Phnom Penh Urban Park Thermal Dataset: LST, NDVI, and Cooling Metrics from OpenStreetMap Polygons (2022–2024)

Phnom Penh Urban Park Thermal Dataset: LST, NDVI, and Cooling Metrics from OpenStreetMap Polygons (2022–2024)Asian cities.Phnom Penh Urban Park Thermal Dataset: LST, NDVI, and Cooling Metrics from OpenStreetMap Polygons (2022–2024)

This dataset provides a polygon-level thermal performance assessment of urban green and open spaces in Phnom Penh, Cambodia, derived from remote sensing and geospatial analysis. The dataset includes 128 OpenStreetMap (OSM) leisure polygons (≥0.3 ha) and integrates multi-source satellite observations with spatial metrics to quantify urban park cooling performance in a tropical secondary-city context.

Land surface temperature (LST) is derived from Landsat 8 and 9 Collection 2 Level-2 products, composited across two dry seasons (November–March, 2022–2024). Vegetation density is measured using Sentinel-2 NDVI composites aligned to the same temporal window. For each polygon, the dataset reports mean LST, surrounding buffer LST (300 m), cooling intensity (CI), and cooling distance (CD), along with structural attributes including area, shape index, and compactness.

Cooling intensity (CI) is defined as the temperature difference between the surrounding buffer and the polygon interior, while cooling distance (CD) captures the spatial extent of measurable cooling effects using a 30 m ring-based approach. The dataset also includes NDVI-based subgroup classification and derived indicators used in regression and spatial autocorrelation analyses.

This dataset enables reproducible analysis of urban thermal dynamics in data-scarce environments and supports comparative studies of green infrastructure performance in rapidly urbanising Southeast Asian cities.