Sustainable Future for Transport and mobility (Wednesday, 19th 14:15 – 15:15)
Session Chair: Ziqi Li
Dominant Trip Purposes within a Dockless Bicycle Sharing System in London
James Todd, Shunya Kimura, Oliver O’Brien and James Cheshire
Abstract: Implementing a zero-inflated multilevel negative binomial regression model, this research identifies the relative increase in dockless bicycle sharing journey destinations within close proximity to a range of built environment factors that enable inferences of trip purposes. Results show that among users of the JUMP system in London there is nearly a fourfold increase in bicycle drop-offs in locations within close proximity to train stations, showcasing significant multi-modal first and last mile use, in addition to a significant proportion of users likely to be students using the mode to get to university buildings.
Tools for planning future Active Travel networks
Chris Larkin
Abstract: To reduce carbon emissions in the transportation sector, the UK aims to increase cycling rates for short journeys. However, the UK has a disjointed and undeveloped cycle network which is unable to facilitate this increase in its current state. This project aims to study the cycling infrastructure connectivity within the UK and internationally to determine if the UK is able to host a developed cycle network and where development would be required. This abstract presents the author’s PhD project, to be developed over the next three years and intended to provide evidence for the future of UK cycle network development.
Analysing Connected Car Data to Understand Vehicular Route Choice
Elliot Karikari, Manon Prédhumeau, Peter Baudains and Ed Manley
Abstract: This paper motivates the use of GPS traces to better understand the complexities of driver behaviour and movement, and introduces a new car trajectories dataset. This dataset, provided via the ESRC Consumer Data Research Centre, consist of connected car data for 50,000 vehicles over one month. We propose to analyse route choice through six cardinal statistical measures including travel distance, travel time, stop time, number of turns, angular deviation, and sinuosity. We report preliminary results on a 450 trips sample and aim to extend the analysis to the entire dataset to better understand how individuals navigate in their daily journeys.
Tongxin Chen, Xiaowei Gao and Tao Cheng.
Abstract: Investigating the association between human mobility and urban deprivation helps to understand the disparate routines of urban residents with different socioeconomic vulnerabilities. Though lots of research have revealed the difference in the population’s mobility behaviours impacted by social distancing measures during the COVID-19 pandemic since 2020, limited analytics focuses on the inequality in mobility recovery patterns of urban residents in the post-pandemic era. Using a large-scale geo-big data set (mobile phone GPS trajectories), we calculated the associations between the measured mobility recovery rate and urban deprivation indices (seven categories) in 4835 London communities (LSOAs) during the first four months of 2022. We show that mobility recovery is associated with urban deprivation (particularly the ‘Barriers to Housing and Services’ deprivation index) over the observed post-pandemic period. The results further demonstrate that the residents from higher deprived/vulnerable communities are likely to obtain lower mobility recovery rates in London.
Urban Sustainability (Wednesday, April 19th 14:15 – 15:15)
Session Chair: Jing Yao
Philip Home, Doreen Boyd and Sabrina Li.
Abstract: Sustainable urban cities contribute positively to the promotion of active travel. Studies have shown the increasing access to greenspace improves health, however little is known about the impacts of biodiversity on active travel. We hypothesise an increased degree of biodiversity will be positively associated with increased cycling volume. We combine several environmental, ecological, and urban form datasets to understand the association between environmental characteristics including biodiversity and recreational cycling in Newcastle. Recreational cycling data was acquired from the Strava Metro dataset in 2018. After data processing, a linear mixed-effects model will be used to determine significant associations between cycling volume and independent variables. This research will contribute to our understanding of the contribution of nature and biodiversity towards public health, pertinent due to the ongoing biodiversity crisis and growing urban populations.
Bart Breekveldt and S.M. Labib
Abstract: Urban greenspace accessibility (UGA) has positive effects on people’s health, well-being, and overall sustainability of urban areas. Existing studies inconsistently use various methods of modeling greenspace accessibility and often focus on specific city contexts. This caused an inadequate understanding of how to model UGA for any urban areas better. Here we tested and compared two widely used spatial accessibility modeling approaches, gravity and (enhanced) two-step floating catchment (E2SFCA) models and propose a reproducible and scalable methodology for modeling UGA using open data. Our results indicated that although both approaches have pros and cons, E2SFCA is better for modeling UGA.
WalkGIS: Exploring Platial Analysis of Leisure Walks via Linked Video Narratives
James Williams, James Pinchin, Adrian Hazzard, Gary Priestnall, Stefano Cavazzi and Andrea Ballatore
Abstract: Extracting rich contextual information from study participants presents an interesting challenge when the expected results are uncertain. This article presents the design of a contextual geographic information system (GIS) to extract platial information from a multimodal data set (audio, video, and GPS) collected during a `think-aloud’ leisure walking study. WalkGIS enables transcriptions, labelling, and platial analysis to be performed within one system, with data being linked and coordinated to form linked video narratives.
Jonathan Olsen, Fiona Caryl, Melody Smith, Paul McCrorie and Rich Mitchell
Abstract: There has been a recent and renewed interest in planning policies that focus on living locally, specifically concepts such as 20-minute neighbourhoods. Despite the adoption of the 20-minute neighbourhood policy in a large number of cities globally, there is surprisingly little evidence of whether having access to important facilities and amenities required for daily locally living are associated with more time spent living locally. That was the aim of this study. There was variation in the built and natural environments within 800m of children’s homes between those living in the most/least deprived and between urban/rural areas of Scotland.
Levelling up: Health and Inequality (Wednesday, April 19th 15:45 – 16:45)
Session Chair: Mingshu Wang
Levelling Up Health: A Review of UK Datasets on Place-based Measures of Health Inequality
Rob Davidson and James Cheshire.
Abstract: The UK government has committed to reducing health disparities between places in the UK by 2030 as part of its Levelling Up agenda. Spatial data can be used to better understand a place’s health, social-economic outcomes, and the relationship between them. Knowledge of what data is out there is key to track the government’s progress, as well as to inform the policy decisions that will enable that progress to happen. This paper forms the beginnings of a comprehensive resource of UK spatial datasets that can be combined, classified, and analysed in the effort to reduce geographic health inequalities.
Inequalities in spatial accessibility to care home in Scotland
Zhiqiang Feng
Abstract: This paper mapped spatial accessibility to care homes for older people in Scotland and examined its variation by urbanity and deprivation. The care home data was collected from Care Inspectorate and road networks and other data were collected from Ordnance Survey and governmental agencies. The three step float catchment areas approach was used to quantify spatial accessibility. It was found that there are significant inequalities by urban rural types and deprivation. The accessibility of care home to general practice and green space was also varied over urban and rural types.
**Spatial Microsimulation and Levelling Up: The Importance of Fine Scale Food Consumption Data
William James and Nik Lomax
Abstract: There are considerable spatial inequalities in food consumption which are associated with health issues such as obesity, cardiovascular disease and cancer. These issues need to be tackled as part of the levelling up agenda to improve health outcomes in deprived areas. Whilst traditional data sources are insufficient to capture the fine scale variability of food consumption, Spatial Microsimulation can be used to estimate consumption at the local level. In this study, we show how individual food diary data can be combined with Census data to derive fine scale estimates. This will enable the targeting of resources, funding and support.
*A data-driven approach to creating a Built Environment Context and Change Atlas (BECCA)
Jessica Hepburn, Rich Mitchell, Jenna Panter and Fiona Caryl.
Abstract: Built Environment Context and Change Atlas (BECCA) aims to produce an online map quantifying built environment change between 2016 and 2020 across Great Britain. Composite variables will be created using a range of data sources to describe what type of place an area is and how this has changed through time. Linking this data to individuals’ health in turn can highlight influences as well as direct future planning and policy. BECCA will available online on a public-facing website for researchers, the public, and private organisations to use to help support decision-making.
Levelling up: Deprivation and Vulnerability (Wednesday, April 19th 15:45 – 16:45)
Session Chair: Diego Pajarito Grajales
Where is Tweeting What about London?: Investigating Discontent in Left Behind Places
Rachael Sanderson, Rachel Franklin, Danny MacKinnon and Joe Matthews.
Abstract: The “levelling up” agenda set out by the UK government was designed to tackle entrenched regional patterns of social inequality, which have been associated with “left behind places”. These places are united through discontent resulting from being ‘left behind’ in regional development, experiencing stagnation whilst other, often metropolitan, areas experience rapid growth. Alternative datasets, including social media data, have been identified as possible ways to evaluate this discontent. This study will use a specific case study, perspectives of London from outside London, to evaluate how this dataset can be used to understand the d:ynamics of levelling up.
Creating an energy deprivation classification at small areas in England and Wales
Meixu Chen, Caitlin Robinson and Alex Singleton.
Abstract: In response to the increasing energy crisis, this study concentrated on the problems of energy deprivation in small areas of England and Wales. We create an energy deprivation classification from multiple measurements within four domains based on the extensive literature on vulnerability to energy poverty. By applying a commonly used k-means clustering method, we discover nationwide spatial disparities and distinctive characteristics of energy deprivation at fine spatial units.
Identifying areas at highest food insecurity through open data: the “Priority Places for Food Index”
Francesca Pontin, Peter Baudains, Emily Ennis and Michelle Morris.
Abstract: Open data provide the opportunity to be reactive to “shocks to the system”. This paper presents the methodology behind the construction of the “Priority Places for Food Index” and accompanying interactive geo-visualisation, in response to growing food insecurity in the UK associated with the increase in cost-of-living. The Priority Places for Food Index highlights drivers of food insecurity across seven domains around access and affordability of food at a small area geography. We highlight how ongoing engagement with policy makers and key stakeholders has driven the development of the current and planned future iterations of the index.
Abdulaziz Ibrahim, Andy Newton and James Hunter
Abstract: Previous studies have identified clusters of crime in sub-urban areas in Europe and North America. There is limited evidence of whether similar patterns exist in Sub-Saharan Africa. This paper explores the spatial patterns of crime in a typical suburban area in Abuja-Nigeria. To date, the challenge is georeferenced datasets are unavailable at small scale. This study, therefore, utilises fieldworkers to carry out a door-to-door survey to capture incidents of crime at property level (N>3000). Preliminary analysis demonstrated that, as observed in Europe-American studies, crime clusters at sub-urban context in Nigeria. Future work will explore explanatory variables for the identified patterns.
Energy (Thursday, April 20th 09:30 – 10:30)
Session Chair: Qunshan Zhao
Athanasia Apostolopoulou, Doreen Boyd, Carlos Calderon, Stefano Cavazzi, Robin Wilson and Carlos Jimenez-Bescos.
Abstract: The reconfiguration of everyday human activities is considered necessary to tackle the challenge of climate change. Building stock is one of the major contributors to urban energy consumption and targeting its energy performance through upgrade could help policy makers to reach the energy reduction goals. Given the rapidity of climate change, Urban Building Energy Modelling (UBEM) is needed. However, gathering the essential data for UBEM is a challenge, and the simplification of its process has become one-way solution. This study focuses on the Level of Generalization (LoG) of the building footprint (individual, local, district), that is needed. Preliminary results show the higher the LoG, the higher the underestimation of energy demand.
Energy Efficient Homes: The Social and Spatial Patterns of Residential Energy Efficiency in England
Bonnie Buyuklieva, Adam Dennett, Nick Bailey and Jeremy Morley
Abstract: Poor energy efficiency of homes is a major problem with urgent environmental and social implications. Housing in the UK is relies heavily on fossil fuels for energy supply and has some of the lowest energy efficiency in Europe. We explore spatial variations in energy efficiency across England using data from Energy Performance Certificates (EPCs), which cover approximately half of the residential stock (14M homes between 2008-22). We examine variations between authorities after accounting for the composition of the housing stock in terms of its ‘fixed’ characteristics of property type, building age and size. We explore variations in terms of geographical and social context (region, urban-rural and deprivation), which gives a picture of the scale of the challenge each faces. We also examine variations in relation to the more readily upgraded factors, such as glazing types, and in relation to local participation in improvement programmes which gives some insight into local actions or progress achieved.
Evaluating Access to Solar Energy in light of a Just Energy Transition
Chiem Kraaijvanger, Juliana Goncalves and Trivik Verma
Abstract: Recent growth in residential solar PV systems in cities has largely contributed to decarbonizing our energy systems. However, the costs and benefits of this transition are not always equitably distributed. Socioeconomic variability has left disadvantaged social groups unable to access the benefits provided by solar PV systems and the stimulative policy measures associated to these systems. To enhance an equitable distribution of future solar PV resources, solar energy policy will need to be more considerate of its distributional impact. This research applies a socio-spatial perspective to understand and evaluate the differences in accessibility to solar PV
systems for various social groups in the urban environment.
Cameron Ward, Caitlin Robinson, Alex Singleton and Francisco Rowe
Abstract: The UK’s residential sector has not decarbonised due to an inefficient housing stock with a reliance on natural gas. This paper uses sequence and clustering algorithms to identify four consumption trajectory groups within English and Welsh neighbourhoods between 2010-2020. These are labelled “Very High to High Consumption”; “High to Medium Consumption”; “Medium to Low Consumption” and “Low to Very Low Consumption”. The results help for the location of electrification and retrofitting policies. Targeting areas of continual-high consumption will accelerate decarbonisation. Whereas targeting areas who under consume and may be in fuel poverty will enhance the occupants well-being and finances.
Sustainability and Conservation (Thursday, April 20th 09:30 – 10:30)
Session Chair: Bruce Gittings
Advances in Landscape Connectivity Assessment for Species Conservation
Matthew Dennis, Jonny Huck and Ewan McHenry
Abstract: Metrics of functional connectivity are necessary to understand the influence of habitat loss and fragmentation on biodiversity outcomes. Effective metrics must capture, three landscape characteristics: i) habitat availability, ii) probability of movement between habitat patches and iii) habitat quality. Patch area has generally been used as a surrogate for quality such that a bias towards fewer larger patches (mirrored within conservation science) exists in connectivity work. We argue that this approach neglects species of conservation concern in highly fragmented landscapes that may persist where dispersal and habitat availability override minimum patch size requirements. We provide solutions to address this bias.
Hollie Blaydes, Emma Gardner, Duncan Whyatt, Robert Dunford, Simon Potts and Alona Armstrong.
Abstract
Solar parks could support insect pollinators in present day landscapes if located and managed appropriately. However, the role of solar parks under future land use change has not been explored. We use a GIS and pollinator model to predict bumble bee density inside solar parks and surrounding landscapes to address this knowledge gap and as part of this, require present day and future landcover maps of Great Britain. However, available future land use maps are coarser spatial resolution and include fewer categories than present day landcover maps. We therefore present the challenge of downscaling coarse resolution future maps for use with a pollinator model, using resampling and conditional overlay approaches
Kristina Wolf, Richard Dawson, Jon Mills, Phil Blythe, Jeremy Morley and Arnab Nandi
Abstract: Citizen reports and social media images that capture the aftermath of natural disasters contain important information for emergency responders. Currently, these data sources are not fully integrated into existing systems or require labour-intensive user input, which can be challenging in critical situations. In this paper, we apply computer vision services to publicly available imagery to derive meaningful information, extract objects and create text descriptions. This research builds on our previous work and enhances available hazard maps with (near) real-time weather and traffic information. Through this geospatial-based workflow, we aim to reduce climate hazard reporting friction and support operational response to incidents.
TRAP: A Novel Road-level Spatial Interpolation to Improve Estimation Errors of Air Pollution
Hyesop Shin
Abstract: Spatial interpolation techniques have been used in air pollution studies to generate area-level estimates. Despite the benefits of a mathematically sound concept, rapid implementation, and user-friendly software, interpolation suffers in areas with a low number of monitoring stations and when the built environment is ignored. The purpose of this study is to introduce TRAP, a nearly finished R package that is a new road-scale spatial interpolation method that uses road weighting. The NO2 results in Seoul showed a small variation during the summer, but large daily variations during the winter. The road-overlaid outcomes gave improved results relative to the roadside measurements.
Accessibility (Thursday, April 20th 11:00 – 12:00)
Session Chair: Jing Yao
Corin Staves, Sm Labib, Irena Itova, Rolf Moeckel, James Woodcock and Belen Zapata-Diomedi
Abstract: Accessibility is a key instrument for assessing active mobility. However, accessibility measures often suffer from biases due to spatial aggregation, isochrones with arbitrary cut-offs, and distance-based cost functions that ignore the route conditions. Previous literature has addressed these issues individually, but not holistically. This study applies the MATSim framework to efficiently route between billions of origin-destination pairs to calculate fully disaggregate Hansen accessibilities using cost functions sensitive to network quality. With examples of greenspace and foodstore accessibility, we demonstrate the potential of this method for providing policy-relevant insight into the suitability of the built environment for active travel.
Using Machine Learning to Predict Perceptions of a Motorbike Ban in Hanoi
Minh Kieu, Alexis Comber, Eric Wanjau, Kristina Bratkova, Thi Thuy Hang Nguyen, Thanh Bui Quang, Huu Phe Hoang and Nick Malleson
Abstract: The dependence on motorbikes has contributed to severe traffic problems in Hanoi, Vietnam. Policymakers have considered a controversial ban on non-electric motorbikes in parts of the city in an effort to reduce congestion and pollution. However, understanding of individual perceptions on critical transport policies, such as this potential ban, is lacking. This paper applies a machine learning algorithm (XGBoost) to a bespoke travel survey to better understand how residents perceive a potential motorbike ban and how their perceptions might change under different policy scenarios. Our results suggest that prior awareness of the ban and shorter distances to public transport both increase peoples’ favour.
Clara Peiret-García, Rachel Franklin, Alistair Ford and Joe Matthews
Abstract: Providing equitable and more sustainable access to basic amenities is key to cutting carbon emissions and increasing social equity in cities. This paper applies a data-driven approach to generate an activity-based accessibility index for British cities. To do so, we employ a two-step approach. Using a self-organising map, we first generate behavioural profiles based on the British Time Use Survey. We then use the resulting clusters to determine the weight of the different amenities in our accessibility index. Our preliminary results seem to provide a promising avenue for applying data-driven methods to generate weighted accessibility scores. Future work will incorporate Census 2021 data in order to better characterise neighbourhoods in terms of both accessibility and their demographic characteristics.
Louise Sieg and James Cheshire
Abstract: This paper outlines the development of a classification for a new set of optimised spatial units for the aggregation of mobile phone in-app data. We create regions which aim to strike balance between disclosure control and data granularity, and attribute to each region a time classification reflecting the region’s dominant time of mobile phone activity. We aim to provide a methodology for the making of bespoke geographical regions and classifications for the use and dissemination of sensitive granular data in research.
Spatial Statistics (Thursday, April 20th 11:00 – 12:00)
Session Chair: Guy Solomon
René Westerholt
Abstract: Geospatial hotspot statistics measure concentration in space. One of the most popular such methods is the Getis-Ord statistic, a measure for revealing geographical structure in the tails of an attribute value distribution. The original 1995 paper introducing this measure includes an interpretation of how the variance of the spatial weights employed feeds into the local measures, but the corresponding mathematical expression can be shown to be erroneous. This paper corrects this expression and discusses the implications of the revised term for the link between the Getis-Ord statistic and the local variances of the spatial weights.
Geographically Weighted Cronbach’s Alpha
Sui Zhang and Ziqi Li
Abstract: As individual’s cognitions and behaviours are affected by where they live, the reliability of responses to tests or scales may vary with location. In this paper, we develop a local version of Cronbach’s alpha, geographically weighted Cronbach’s alpha, to investigate how the reliability of the measure varies spatially. Two demonstrations of exploratory applications of the three-wave geocoded measure of “social capital index” in the Baltimore metropolitan area, USA, was also performed in this paper.
Alexis Comber, Paul Harris and Chris Brunsdon.
Abstract: This paper proposes a novel spatially varying coefficient model for spatial regression using General Additive Models (GAMs) with Gaussian Process (GP) splines parameterised with observation locations. The brand leader in this area is probably Multiscale GWR (MGWR) models but these have a number of theoretical and technical limitations. Here, a GAM with GP spline model and a MGWR model were applied to simulated spatial datasets with varying degrees of spatial autocorrelation. The GAM was shown to perform better than MGWR under a range of fit metrics. Some unresolved issues are discussed such as model calibration or tuning of knots and spline smoothing parameters.
Integrated urban and transport modelling: the HARMONY Strategic Simulator – Oxfordshire application
Fulvio D. Lopane, Eleni Kalantzi, Tianqu Shao and Michael Batty.
Abstract: The HARMONY Model Suite is an integrated set of models aimed at capturing the dynamics and mobility patterns of metropolitan areas; it is structured over three different levels: 1) Strategic (long-term) demographic land-use transport models, 2) Tactical (mid-term) people and freight activity-based models and 3) Operational (short-term) multimodal network models. This paper presents the Oxfordshire application of the HARMONY Strategic Simulator: a New Housing Development (2030) scenario. This application involves a demographic forecasting model coupled with a regional economy model, both informing a land-use transport-interaction model and a residential land development model in the HARMONY web-based modelling platform.
Socioeconomics and Geodemographics (Thursday, April 20th 14:15 – 15:15)
Session Chair: Petrus Gerrits
Exploring the effects of socioeconomic factors on voter preferences: A case-study of France 2022
Niloufar Pourshir Sefidi and Peter Mooney
Abstract: Understanding the reasons behind voter preferences in elections is time-consuming and resource intensive using traditional methods such as surveys. In this paper, we describe a flexible statistical modelling approach to investigate the effect of different socioeconomic factors on voters’ preferences for the 2022 French presidential election. Our results show that socioeconomic factors have a significant effect on the voters’ preferences for the French presidential election in 2022. The methodology developed can be used to investigate these effects for elections in other countries where appropriate socioeconomic and spatial data is available.
Intersecting the location and the geodemographic context of museums at the national scale
Andrea Ballatore, Stefano De Sabbata, Jamie Larkin and Fiona Candlin
Abstract: The geography of the cultural sector concerns the location of producers, consumers, and venues of the Cultural and Creative Industries (CCIs) to answer questions about their
development and dynamics. Considering the case of the UK museums, we use national data from the Mapping Museums project to study their geodemographic context at the Local Authority District level. Across the UK, we observe the distribution of different types of museums according to their governance type and subject matter in areas belonging to different geodemographic categories. Both in terms of simple counts and divergence from expected values, different types of museums show vastly different distributions and trends, reflecting the variety of the sector.
Fangzhou Zhou, Tao Cheng and Xianghui Zhang
Abstract: This study attempted to explore the place of origin of people who paid their respects to Queen Elizabeth II. Using Mobile Apps GPS dataset, we went backwards to the starting point of peoples’ trips who came to queue in the site, then origin information based on the local authority districts level of these people was found. The results are visualized on the UK map. According to the results, we can know people’s emotional intensity for the Queen in different regions.
Hosting major events: understanding the impacts of short-term rentals in Glasgow during COP26
Yang Wang, Nick Bailey, David McArthur and Mark Livingston
Abstract: Existing research analysing properties available for short-term rental on Airbnb focuses on long-term impacts. There is still a gap to examine abrupt market growth due to mega events like COP26 in Glasgow. By employing Airbnb daily scrapings, this paper summarized spatiotemporal trends and patterns of new Airbnb listings and their impact on neighbourhoods with different social
characteristics. We found growth in both deprived and affluent areas during the event. The growth in
deprived neighbourhoods, with a smaller Airbnb market in 2020, is consistent. The benefit of
responding to the surge in demand, however, is more enjoyed by affluent neighbourhoods.
Urban Structure and Form (Thursday, April 20th 14:15 – 15:15)
Session Chair: Kristina Wolf
**Inferring characteristics of urban structure through the variability in human mobility patterns
Carmen Cabrera-Arnau, Chen Zhong, Michael Batty, Ricardo Silva and Soong Kang.
Abstract: The polycentric city model has gained popularity in spatial planning policy, since it is believed to overcome some of the problems often present in monocentric metropolises, ranging from congestion to difficult accessibility to jobs and services. However, the concept ‘polycentric city’ has a fuzzy definition and as a result, the extent to which a city is polycentric cannot be easily determined. Here, we leverage the fine spatio-temporal resolution of smart travel card data from multiple cities to infer urban polycentricity by examining how a city departs from a well-defined monocentric model.
Ki Tong and Alessia Calafiore
Abstract: Noise pollution is problematic in highly urbanised areas, impacting adversely on active commuting experiences. The potential rippling effects of noise pollution on transportation choices inform the study’s intent. Using London as a case, the study explored the association between noise pollution, urban forms and functions and the mode of transportation to work. Results show cyclists are more vulnerable to high noise levels, whereas pedestrians are more exposed to mid-range noise levels, with urban density and diversity positively correlated with noise pollution.
Input-output analytics for urban systems: explorations in policy and planning
Griffith Rees, Guy Solomon, Bowen Zhang and Alan Wilson
Abstract: We develop an inter-region input-output economic model of future challenges and appropriate policies. We initially focus on the UK—developing an analytic tool for the ‘levelling up’ agenda: to examine the characteristics of economically ‘good’ versus ‘poor’ cities and to explore policies aimed to shift cities from poor to good. Our method has potential application to equivalent challenges at varying scales and in other countries. The analysis is distinctive: estimate 48 city input-output accounts and the trading relationships between them via a novel incorporation of travel costs. Temporal extensions facilitate a scalable method to model future scenarios of policy outcomes.
Wataru Morioka, Atsuyuki Okabe and Mei-Po Kwan
Abstract: The objective of this study is to understand the spatial impact of the pandemic on various kinds of food retailers with a focus on population change. To achieve this goal, we examined the spatial similarity between the distribution of food retail outlets and that of the space-time population before and during the COVID-19 pandemic in a central area of Yokohama, Japan. The similarity was measured by using the kernel density estimation on a street network and the Kullback–Leibler divergence. The results showed that the impact of business closures varied by the type of retail outlets.
Small Areas Statistics (Thursday, April 20h 15:45 – 16:45)
Session Chair: Craig Robson
Yannick Oswald, Nicolas Malleson and Keiran Suchak.
Abstract: Global problems, such as pandemics and climate change, require rapid international
coordination. One notable example was the imposition of ‘lockdown’ policies in response to the COVID-19 pandemic in early 2020. Here we build an agent-based model of this rapid policy diffusion, where countries constitute agents and with the mechanism for diffusion being peer mimicry. We utilize data assimilation to constrain the model against observations in ‘real-time’. We find that the model is able to predict the policy diffusion relatively well and that the data assimilation improves the fit to the data.
A pedestrian evacuation ABM in a complex environment based on Bayesian Nash Equilibrium
Yiyu Wang, Jiaqi Ge and Alexis Comber
Abstract: This research proposes an updated evacuation model that incorporates Bayesian Nash Equilibrium (BNE) within a multi-agent system. It augments the rationality of pedestrian decision-making processes and improves the evacuating behaviours of pedestrians. Recent research found that BNE pedestrians were able to evacuate more quickly by avoiding clogged areas, closely matching the pedestrian evacuating behaviours in reality. This paper extends this work by introducing a number of physical barriers in the model in order to evaluate whether and how BNE affects the pedestrian evacuation process in more complex scenarios. It provides a detailed introduction of the updated simulation model and identifies several potential research directions.
Towards Inclusive Urban Planning: An Elderly-Focused Agent-Based Model of West Midlands
Manon Prédhumeau and Ed Manley
Abstract: The UK population is ageing. This demographic trend is challenging for existing infrastructure and services because older people have specific daily activities and mobility patterns. Moreover, the older population is highly heterogeneous regarding demographics, health conditions and transportation access. While agent-based models are increasingly used in urban planning, they generally ignore this heterogeneity in elderly populations. This paper presents an agent-based model to support urban planning related to seniors in the West Midlands county, UK. The model is built and validated using mainly open data. Its potential is illustrated on the example of elderly healthcare accessibility.
Modelling current and future Lyme Disease risk in urban greenspaces: an agent-based approach
Kirsty Watkinson and Jonathan Huck
Abstract: Green infrastructure (GI) is an important promoter of urban biodiversity. However, improper GI planning can lead to negative outcomes (disservices) such as the increased transmission risk of Lyme Disease (LD) to humans. We describe an agent-based approach to model LD risk under current and future GI scenarios. We model the interaction of humans, ticks, and deer within GI in order to estimate LD risk to humans under different landscape scenarios. This will provide new insight into LD risk and enable policy makers to better understand how to plan GI to maximise its benefits to people and the planet.
GeoData Types (Friday, April 21st 09:30 – 10:30)
Session Chair: David Forrest
Exploring the evolution of GIS research using bibliographic data
Robert Berry, Caitlin Hafferty, Scott Orford and Lucy Clarke
Abstract: This paper provides new insight into the evolution of geographical information science and systems (GIS) research via a computational analysis (in R) of over 100,000 bibliographic records (from 1970 to 2021) downloaded from Scopus. We conduct an exploratory analysis of the data, then attempt to discover the thematic/topical structure of the GIS literature using the Structural Topic Model (STM) framework. We show how topics in GIS have evolved and discuss how our findings contribute to the understanding of the evolution and trajectory of GIS research. We conclude by highlighting the limitations of the approach and explaining our future research plans.
Residential Telephone Directories and Geodemographic Change in Britain
Nikki Tanu, Maurizio Gibin and Paul Longley.
Abstract: Historical British telephone directories have recently availed themselves as disaggregated sources of population data for time periods not covered by other popular sources. Having devised a comprehensive processing pipeline to prepare these data for further analyses, this paper explores what they could uncover about British historical demography. For one, the differential uptake of telephones across space and time, owing to changes to its affordability and applications, reflects socioeconomic disparities. Being a data source that spans decades, it also lends its application to the study of intergenerational inequalities through the unit of family groups, as indicated by surnames that subscribers bear.
Exploring Historical Links between Scotland and India using Geoparsing
Chandramauli Tyagi and Bruce Gittings.
Abstract: A significant amount of spatial information can be derived from unstructured datasets available in web pages, e-books, and digital archives. Geoparsing is one such concept that is very useful in extracting spatial data from any unstructured text source. Geoparsing complemented with Natural Language Processing algorithms can effectively automate this process of identifying and geo-tagging the extracted spatial data. The research illustrates the power of geoparsing by extracting place names from a corpus of biographies of famous Scots who travelled to India from the 18th to the early 20th Century to give an impression of the spread of the Scottish diaspora at that time.
A reproducible approach to generating synthetic spatial data for teaching and learning purposes
Paddy Gorry and Peter Mooney
Abstract: Whilst there is an ever increasing amount of openly available spatial-data for teaching and learning purposes teachers and students often wish to generate their own synthetic spatial datasets. Such datasets can be used to test algorithms, test software code, assist in visualisation development, and be used as unseen datasets for student assessment and learning. We describe the development of a Python-based software tool called RADIAN ((RAnDom spatIal dAta geNerator)) for generating simple synthetic spatial datasets.
GIS and Vulnerability (Friday, April 21st 09:30 – 10:30)
Session Chair: Sabrina Li
Gendering the research pipeline: A quantitative feminist geographical approach
Laura Sheppard and Jon Reades
Abstract: Women and gender minorities are underrepresented in positions of seniority in higher education (HE). Research on gender in HE varies in scale and methodology from large global surveys to smaller projects, with a noticeable gap of the experiences of ECRs and doctoral students, and how they vary from discipline-to-discipline, institution-to-institution, and department-to-department. We propose a new agenda by drawing together quantitative and feminist geography to focus on different ‘platial’ scales within HE, such as expanding the definition of gender and considering intersectionality, using open-source data, and using quantitative geography methods to develop a multi-scalar understanding of these dynamics.
Enhancing Urban Design Through Geodata and Machine Learning
Alexander Mackay and Diego Pajarito Grajales.
Abstract: Machine learning (ML) methods have seen surprisingly little application in the innovation-driven field of urban design. Research by Boim, Dortheimer, and Sprecher (2022) presented a novel use of ML to generate alternative urban plans considerate of existing local practices, using data extracted from a GIS package to train a Conditional Generative Adversarial Network (CGAN) model. This paper extends that work with a novel dataset built from open geospatial data from Glasgow with results validated using additional quantitative, and qualitative validation methods. The results show the CGAN model is capable of producing geographic and contextually sympathetic urban design proposals with output quality confirmed by result validation.
Andreas Keler, Sasan Amini, Johannes Lindner and Klaus Bogenberger
Abstract: The Digital Twin Munich project (DZ-M) aims to depict complex urban environments through the use of static and dynamic components, and their semantic relationships. The project focuses on the development of a street network model and urban mobility simulation, utilizing the open source microscopic traffic flow simulation software SUMO. The transport demand is provided by the VISUM model of the city of Munich, and the data structure developed is compatible with standards such as OpenStreetMap, OpenDrive, CityGML, and GTFS. The project also includes the use of physical VRU simulators for data collection purposes, and the integration of these simulations into a 3D VR environment in Unity.
Assessing the levels of loneliness and satisfaction with accommodation among older adults in England
Andrea Nasuto, Richard Dunning and Les Dolega
Abstract: The population of England is rapidly aging and policymakers face a challenge in providing suitable housing for older adults to “age in place.” Two major areas of concern are housing and loneliness. There is a lack of understanding of differences in housing satisfaction and loneliness among older adults, as well as their spatial distribution. This research provides estimates of housing satisfaction and loneliness in the population aged 50 and older at the LSOA level in England, and identifies the key drivers of these phenomena through the use of small-area estimation methods.
Comments are closed