A first-of-its-kind AI study surveyed over 20,000 parks across the world to pinpoint how parks at the periphery of cities were failing citizens.
Huge inequality between inner-city and suburban parks across the world could be threatening wellbeing globally, a new study from King's College London and Nokia Bell Labs suggests.
In a first-of-its-kind study, researchers used AI Large Language Models to study how 23,000 parks across 35 cities on five continents support wellbeing across several categories, including cultural, natural, social, and physical activity provisions. Using a novel scoring mechanism, the team found that inner-city parks in all areas of the world consistently scored higher across all categories compared to those at the periphery of cities.
This is despite lower restrictions on space in those areas, and scores were particularly low for social interaction, which, while globally low, were particularly poorly represented in suburban parks.
The suburban dream is built around this foundation of freedom of space, and the assumption would be that city planners could make advantage of the availability of space and use that to design parks of good quality. But the exact opposite is true. No matter where you live in the world, if you live at the edge of a city, your local park is much more likely to be worse than your inner-city neighbours."
Dr Linus Dietz
Dr Linus Dietz, Lecturer in Computer Science Education at King's College London and first author of the study said, "The suburban dream is built around this foundation of freedom of space, and the assumption would be that city planners could make advantage of the availability of space and use that to design parks of good quality.
"But the exact opposite is true. No matter where you live in the world, if you live at the edge of a city, your local park is much more likely to be worse than your inner-city neighbours, and because we know that good quality parks are vital for wellbeing, you're likely to feel that."
Local green spaces such as parks have continuously been linked to physical and mental health benefits, including lower mortality, with the NHS's green social prescribing programme encouraging people to engage with parks and nature as part of their healthcare.
By providing quantitative proof of the 'savannah trap', where limited funding for parks encourages designers to fill large spaces with featureless swathes of grassland, for the first time, the work highlights some of the challenges to suburbanites' health globally.
To do this, the team first mapped all features such as benches, ponds, and sports courts in each park and linked them to one of five categories known to support wellbeing (cultural significance, natural appreciation, environmental engagement, social interaction, and physical activity) using a Large Language Model. A specific statistical model was then created for each country, which looked at the density of each park's amenities and compared them against a cultural baseline to account for parks of differing sizes and generate a quality score.
The analysis is also the first to supply data-based insight on the state of urban parks, giving city workers a new understanding so they can intervene to make parks better.
The study helps us think about what it is we're doing and how and how some activity categories have been neglected."
Oliver Hayes, Parks Development Manager at the London Borough of Hackney
Historically, small teams of planners have taken infrequent surveys of the quality of parks and mapped them against pre-existing suppositions of how to improve them. Using this study, professionals can now pinpoint underperforming parks and the areas they are underperforming, laying the groundwork for more targeted interventions.
Oliver Hayes, Parks Development Manager at the London Borough of Hackney, said of the work "The study helps us think about what it is we're doing and how and how some activity categories have been neglected."
In addition to finding large disparities between parks at the centre and periphery, they also found that parks could be used to identify urban inequality hotspots, such as Copenhagen and Rio de Janeiro. These results are likely reflective of Copenhagen's approach to city planning in the past three decades, leading to gentrification and the impact of favelas and their comparative lack of funding in areas like Brazil.
