Satellite Thermal Data Reveal the Exact Park Size That Cools Cities Best

Cities around the world are heating up faster than ever, and one of the most relied-upon defenses against urban heat is also one of the oldest: the humble park. But how much do urban parks actually cool their surroundings, when do they cool them most effectively, and how big does a park need to be to deliver the best return on scarce urban land? A new study published in Theoretical and Applied Climatology offers some of the most precise answers yet, drawing on high-resolution thermal measurements from an instrument mounted on the International Space Station to track park cooling through an entire year of changing seasons.

The research team, led by Shilong Yao and Lili Lin of Minnan Normal University in China, together with Zhenbang Hao of Zhangzhou Institute of Technology and Christopher J. Post and Elena A. Mikhailova of Clemson University, focused their attention on Zhangzhou, a prefecture-level city in Fujian Province in southeastern China. That choice matters. Most park cooling studies to date have concentrated on massive metropolitan centers such as Beijing, Hangzhou, or Los Angeles, leaving small and mid-sized cities—where the majority of the world’s urban population actually lives—poorly characterized. Previous assessments also tended to treat park cooling as a static phenomenon, capturing a snapshot during a single heat wave rather than the full seasonal and daily rhythm of how green space interacts with the urban thermal environment.

To fill that gap, the team turned to ECOSTRESS, the ECOsystem Spaceborne Thermal Radiometer Experiment on Space Station. ECOSTRESS measures land surface temperature—the radiative temperature of the ground itself rather than the air a few meters above it—at a spatial resolution fine enough to distinguish individual parks from the concrete and asphalt around them. Because it flies on the International Space Station rather than a traditional sun-synchronous orbit, it samples different locations at varying times of day and night, giving researchers a rare window into the diurnal cycle of urban heat. In total, the team assembled 23 observation periods spanning all four seasons, calculating land surface temperature inside each park and in the surrounding buffers to quantify how far and how strongly the cool air effect extends outward.

The findings reveal a striking seasonal architecture to park cooling. Across all parks, the mean cooling effect was strongest in spring, when park surfaces averaged 0.46 degrees Celsius cooler than their surroundings, and more than 60 percent of all daytime observations recorded lower temperatures inside parks than outside them. That springtime dominance carried through to the spatial footprint of the effect: the park cooling area, the park cooling intensity, and the park cooling gradient—a measure of how steeply temperature rises as you move away from a park’s edge—all reached their highest values in spring. In practical terms, parks in Zhangzhou both cool their interiors more deeply and spread that cooling farther into surrounding neighborhoods during spring than at any other time of year.

Summer told a different story, and one with major implications for heat planning. While the raw cooling footprint shrank compared with spring, the park cooling efficiency—a metric that captures how much cooling a park delivers per unit of its own area—peaked in summer, exactly when urban heat stress is most dangerous. In other words, during the season when cooling matters most for human health, parks operate at their highest thermodynamic efficiency, wringing the greatest benefit from every hectare of green space even as their absolute cooling intensity moderates. This decoupling of capacity from efficiency across seasons is precisely the kind of dynamic that static, single-season studies have missed.

The seasonal pattern was not just a matter of vegetation cycling through growth and dormancy. When the researchers ranked the environmental drivers of each cooling metric, they found that the factors dominating in spring and summer were fundamentally different from those that mattered in autumn and winter. During the warm seasons, vegetation cover and water bodies within and around the parks were the primary forces behind cooling intensity and cooling gradient, reflecting the powerful combination of evapotranspiration—the process by which plants pump water vapor into the atmosphere, carrying heat away with it—and the high heat capacity of water surfaces. Come autumn, tree canopy height became the more influential variable, as taller canopies continue shading the ground and modulating radiation long after peak growing season has passed. And in winter, when vegetation activity is at its low point, the surrounding urban fabric took over: building height, building density, and the extent of construction land around the parks became the strongest predictors of how parks behaved thermally.

Perhaps the most practically valuable result of the study concerns sheer size. Park area and perimeter together accounted for more than 60 percent of the relative importance in explaining both the park cooling area and the park cooling efficiency, confirming that the physical dimensions of a park remain the dominant lever on how far its cooling influence reaches and how efficiently it operates. But the relationship is not a simple case of bigger is always better. By analyzing where efficiency peaks along the size spectrum, the team identified an optimal park area range of roughly 25 to 33 hectares for efficient cooling in Zhangzhou. Parks smaller than this range cannot generate or sustain a substantial cool island that escapes their boundaries, while parks substantially larger begin to show diminishing returns, delivering less additional cooling benefit per hectare of land invested.

That threshold carries real weight for planners wrestling with competing demands for urban land. In fast-growing cities across China and much of the developing world, every hectare reserved for green space is a hectare not available for housing, commerce, or infrastructure. A 25-to-33-hectare target—comparable to 35 to 45 standard football pitches—gives planners an evidence-based sweet spot: large enough to generate meaningful, spreading cooling, but not so large that the marginal cooling return fails to justify the land cost. It also suggests that a portfolio of mid-sized parks may outperform a single sprawling green space, and that small pocket parks, while valuable for other reasons, should not be counted on to deliver measurable district-scale temperature relief.

The methodological approach also pushes the field forward. Rather than relying on a single cooling metric, the study calculated both maximum cooling impact measures and cumulative metrics—the total cooling experienced across a park’s entire influence zone—capturing the trade-off between how deeply a park cools at its core and how widely that cooling spreads. The drivers were analyzed with relative-importance methods that can attribute explanatory power across correlated variables, an essential step when factors like vegetation cover, water extent, and park size inevitably overlap. And by repeating the analysis across 23 ECOSTRESS observation periods, the team ensured that their seasonal conclusions rest on a dense temporal record rather than a handful of favorable overpasses.

The broader significance of the work lies in who it speaks to. Heat is increasingly recognized as the deadliest weather hazard, and the burden falls disproportionately on cities that lack the resources of wealthy megacities. Prefecture-level cities like Zhangzhou sit squarely in that demographic, growing rapidly, warming steadily, and historically underrepresented in the urban climate literature. By showing when parks cool most, what drives their performance in each season, and how big they should be, the study hands these cities a blueprint. Plant for spring strength, design for summer efficiency, preserve tall canopy for autumn shade, and account for winter’s urban-morphology dominance—and above all, aim for the size that buys the most cooling per hectare. As global temperatures continue to climb, that kind of precision in green-space planning could become one of the cheapest and most effective climate adaptations a mid-sized city can make.

Subject of Research: Seasonal variations and size thresholds of urban park cooling effects measured with satellite land surface temperature data

Article Title: Multi-temporal assessment of seasonal variations and size thresholds of urban park cooling using ECOSTRESS observations

Article References: Yao, S., Hao, Z., Post, C. J., Mikhailova, E. A., & Lin, L. (2026). Multi-temporal assessment of seasonal variations and size thresholds of urban park cooling using ECOSTRESS observations. Theoretical and Applied Climatology, 157(11), Article 715. https://doi.org/10.1007/s00704-026-06644-7

Image Credits: AI Generated

DOI: 10.1007/s00704-026-06644-7

Keywords: urban heat island, ECOSTRESS, land surface temperature, urban parks, cooling efficiency, Zhangzhou, seasonal variation, remote sensing, urban planning, evapotranspiration, green infrastructure, prefecture-level cities