Beyond Canopy Targets: Pedestrian-Level Evidence for Urban Heat Adaptation Policy

• In Event: Extreme Heat, Health, and Urban Adaptation

Abstract

Canopy targets help governments track urban greening, but assessing heat adaptation also requires evidence about thermal conditions where people use public space. This study examines two requirements for that assessment: whether a summary measure captures informative differences in local canopy configuration, and whether the resulting predictions transfer to parks outside the training sample.

The analysis uses 2,368 paired sun and shade air-temperature observations collected at recurring locations in nine Greenwich Village parks across four survey dates, with 2,345 unflagged records in the primary sample. Date-specific regressions, comparisons matched on sky view factor (SVF) and local imperviousness, and spatially grouped cross-validation assess the information added by the difference between upper and lower angular gap fractions.

On the spring and summer survey days, angular configuration adds explanatory information beyond overall openness and local context. Matched observations with higher angular gap contrasts have mean sun–shade temperature differences approximately 0.45°C and 0.41°C larger, respectively. The relationships persist across covariate and sample checks. Predictive gains are more limited across parks: adding the contrast reduces pooled root mean squared error by 7.1% when sampling locations are withheld and 2.0% when entire parks are withheld. The difference between validation settings exceeds the gain from the additional indicator.

The findings contribute to policy performance measurement by distinguishing the information needed to characterize local environments from the evidence needed to extend a model’s use. They support supplementing canopy inventories with repeated field assessments and validation at the intended geographic and temporal scale. A larger sun–shade contrast alone does not establish cooler absolute conditions or greater comfort. Evaluating accessible thermal relief therefore requires connecting physical measurements with evidence on pedestrian experience and use.

Author

• Yeji Choi, New York University
  Presenting Author