Kirsty Watkinson

Dr. Kirsty Watkinson


Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
2 January 2023 in Cartography and Geographic Information Science
Verified
Journal Article
Closed access
KWKirsty Watkinson
Kirsty Watkinson
JHJonathan Huck
Jonathan Huck
AHAngela Harris
Angela Harris
Volunteered geographic information (VGI) offers a solution to inequalities in authoritative map data that can limit our response to humanitarian crises. However, sustaining voluntary contributions of map data can be difficult and hybrid machine learning-VGI (ML-VGI) workflows developed to encourage sustained volunteer contributions have been demonstrated to be insufficient. Gamification can be used to encourage volunteers to map for longer, however evaluations of gamification to increase humanitarian mapping contributions are rare. Here we develop a gamified humanitarian ML-VGI mapping platform (“Map Safari”) and evaluate the use of game elements to encourage sustained volunteer contributions without reducing contribution quality. Our results suggest that gamification makes mapping more fun, particularly for first time mappers, without degrading map data quality. Competition is demonstrated to be important for encouraging enjoyment of game elements and increasing map data contributions. Future gamified mapping platforms should emphasize competition and ensure there are enough game elements to make platform use feel game-like. This research demonstrates that gamification can be used to encourage continued voluntary contributions of map data thereby increasing the amount of map data available to humanitarian organizations.

13 June 2022 in Annals of the American Association of Geographers
Verified
Journal Article
Closed access
KWKirsty Watkinson
Kirsty Watkinson
JHJonathan Huck
Jonathan Huck
AHAngela Harris
Angela Harris
Volunteered geographic information (VGI) is often cited as a potential solution to persistent global inequalities in map data, particularly in areas undergoing humanitarian crises. Poor volunteer engagement, slow data production, and low-quality outputs have limited progress, however, and can unintentionally exaggerate inequalities. Hybrid machine learning–VGI (ML–VGI) frameworks can help to overcome these challenges through a combination of workflow automation and purposive human input, but the use of these workflows is rare in practice. Here, we implement an ML–VGI framework (Centaur VGI) and undertake a detailed comparative usability assessment against an existing, widely used VGI mapping platform to demonstrate its potential to improve volunteer engagement, mapping speed, and data quality. Our results suggest that through automated building, searching, and labeling, the Centaur VGI platform provides greater usability, quicker data production, and improved data quality for most users. Consequently, we provide the first evidence that hybrid ML–VGI approaches can be used to facilitate increased public participation in humanitarian building mapping efforts and thus help reduce global inequalities in map data.

23 September 2025 in Nature Cities
Verified
Journal Article
Open access
PDF file
SGSara L. Gandy
Sara L. Gandy
JHJessica L. Hall
Jessica L. Hall
GPGrace Plahe
Grace Plahe
KWKirsty Watkinson
Kirsty Watkinson
DJDavid Johnson
David Johnson
LGLucy Gilbert
Lucy Gilbert
Within- and between-city contexts and interactions shape our experiences of city life. However, a gap in understanding is how the wider landscape context of cities and the interactions with hinterlands influence urban ecology and health hazards. Using a meta-ecosystem framework, we fill this gap for the tick-borne Lyme disease ecological system by comparing the tick and Lyme disease hazards of urban and hinterland sites for 16 UK cities. We discover that the environmental hazards of ticks and Lyme disease of urban greenspaces are two- and threefold lower, respectively, than those of hinterland woodlands. Crucially, urban tick and Lyme disease hazards are shaped by tick abundance and the landcover (woodland and built-up) of hinterlands, but not of cities themselves. This highlights how rural–urban interactions form meta-ecosystems, and urban eco-epidemiology can depend on the characteristics of the surrounding rural landscape. Therefore, to better understand urban ecological processes and to mitigate disease risk in cities, it may be necessary to consider environmental factors in the hinterland such as landcover and disease hazard outside cities.

15 October 2025 in Journal of Environmental Management
Verified
Journal Article
Open access
PHPeng Huang
Peng Huang
YSYue Sun
Yue Sun
KWKirsty Watkinson
Kirsty Watkinson
ACAbigail Coole
Abigail Coole
RBRichard D. Bardgett
Richard D. Bardgett
DJDavid Johnson
David Johnson
Green spaces have potential to contribute substantially to carbon (C) capture and storage in urban environments. Yet, the impact of urbanisation and land cover change on key pools of C and nitrogen (N) in soils and soil organic matter (SOM) fractions is poorly resolved, which limits the ability to optimise management of urban environments. Here, we investigated the associations among urbanness and land cover change with C and N pools in bulk soil and SOM fractions in green spaces. Soils were sampled along an urbanness gradient from green spaces that contained three land cover classes (woodland, grassland and a transition area between each) across the metropolitan borough of Greater Manchester, northern England. We found that soil C and N concentrations in green space generally increased from rural to urban areas and from woodland to grassland. Plant species diversity plays a key role in connection between urbanness and soil C and N sequestration. We also found a significant interaction between land cover types and urbanness, which positively influenced soil C and N storage. Green space in the most urbanised areas contained higher mass percentage of particulate organic matter (POM) fractions compared to green space in the least urbanised areas, which indicated that urban soil is much more labile. However, the concentrations of C and N in these two fractions was also interactively mediated by the urbanness and land cover types. These findings demonstrate that urbanness and land cover types have significant effects on key pools of soil C and N and emphasise the importance of considering these pools for future urban planning and design amidst ongoing urbanisations to meet the target of carbon neutrality and sustainable living environment.

21 December 2023 in Journal of Spatial Information Science
Verified
Journal Article
Open access
PDF file
KWKirsty Watkinson
Kirsty Watkinson
JHJonathan Huck
Jonathan Huck
AHAngela Harris
Angela Harris
As a consequence of their reliance on a scarce volunteer resource, humanitarian mapping organizations must prioritize their mapping activities. For mapping in anticipation of a crisis or mapping in support of long-term crises, the only method available to organizations is an estimation of the "completeness" of the map, with organizations directing volunteers to map areas where data are missing. Whilst this method is suitable for organizations that focus on general map improvement, for those who create data for a specific reason (e.g., drinking water provision) the method is sub-optimal. In this article, we present a new method of humanitarian mapping prioritization, that considers the purpose of map data collection. The method identifies locations where contributions by volunteers are expected to have the biggest impact on the desired use of the map data and therefore maximizes the value gained from volunteer contributions. We explain our method using the example of measuring distance to healthcare and demonstrate its superior ability to consider the context of map data over generic estimations of map "completeness". Our method provides humanitarian mapping organizations with an easily reproducible and low cost method and an opportunity to make better informed decisions about mapping prioritization, when the purpose of map data collection is known. Using our method, organizations will be able to maximize the value gained from a scarce volunteer resource and increase the efficiency of humanitarian map data production.

21 May 2026 in medRxiv
Verified
Preprint
Open access
PDF file
SGSara L. Gandy
Sara L. Gandy
GPGrace Plahe
Grace Plahe
JHJessica Hall
Jessica Hall
KWKirsty Watkinson
Kirsty Watkinson
SGSwapna Guntupalli
Swapna Guntupalli
LGLucy Gilbert
Lucy Gilbert
Introduction: Socioeconomic deprivation is often associated with poorer health outcomes, but some studies suggest the opposite for Lyme disease. Here we test two hypotheses to explain this: differences in (i) local landcover of high risk habitats such as woodlands (landscape hypothesis) and (ii) outdoor recreation in such habitats (behaviour hypothesis). Methods: We analysed reported Lyme disease incidence data for 824 data zones in the city of Glasgow, UK, against deprivation rank (based on indicators relating to income, employment, health, education, crime and housing). We then tested how these relate to woodland cover and indices of urban greenspace usage (per capita and per ha of greenspace). Additionally, we measured Lyme disease hazard (density of infected ticks) in 32 greenspaces and tested relationships with deprivation, woodland and greenspace usage. Results: More advantaged data zones (data zones with low deprivation rank) had higher Lyme disease incidence. These areas had more woodland and woodland cover was positively correlated with both Lyme disease incidence and hazard. Deprivation did not correlate with greenspace usage, nor did greenspace usage correlate with Lyme disease incidence. Intensely used greenspaces had lower infected tick densities, consistent with a human disturbance effect on wildlife that carry ticks. Conclusions: Differences in woodland cover, but not outdoor recreation behaviour, can help explain our finding of higher Lyme disease incidence in more advantaged areas. However, to further test the behaviour hypothesis, we need more detailed data on outdoor recreation activity per capita both locally and in rural areas, as well data on mitigation behaviours.