Urban green space metric choice and weighting matter for assessing cooling benefits for human health

Extreme heat is a growing source of health-related risk in cities, and urban greening is widely promoted as a strategy to reduce its adverse consequences. However, urban green space can be measured in very different ways, and it remains unclear whether commonly used metrics are interchangeable when estimating health-relevant cooling benefits. Here we test how green space metric choice affects inferred heat-mortality attenuation in Paris using a harmonised arrondissement-day panel for the 2008–2017 summer seasons, combining daily mortality, 100-meter resolution UrbClim heat fields, and three urban vegetation measurement approaches: street-level green view index (GVI), satellite-derived normalized difference vegetation index (NDVI), and local planimetry-based vegetation cover (IMU). We estimate conditional time-series distributed lag non-linear models across nine heat indicators. Greener arrondissements consistently show lower heat-related mortality risk: across all 45 heat-metric and greenness-metric combinations assessed. Street-level GVI yields the strongest central attenuation estimate, followed closely by NDVI, and finally by IMU. For example, for mean daily wet-bulb globe temperature, the central estimate for the attenuation in heat-related mortality risk is 18.5% with GVI, compared with 16% for NDVI and 14% for planimetry-based vegetation. Despite uncertainty around individual effect sizes, the consistent directional ordering across all three metrics suggests that green-space indicators are not interchangeable in heat-health research. Metric and weighting choices made upstream—in how vegetation is measured and spatially aggregated—can materially alter estimated health burdens and, in some cases, even reverse their sign, with direct consequences for policy recommendations. We treat these findings as hypothesis-generating: more systematic investigation across cities and climate contexts is needed to establish which representations of urban greenness best capture heat-relevant cooling benefits.