Open spatial science as routine practice: Urban Data/Code contributions from GISRUK 2024

Introduction

Welcome to this (first) Special Issue of Environment and Planning B (EP-B) devoted to Urban Data/Code contributions. The issue presents a curated set of articles arising from the 32nd GIS Research UK (GISRUK) conference, ¹ held at the University of Leeds (UK) in April 2024. Since 1993, GISRUK has hosted an annual conference, usually drawing over 200 researchers from the United Kingdom, Europe and beyond. Each conference showcases research that contributes new theories, methods and insights across a range of spatial analysis problems and applications. GISRUK is committed to supporting Early Career Researchers (ECR) through numerous workshops, career development opportunities and ECR-led events. This commitment is even codified into its constitution. The national committee retains a dedicated ECR representative and its chair must not have attained full professor status.

A characteristic of most GISRUK submissions, even commentaries and perspective pieces (cf. Chris Brunsdon’s 2023 keynote on Bayes, Ulam and Missingness), is that they are underpinned by a new and interesting dataset and/or computational technique. More often than not, this ends up finding its way into some sort of computer code, and many GISRUK authors publish their code and data contributions via open platforms (Figshare, Dataverse, GitHub, CRAN, PyPI ). While the imperative for open science practice is generally taken-for-granted in the GISRUK community, the ‘hard’ rewards for this endeavour have historically been in short supply (Arribas-Bel et al., 2021). At the 2024 GISRUK conference, however, a small grant award part-funded by OSGeo:UK (OSGeo, 2026) was initiated for the first time to support the development of open source tools from standout GISRUK papers.

Given this background, EP-B seemed like an ideal choice when thinking about possible target journals for a special issue resulting from the Leeds conference. The Urban Data/Code section provides a dedicated venue for researchers to receive credit for their data and code contributions (Arribas-Bel et al., 2021). The section publishes descriptions of two types of openly released artefacts: open data products – datasets and surrounding services such as APIs, web maps or dashboards; and software that performs analysis operations in a reproducible way.

The six articles that feature in this special issue are a perfect illustration of the diversity and quality of data-driven GIScience at the GISRUK conference. Below we group the articles in three main themes and briefly describe each of them, focusing on their relevance to open science practice. As is required by Urban Data/Code submissions, each paper is a concise description of the artefact, accompanied by code repositories with further detail for the interested reader. All published articles are available as Free access by default.

Fundamental GIS research

Three articles might be classed as purely methods-based, or at least are dedicated to spatial analytic operations particular to GIScience.

Huck (2026) describes a method for partitioning complex polygons into irregular grids of equal area rectangles. This class of subdivision is useful in instances where the desired units of analysis are smaller than the available small-area geographies. The algorithm for determining how the subdivisions are effected is adapted from a Python library (Pmav99, 2014) and is made available in the fully open source QGIS via a plugin. Since the release of this QGIS plugin, Huck (2026) notes that a similar tool has been added to ArcGIS Pro, but with little documentation and no explanation of the underlying algorithm (ESRI, 2024).

Beecham et al. (2026) article presents an R package for automatically generating gridmap layouts – small multiple data graphics in which each item is a spatial unit of regular size, laid out with an approximate geographic arrangement for use in multivariate data visualisations (e.g. Beecham et al., 2021). The package uses linear programming and the open source ompr library (Schumacher, 2023) for constraining allocated layouts on distance distortion and handling parameters designed to preserve key geographic features in the abstracted layouts.

Presented in Deakin et al. (2026) are two approaches to simplifying road network geometries – reducing multiple parallel roads into a single representative transport corridor. Simplifying networks in this way eliminates visual clutter, aiding visual data analysis; the reduced file size also greatly eases the process of deploying network data in web-based interactive tools. Both approaches are implemented using the Parenx Python library (Deakin, 2024), and the authors compare theirs with another recent network simplification approach (Fleischmann et al., 2026), again released open source with accompanying code demonstrations.

Data-driven approaches to mobility science

Increasingly prominent amongst applied research at GISRUK are studies that combine open data and methods to quantify the viability and fairness of alternative transport modes to the car.

Credit et al. (2026) article presents a method for approximating urban ‘walkable’ accessibility to key services and amenities for individual census blocks across the United States. The approach combines open data with traditional gravity models and achieves scores very similar to those published as a proprietary data product (Walk Score, 2011). Comparison of these two sets of scores enables some speculative insights into how the ‘closed’, proprietary Walk Score estimates might have been generated, and the relative importance of certain parameters and arbitrary weightings. The authors also point to exciting future directions in validating scores against large-scale observational data.

A challenge, when quantitatively evaluating transport planning interventions that better deliver non-car alternatives, is how to represent fairness. Is a fair transport system one that maximises spatial equality of access? That maximises utility – accessibility for the greatest number of people? Or one that maximises benefits for the most vulnerable? Nelson et al. (2026) present an open source package that evaluates public transport provision on these different definitions, with an illustrative case study of Cape Town, South Africa. In directly incorporating theories of spatial justice into quantitative accessibility measures, the work is ambitious and distinctive; the early paper and presentation won the CASA best paper prize for Spatial Analysis, ² at the conference.

Open data products

The final paper linked to this special issue (Morgan, 2026) provides a characteristic example of the kinds of data products generated through GISRUK research. Morgan (2026) describes a project in which an array of observational, derived and modelled datasets are combined to estimate neighbourhood-level carbon footprints in Great Britain. In addition to the neighbourhood-level footprint estimates, the authors provide a web-based front-end that supports easy exploration and bulk download.

Conclusion

These six articles present only a subset of the 50+ research papers presented at GISRUK 2024. While each article makes a substantive methodological or empirical contribution, that they follow open science protocols is reassuringly routine amongst GISRUK contributions. Worth noting, however, is that not all open science practices are enthusiastically adopted. As is the case across quantitative geography, it is rare to find GISRUK papers that document analysis protocols in advance as pre-registered reports (Nosek et al., 2018). This may be for good reason: many applied spatial analysis studies involve repurposing data to investigate spatial processes that are often fluidly defined, where there is a high level of analytic uncertainty (Beecham and Lovelace, 2023) and where standard inferential statistics cannot be easily applied (Goodchild and Li, 2021). It will be interesting to observe whether this changes. At a time when data and modelling workflows are becoming increasingly automated, mechanisms for formalising research designs or at least articulating intentions with recourse to evidence-backed theory (Wolf, 2023) may assume greater importance.

We are grateful to the authors for sharing their excellent open science contributions, to the GISRUK community for championing open science practices, and as ever to EP-B’s reviewers for supporting authors in expressing their work in a clear, rigorous and accessible way. Finally, we would like to thank EP-B for initiating the Urban Data/Code section and providing a mechanism for promoting the kind of data-driven activity that is foundational to contemporary urban analytics practice.