Examining Urban Heat Island Effect and Its Public Health Implications with Remotely Sensed Data

Abstract

The Urban heat island (UHI) as a byproduct of urbanization has long been studied utilizing remote sensing technologies. However, issues remain to be addressed. Land surface temperature (LST) as the indicator of surface UHI can be retrieved from remotely sensed data, but its accuracy is limited as existing studies neglect the neighboring effect. Further, while LST serves well as an indicator of surface thermal condition, it lacks the ability to reveal human heat stress, which is an environmental hazard that can seriously affect productivity, health or even survival of individuals. Although human heat stress has long been studied and can be quantified by many heat stress indices, it has never been explored across continuous spaces. Aiming to address these issues, the objectives of this research include: (1) taking into account the neighboring effect during LST retrieval using a moving window method; (2) revealing human heat stress with remotely sensed data; and (3) exploring the relationship between heat stress and land cover composition and configuration. My results indicate that the accuracy of LST estimation is improved when neighboring effect is considered. Discomfort index (DI) as an indicator of human heat stress can be retrieved from remotely sensed data, and its spatial distribution and relationship with land cover composition is largely affected by relative humidity. Spatial configuration of different land covers has an impact on DI, which may provide insights for policy makers and urban designers on mitigating hazardous environmental effect brought by urbanization.