The impact of urban physical environments on cooling rates in summer: Focusing on interaction effects with a kernel-based regularized least squares (KRLS) model

Yeri Choi, Sugie Lee

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

Rapid urbanization causes changes in land-cover and land-use patterns, and leads to warmer urban temperatures; this is defined as the urban heat island (UHI) effect. The intensity, duration, and frequency of heat waves have increased in recent years, posing a serious threat to urban populations. This study examined the relationships between urban physical elements and the cooling rate (CR) around sunset in the summer of 2016, using high-resolution climate data from 218 automatic weather stations (AWSs) around Seoul, South Korea. To detect possible nonlinearities and interactions among predictors, the Kernel-based regularized least squares (KRLS) estimation approach was adopted for empirical analysis. Along with the KRLS model, traditional ordinary least squares (OLS) analysis was also conducted mainly for comparison purposes. The results showed that urban elements, including both land-cover and three-dimensional built environment characteristics, had a significant influence on CR. In addition, significant interacting behaviors were also found. Our results indicate that a more comprehensive approach to understanding both the singular and compounding effects of various urban characteristics can help ameliorate outdoor thermal environments by enhancing summertime CR.

Original languageEnglish
Pages (from-to)523-534
Number of pages12
JournalRenewable Energy
Volume149
DOIs
StatePublished - 2020 Apr

Keywords

  • Air temperature
  • Built environment
  • Cooling rate
  • Kernel-based regularized least squares model
  • Thermal environment
  • Urban heat island effect

Fingerprint

Dive into the research topics of 'The impact of urban physical environments on cooling rates in summer: Focusing on interaction effects with a kernel-based regularized least squares (KRLS) model'. Together they form a unique fingerprint.

Cite this