Global maritime networks are highly complex and dynamic systems, organized around hubs and corridors that form hierarchical structures essential to international trade. Hubs and corridors are conceptually interconnected; however, their identification has often relied on distinct analytical approaches, with hubs typically defined by connectivity-based metrics and corridors by patterns of spatial interaction and flow concentration. Although resilience studies recognize the interdependence of hubs and corridors, existing approaches often rely primarily on hub-centered data, offering limited empirical insight into how corridor-level flow dynamics evolve under external disruptions such as pandemics or economic shocks. To bridge this gap, this study introduces a Maritime Hub–Corridor framework that integrates hubs and their inter-hub linkages into a hierarchical network system, developed through a structured, sequential analytic procedure. Applying this framework to Port-of-Call data, we use three complementary indices—connectivity, freight capacity, and spatial interaction—to capture distinct dimensions of hub functionality. Rather than treating these as isolated computed measures, we use them to screen candidate hubs, evaluate them against a Cumulative Flow Contribution Curve (CFCC) criterion across all dimensions, identify the resulting hub set and hierarchy, and grouped interconnected hubs into corridor clusters. Findings reveal that maritime networks organize as differentiated hub-corridor systems across four spatial types, exhibiting multi-scale hierarchies. Additionally temporal analysis reveals the adaptability and contrasting resilience mechanisms of hub-corridors to socio-economic disruptions: Suez consolidation versus Panama fragmentation. This research contributes a novel analytical approach for identifying hub-corridors structures systematically, characterizing multi-scale maritime network organization, and understanding resilience mechanisms under major disruptions.
The science output of a region may depend on various intraregional and interregional factors. A Spatial Durbin Model was used to estimate the impacts of regional factors as well as spillovers from other regions on science output in European NUTS2 regions. Three different weightings of interregional impact were tested to capture geographical proximity: the inverse distance, the inverse distance squared, and a binary distance-threshold specification that excluded distant regions. The main findings are that research and development (R D) employment in higher education institutions in the region and in surrounding regions has a positive impact on science output, while R D employment in high-tech industries in the region has a negative impact. There are also intraregional and interregional effects of disciplinary specialization, with less specialized regions exhibiting higher science output. Regions with older universities exhibit higher within-region scientific output, although the total spatial effect of university age is not statistically robust once spillovers are accounted for.
Urban planners are increasingly tasked with strengthening neighborhood economies and encouraging more connected communities. Central to this objective is ensuring equitable access to essential amenities such as grocery stores, parks, and hospitals. In recent years, shared e-micromobility services such as publicly available e-bikes and e-scooters have gained popularity as a means of local transport, offering potential to extend access. However, their contribution to the 15-Minute City framework across cities is not well understood. In this study, we assess if shared e-micromobility services expand access to essential amenities (i.e., supermarkets, schools, hospitals, parks, and entertainment venues), across five international cities. Observed e-micromobility trips are used to generate city-level travel time surfaces, which are compared with counterfactual walking times for the same trips. Results suggest access gains from e-micromobility are uneven, largely determined by urban form, distribution of amenities, and distance from city center. Cities and neighborhoods with less compact urban form and more dispersed land use tend to experience the largest gains in access with e-micromobility, increasing the proportion of the city reachable within 15 min by up to threefold in some cities. This work provides evidence to inform policies to better integrate new forms of local-scale transport (i.e., e-micromobility services), to improve access to essential services within the 15-Minute City framework.
This study develops a Fractional-Order Reaction–Diffusion (FO-RD) model to capture the spatiotemporal dynamics of the adoption of Remote Healthcare Systems (RHS) in geographically isolated regions. By integrating Caputo fractional derivatives (CFDs) into a classical RD framework, the model incorporates memory effects and non-Markovian dynamics that characterize human behavioral responses to healthcare interventions. The model comprises three interacting compartments: unaware population, adopter population, and healthcare need, which evolve through socially driven local reactions and spatial diffusion. Theoretical stability analysis establishes conditions for local and global Mittag-Leffler stability of the equilibrium points. Numerical simulations validate the theoretical results and demonstrate how the FO modulates memory strength in adoption dynamics. Specifically, lower values of the fractional order (stronger memory) lead to slower initial adoption but more sustained long-term adoption. Sensitivity analysis highlights the influence of key parameters such as social contagion rate, external influence, adopter mobility, and relapse rate. The model serves as a quantitative tool for predicting adoption patterns, identifying geographical tipping points, and evaluating policy interventions aimed at enhancing healthcare equity in remote areas.
Spatial panel models with heterogeneous coefficients provide a flexible framework for capturing region-specific spillovers, but their practical use is often constrained by two challenges: (i) existing estimators primarily justified under large- T asymptotics, leaving unit-specific effects imprecisely estimated in the short panels common in regional science; and (ii) allowing spatial dependence and slopes to vary across units yields a high-dimensional parameter space that is difficult to estimate efficiently. This paper proposes a hierarchical Bayesian spatial model with heterogeneous coefficients that addresses both issues through hierarchical priors and an integrated Hamiltonian Monte Carlo estimation framework. The hierarchical prior induces partial pooling across units, shifting inference toward low-dimensional population hyperparameters while stabilizing unit-level estimates in finite samples, particularly when N>T . Furthermore, we develop scalable evaluation of the log-determinant term using a power-series expansion and Hutchinson stochastic trace estimation, and we exploit sparse matrix-vector multiplication and parallel computing to improve computational efficiency. Comprehensive simulation experiments demonstrate that the proposed Bayesian framework delivers reliable finite-sample inference and improved estimation accuracy relative to maximum likelihood in short panels. The results also show that the heterogeneous spatial dependence is more sensitive to the density of the spatial weights matrix than slope coefficients, offering practical guidance for applied research.
Complex urban systems are shaped by the ongoing interaction between regulatory rules and emergent spatial dynamics. In planning theory and urban modelling, regulation is often conceptualized as an external instrument designed to steer or correct self-organizing processes. This paper challenges that separation by reconceptualizing regulation as an endogenous process generated within the dynamics it seeks to govern. The paper introduces a divergence-driven framework in which regulation evolves through the persistent gap between regulatory intentions and realized spatial outcomes. Rather than being treated only as policy failure, divergence is understood as a generative condition that produces recursive feedback between regulation and urban development. By developing a formal representation of this process, the paper shows how regulatory adjustment and urban dynamics may co-evolve through ongoing interaction. The contribution is primarily theoretical and methodological, offering a formal account of regulation as endogenous feedback and reframing divergence as a structural driver of adaptive governance in complex urban systems. This perspective shifts the focus from the pursuit of control and equilibrium toward the management of uncertainty and persistent misalignment.
Biodiversity-friendly farming requires spatially coordinated land use to effectively deliver landscape-scale environmental benefits such as biodiversity and other ecosystem services. This paper argues that addressing such coordination challenges requires a stronger integration of ecological, economic, and spatial perspectives within geographical science. Using the Dutch agri-environmental scheme (ANLb) as an empirical case, the paper examines the role of farmer collectives as an institutional innovation aimed at improving spatial coordination of conservation measures on privately owned farmland. Since 2016, these collectives have been responsible for coordinating agri-environmental management across regions, aligning farm-level decisions with landscape-scale biodiversity objectives. However, evidence indicates that spatial coordination remains insufficient to meet these objectives, limiting ecological effectiveness at the national scale. This ‘implementation gap’ is closely linked to economic and behavioral factors, including the voluntary nature of participation, misaligned incentives, and practical constraints at the farm level. The effectiveness of collective approaches depends on achieving sufficient scale and intensity of participation, supported by adaptive and spatially differentiated policy instruments. In conclusion, farmer collectives offer a promising pathway towards more effective landscape-scale biodiversity governance, but their potential depends on closer integration of ecological, economic, and geographical considerations in both research and policy design.
This paper examines the relationship between crop diversity, organic farming, and agricultural efficiency across European regions using data from the Farm Accountancy Data Network (FADN). Technical efficiency is measured via a conditional Debreu-Farrell input-oriented index within a Data Envelopment Analysis (DEA) framework, based on labor, capital, and land inputs. Efficiency scores are analyzed using panel data models with correlated random effects to distinguish within- and between-region variation. We find a statistically significant positive effect of organic farming on efficiency within regions but not across regions, suggesting that local adoption improves productivity even if regional averages do not. In contrast, crop diversity is linked to lower efficiency within regions - a counterintuitive result that possibly reflects adaptation to external shocks rather than productivity aims. Taking into account the strong spatial dependence, we apply spatial panel models to a balanced subsample, confirming the robustness of our results. Climate variables further inform the analysis: temperature variability reduces efficiency, while higher average temperatures support it. The negative association between crop diversity and efficiency highlights the dual role of diversification as a resilience strategy and a possible productivity trade-off, which deserves further research.
In today’s increasingly mobile world, the dynamics of connectivity and accessibility are reshaping our understanding of movement. Mobility extends beyond physical travel, acting as a form of social currency deeply intertwined with broader societal structures and inequalities. Distinct daily mobility patterns allow different groups to navigate and experience urban spaces in diverse ways. This study aims to understand daily mobility profiles across cities and examine whether these profiles differ between locations or reflect similar patterns in different urban contexts. We analyze the daily movements of approximately 13,000 individuals in Auckland, New Zealand, and 23,500 individuals in Rio de Janeiro city, Brazil, using individual daily mobility data and a clustering algorithm to identify four distinct mobility groups. Our findings suggest a broader, more universal pattern of mobility-based divisions. While these groups show moderate mixing in residential areas, they are more segregated in the places they visit. This suggests that daily mobility, like other forms of capital, has the potential to reinforce (spatial) segregation, as individuals who live in close proximity often frequent different parts of the city, limiting exposure between members of different mobility groups.
Traditional spatial optimization problems such as the maximal coverage location problem and the p-median problem are NP-hard, which makes it difficult to obtain optimal solutions in general. To address this computational difficulty, metaheuristics have been widely deployed. However, they usually operate in spatially blind contexts that do not explicitly account for geographic structures, tending to converge to suboptimal solutions. This study proposes a new hybrid framework named GAT-Metas that integrates a Graph Attention Network (GAT) with metaheuristics to overcome these limitations. GAT-Metas trains GATs in a supervised proxy prediction task to learn an interpretable map of spatial potential. The learned spatial potential is then used to probabilistically bias both initial solution construction and neighborhood search within the metaheuristic and remains fixed during each metaheuristic run. We validate the framework using synthesized grid instances and a case study on optimal placement of shared e-scooter parking zones in Seoul. The experiments show that the GAT-SA and GAT-TS outperform their purely random-initialized counterparts and, for some large instances, even time-limited commercial Mixed-integer Programming baselines, yielding statistically significant improvements and more stable solutions. By combining deep learning’s pattern recognition ability with the exploratory power of classical optimization methods, the GAT-Metas offers a methodological advance for solving complex urban siting problems more effectively, reliably, and interpretably.
The debate over administrative and experiential definitions of neighborhood boundaries has been a longstanding concern in urban geography and planning. While a substantial body of empirical work has examined administrative, behavioral, and perceptual representations of neighborhoods, these approaches have often been studied separately, limiting understanding of how they differ and relate to one another, particularly among underrepresented groups such as young adults. This study examines how administrative boundaries diverge from individuals’ perceived neighborhoods and activity spaces, based on spatial behavior data from 231 young adults in the Sinchon area of Seoul, South Korea. Results show that administratively defined neighborhoods average about 2.8 km², while perceived neighborhoods average 1.7 km², and activity-based neighborhoods remain below 1.0 km², indicating systematic overestimation by administrative units. These discrepancies are significantly associated with sociodemographic characteristics such as income level, housing type, and student status. The findings suggest the presence of the Uncertain Geographic Context Problem (UGCoP), whereby failure to account for how individuals perceive and interact with space can lead to mismatches between the provision of public services and actual spatial demand. To mitigate this limitation, urban planning practice should incorporate behaviorally relevant spatial data to better align policy interventions with actual spatial behaviors and service demands.
In the context of territorial development, we introduce a novel decision support framework aimed at identifying, characterising, and prioritising spatially explicit strategic corridors, with a focus on their developmental processes. These corridors offer opportunities for directing investments into transport and urban systems from a 'policy-first' perspective. We outline the development process and structure of a domain model for strategic corridors, which is constructed based on both conceptual and behavioural models documented in Unified Modelling Language. This domain model, driven by standards, is designed to translate requirements into technical components, ensuring interoperability while allowing for flexible extensions. Unified Modelling Language models facilitate adherence to standards of the International Organisation for Standartisation’s Technical Commitee 211 and the Infrastructure for Spatial Information in Europe, and leverage best practices in integrating data from diverse stakeholders across various spatial scales. Beginning with policy scenarios, we streamline the required data to a minimum, derive and merge numerous quantitative indicators and utilise spatial data to rank strategic corridors across different policy scenarios. We apply this integrated framework across the entire African continent, identifying, characterising, and ranking 55 terrestrial corridors using policy-driven utility functions and indicator weights. For selected strategic corridors, incorporate feedback from local stakeholders, refine scenarios based on investment priorities, propose subcorridors and conduct recursive characterisation and ranking of them, focusing on areas ripe for intervention and development. This innovative framework serves as a systematic foundation for developing diverse applications, fostering interoperability across domains. It enables the transferability of the decision support framework to assess potential investment opportunities at various spatial scales across different regions of the world.
Geographically weighted regression (GWR) model is popular as it effectively detects spatial non-stationarity of regression relationships, where all coefficients are considered to be varying across geographical coordinates. As an extension of the geographically weighted regression (GWR) model, the multi-type coefficient geographically weighted regression (MTC-GWR) model incorporates not only zero, non-zero constant, and spatially varying coefficients but also coefficients associated with a single geographical coordinate. Based on the local linear estimation method of GWR and adaptive group lasso technique, this study proposes a novel structural identification method that penalizes both regression coefficients and their partial derivatives. This method enables the simultaneous identification of four coefficient types: zero, non-zero constant, single-coordinate-related, and spatially varying coefficients. Within this framework, constant coefficients are averaged across all geographical locations, while coefficients related to a single coordinate direction are averaged along the other coordinate. The performance of model structure identification and the precision of coefficient estimation was evaluated through simulation experiments, which was applied to the analysis of the Dublin voter turnout dataset.
The intensification of the urban heat island effect as a consequence of global warming represents a considerable threat to urban populations, elevating the risk of heat exposure. Studies have investigated the effects of urbanization or greening on heat exposure, while the mitigation of seasonal greening and its interactions with urban typologies at the community level is still understudied. Therefore, this study aims to investigate the variation in urban heat island effect and heat exposure risk across cities, and how urban greening mitigates the adverse effects of urbanization. Using interpretable machine learning models and remote sensing data from eight different US cities in 2020, the study found that urban heat exposure risk is strongly related to the urban heat island effect and is significantly pronounced in summer and fall. There are seasonal differences in the risk reduction associated with urban greening, suggesting that it should take into account the seasonality of vegetation growth. Particular attention should be paid to green spaces and tree canopy height, as these have the best mitigation effects. Taken together, the study examines the impact of urbanization on heat exposure and its interaction with green spaces, with the aim of providing actionable insights for effective landscape strategies that can mitigate urban heat exposure.
The shape of an area is an important geographical concept that is largely defined by its degree of being compact. As a result, much attention and focus has been on the best approach(es) to measure compactness in order to assess the relative characteristics of an area. Geographers have made important contributions to the specification of measures and metrics associated with compactness. In fact, the radial deviation approach proposed by Boyce and Clark (Geogr Rev 54(4):561–572, 1964) remains particularly popular and broadly applied. This paper investigates this measure with respect to properties and theoretical relationships. A GIScience based generalization is possible that extends the radial sampling and approximation notion to a closed form exact approach. Empirical assessment explores the merits of this generalization with respect to implementation and application in GIS. The findings highlight that existing approaches do not accurately estimate average radial distance from a center to shape boundaries, creating bias and misleading compactness measurement. Further, the findings demonstrate that this can be overcome through the proposed generalization and extension relying on the actual average radial distance.
We analyze 58 stated-preference studies published between 2008 and 2023, comprising 417 estimates of willingness-to-pay (WTP) for ecosystem services provided by agricultural landscapes across Europe, Asia, Oceania, and North and South America. The analysis examines how landscape characteristics, ecosystem service attributes, valuation methods, sampled populations, and other study characteristics influence WTP. Results reveal substantial heterogeneity in preferences across regions. WTP is significantly higher when valuation scenarios are framed as avoiding environmental degradation or promoting conservation, compared to scenarios describing further improvements. In addition, WTP is scope-sensitive, with larger perceived landscape changes associated with higher WTP. In weighted specifications, additional variables show systematic patterns, pointing to a potential role of payment vehicle, sample composition, number and type of Ecosystem services (ES) valued and, landscape scarcity in explaining WTP variation.
Spatial disaggregation models may employ a cross-scale approach, namely a regression predicting values of a variable at small area level, although observations are at an aggregated region level. These constitute target and source areas in dasymetric mapping terminology. Existing cross-scale models consider single target variables only, with limited role for ancillary indicators. We consider here more general causally oriented approaches with disaggregation modelling as a constituent feature. The first involves multiple target indicators of a latent neighbourhood health risk factor to predict observed health outcomes. The second, a multiple outcome framework, considers prediction of multiple target indicators from observed health indicators. We use Bayesian inference which adapts straightforwardly to multiple likelihoods and sharing of spatial information via random effects. Two case studies are included. The first considers four adverse environmental characteristics as target indicators for 1002 neighbourhoods (substance use, sexual infection rates, violent crime and drug crime), though they are only observed for 33 London local authorities. Modelled adverse environmental measures are combined in a single neighbourhood latent construct to predict psychosis prevalence. The second case study considers impacts of neighbourhood crime on indicators of neighbourhood social cohesion, so is a social indicator application rather than health risk application. We assess impacts of crime—expected, and confirmed, to be negative—on feelings of belonging to the neighbourhood, neighbourly interactions, neighbourhood trust, and formal volunteering.
Volunteered Geographic Information (VGI) has gained popularity in urban contexts where population density supports mapping efforts. VGI projects are filling critical data gaps. One example is BikeMaps.org, a VGI platform collecting data to overcome major limitations in data on bicycling safety. BikeMaps.org has been active for over a decade and data have been used in multiple studies and have informed planning decisions in several cities. Our goal is to use BikeMaps.org as a case study for considering best approaches to developing and implementing long-term urban VGI projects and outline research needed to advance the effective use of VGI. We discuss lessons learned, including why location matters, the importance of demographic data, challenges of sampling bias, perceptions of data quality, community outreach and uptake, long-term funding, and the importance of data integration. We conclude by outlining a research agenda for urban VGI that should include research on data integration, analysis and flexible assumptions, interface usability, and the social science of equitable participation.
The structure of urban regions evolves with population growth and development, but in doing so inherent risks and dangers emerge. This is especially true in the wildland urban interface where it is not uncommon to find significant clusters of homes with limited egress, where people are forced to exit (or enter) through a restricted number of access roads. Limited points of entry/exit are bottlenecks that may create congestion, but also amplify risk and vulnerability during an hazard event, such as fire, flood, toxic release, severe weather, volcanic eruption, tsunamis, landslide, debris flow, sink hole, unexploded ordnance, improvised explosive devices, or other emergencies when evacuation is necessary. If bottlenecks are obstructed or otherwise inoperable, then neighborhood clearing is hampered. The paper introduces spatial analytics combined with geographic information systems that support assessment and planning efforts associated with hazards that may necessitate evacuation. An assessment of coastal communities in Santa Barbara County, California, is reported to demonstrate current capabilities in identifying neighborhoods that are potentially difficult to vacate and their associated bottleneck street segments that limit egress. The paper outlines future research needs to better support safety and security efforts as well as legal mandates in the assessment and mitigation of evacuation risk and vulnerability.
