Are urban villages cooler than other built-up land?

Details

Widely scattered urban villages and increasingly serious urban heat islands are common problems in highly urbanized regions, especially in developing countries, but the influence of urban villages on urban heat islands remains little understood.

Researchers studied the Guangzhou–Foshan core areas of the Pearl River Delta. They used different methods to retrieve land surface temperature from thermal bands and used the nearest object-oriented method with spectral, texture and shape metrics, ZY-3 high-resolution satellite imagery, and road network data to extract urban villages and other land-use types. They then analyzed the relationship between land surface temperature and land-use types at multiple scales.

The object-oriented method identified five land-use types: vegetation, normal construction land, urban villages, water and unused land. Its overall accuracy of 0.887 and Kappa coefficient of 0.863 were considered satisfactory for subsequent analysis.

The Guangzhou–Foshan core areas showed the most outstanding urban heat island effect across the Pearl River Delta region. About 60.5% of their total area was covered by impervious surfaces, including normal construction land (50.4%) and urban villages (10.1%). The average land surface temperature of urban villages was 1.89–2.97 °C lower than that of normal construction land.

Based on the average contribution index of thermal effect and Pearson's correlation coefficients, urban villages had a relatively lower contribution to the urban heat island and a weaker warming effect than normal construction land, but a higher contribution and stronger warming effect than the other land-use types. On surface-temperature profile lines, urban villages next to normal construction land formed slightly lower temperature valleys, while those close to vegetation and water formed distinct temperature peaks.

The authors say the work improves understanding of the relationship between increasingly serious urban heat islands and widely distributed urban villages and could help local authorities monitor and improve the urban environment in metropolitan regions.