Urban parks are usually discussed as amenities, environmental infrastructure or places for recreation. A new study argues that cities can also evaluate them as preventive health assets, using real-world mobility to estimate how much walking park visits generate and whether those benefits justify the cost of creating and maintaining the space.
Researchers analysed 1.79 million park visits by about 320,000 mobile-phone users across 4,139 parks in Greater Tokyo. They combined those movement records with park characteristics, residential locations, population, land prices and estimates of construction and maintenance costs. The central finding was not that every park produces an attractive financial return. Instead, the value was highly uneven, and some of the smallest neighbourhood parks emerged as the most plausible investments when the calculation was restricted to walking-related preventive health benefits.
About 16% of block parks were estimated to recover their costs within 20 years. Large destination parks rarely did so under the same narrow accounting framework. The study, published in Nature Communications on 5 October 2026, therefore challenges a simple assumption that bigger parks necessarily deliver better health value for money.
Turning actual park visits into a health investment calculation
Evidence linking green space with health is extensive, but urban planners face a harder operational question. Knowing that parks can encourage activity does not tell a city which proposed park is likely to be used, how far people will travel to reach it, or whether the resulting health benefit is large enough to justify scarce urban land and public spending.
Hang Su, Wenjing Li and colleagues approached that problem at park level. Their anonymised mobility dataset covered 1.79 million observed visits made by roughly 320,000 users. Rather than treating all green space as equally accessible, the analysis connected visits to users’ residential areas and examined how park use changed with distance. The researchers then estimated walking-related preventive health benefits and compared them with land, construction and maintenance costs, all expressed in 2019 Japanese yen.
This design matters because it links three pieces that are often analysed separately: who actually uses a park, the health value associated with the walking generated by those visits, and the economic cost of supplying the park. The resulting measure is closer to an investment screen than a general claim about the benefits of nature.
Distance mattered strongly within about seven kilometres
Across the mobility data, estimated walking-related health benefits declined as residential distance from a park increased, with the relationship most relevant within roughly seven kilometres. In practical terms, a park’s value cannot be separated from where potential users live and how likely they are to visit.
Park type changed the overall level of benefits, but distance remained central to the spatial pattern. This gives planners a more realistic alternative to simply counting hectares of green space per resident. Two parks of the same size can have very different preventive-health value if one is positioned near people who are likely to use it and the other is difficult to reach.
The result also reinforces an important distinction between access and provision. A city may technically contain substantial green space while many residents remain too far from the most useful sites to visit them regularly. Conversely, a compact neighbourhood park placed close to homes can create frequent opportunities for routine walking even if it lacks the scale or attractions of a major destination park.
Small parks produced the strongest investment case
The most striking result came from comparing estimated health benefits with costs. Returns varied sharply across individual parks and park classes. Roughly 16% of block parks were estimated to recover their land, construction and maintenance costs within a 20-year horizon when walking-related preventive health benefits were monetised.
Large destination parks, by contrast, rarely recovered their costs within that period if the calculation included only walking-related health benefits. That does not mean large parks are poor public investments. It means the specific health pathway measured in this study was generally insufficient, on its own, to offset their much larger costs.
This distinction is crucial. Large urban parks can provide biodiversity habitat, cooling, flood management, air-quality benefits, recreation, tourism, social space and cultural value. Those benefits were not comprehensively monetised in the study’s walking-focused return calculation. A large park that appears financially unattractive under this metric could still be highly valuable once its wider ecosystem and social services are considered.
Why smaller parks can perform well
Small neighbourhood parks have several structural advantages in this framework. They require less land, which is especially important in an expensive metropolitan region such as Greater Tokyo. They can also be distributed closer to residential populations, lowering the distance barrier that the mobility data showed was associated with declining benefit.
A large destination park may attract many visitors, but its land and development costs can be enormous. A small local park does not need to generate the same absolute number of visits to produce a competitive benefit-to-cost relationship. This is one reason the study’s planning estimates suggested different siting strategies for the two smallest park classes rather than a single rule for all green spaces.
The implication is not that cities should replace large parks with pocket parks. The two types serve different functions. Instead, the evidence suggests that planners should stop assuming that park size is a reliable proxy for preventive-health return.
The study used observed behaviour rather than stated intentions
One of the study’s strengths is the use of anonymised mobile-phone mobility data. Surveys can ask residents how often they visit parks or how far they would travel, but responses are vulnerable to recall error and hypothetical intentions. Large-scale location data provide a different view by capturing patterns of actual movement.
The scale is substantial: 320,000 users, 1.79 million park visits and 4,139 parks create enough variation to compare different park types and locations across one of the world’s largest metropolitan regions. Combining those records with land prices and cost data then makes it possible to move from descriptive visitation patterns to an explicit investment framework.
That framework could be useful beyond Tokyo because many cities already hold land-value, population and park-inventory data, while aggregated mobility datasets are increasingly available. The exact results would not transfer automatically, but the method provides a template for evaluating prospective sites under local costs and travel behaviour.
A useful lesson for fast-growing cities
The findings are relevant to cities where land prices and development pressure make new green space difficult to justify. In South Africa, municipalities face simultaneous demands for housing, transport, utilities, safety, recreation and climate resilience. A framework that expresses part of a park’s benefit in health terms could help make those trade-offs more transparent.
It could also change how planners think about underserved neighbourhoods. If small parks can generate meaningful walking benefits when located close to residents, green-space policy need not depend only on acquiring large tracts of land. Smaller sites embedded in daily movement networks may have a different but complementary role.
That idea fits with other Research Today coverage showing that environmental interventions often depend heavily on location and context. Research on global wet and dry patterns, for example, illustrates why broad environmental averages can hide large spatial differences. The same principle applies at city scale: the value of an intervention can change sharply depending on where it is placed.
The economic return is deliberately incomplete
The study’s narrow definition of benefit is both a strength and a limitation. Focusing on walking-related preventive health creates a transparent metric that can be compared with costs. It also excludes many reasons cities build parks.
The analysis does not attempt to capture the full value of reduced urban heat, stormwater management, biodiversity, air-quality improvements, children’s play, social interaction, mental restoration, aesthetic value or increased nearby property values. It also cannot establish that every observed park visit caused an equivalent increase in physical activity. Some visitors may have walked anyway, while others may have travelled partly by public transport, bicycle or car before entering the park.
Mobile-phone datasets have their own representativeness constraints. People included in such data may differ from residents who are not observed, and location traces cannot reveal every aspect of a visit, such as its purpose, intensity of activity or subjective benefit. The analysis therefore estimates population-level patterns rather than measuring each person’s clinical outcome.
The economic calculations are also specific to Greater Tokyo’s land market, population density, park system and 2019 cost assumptions. A city with cheaper land, lower density or different travel behaviour could produce very different payback estimates. The 16% figure for block parks should not be treated as a universal benchmark.
What the research changes
The most useful contribution is not a claim that small parks are always better. It is a way of treating parks as infrastructure whose use, location, cost and measurable health return can be evaluated together.
That approach creates a more demanding question for urban policy. Instead of asking only whether a city has enough green space, planners can ask which residents are likely to use a particular park, how distance affects that use, what health benefit follows from the resulting activity, and whether the investment is competitive with alternative sites.
In Greater Tokyo, the answer varied dramatically from park to park. A minority of small block parks could recover their costs within two decades through walking-related health benefits alone, while large destination parks generally could not. The result does not diminish the broader value of major green spaces. It shows why urban nature should be evaluated with more precision than size alone can provide.
Source Information
Study: Su, H., Li, W., Li, S., Ke, S. et al. “Quantifying parks as preventive health assets using population-scale mobility data.”
Journal: Nature Communications.
Published: 5 October 2026.
DOI: 10.1038/s41467-026-78106-7
Study design: Observational mobility and economic modelling study combining anonymised mobile-phone data on 1.79 million visits by 320,000 users with records for 4,139 parks, population, land prices and park cost estimates across Greater Tokyo.
Funding: The study reports support from multiple Chinese research and public-sector funding programmes, including the National Natural Science Foundation of China.
Competing interests: The authors declared no competing interests.
