while rising COVID-19 hospitalizations reduced congestion, higher fuel prices in the post-pandemic era did not deter private vehicle use, ultimately increasing congestion.

Why it matters — It demonstrates how the predictive power of specific external factors shifted across different phases of a public health crisis, proving that post-pandemic fuel price hikes failed to suppress private vehicle travel as commuters prioritized infection avoidance over transit.

Caveat: The findings and model performance are based on a single county in California, which may limit generalizability to regions with different transit infrastructures or pandemic policies.

arXiv · pdf

Still cited

Karen Lucas (2012), Transport and social exclusion: Where are we now? 1 of today's items cite it · 59 of 4454 in the archive stand on it

Also published today

These appeared today in journals we track. Their abstracts are not openly available, so we cannot summarise them.