Urban heat islands in Hong Kong: statistical modeling and trend detection
Publication in refereed journal


Times Cited
Web of Science26WOS source URL (as at 30/11/2020) Click here for the latest count
Altmetrics Information
.

Other information
AbstractUrban heat islands (UHIs), usually defined as temperature differences between urban areas and their surrounding rural areas, are one of the most significant anthropogenic modifications to the Earth's climate. This study applies the extreme value theory to model and detect trends in extreme UHI events in Hong Kong, which have rarely been documented. Extreme UHI events are defined as UHIs with intensity higher than a specific threshold, 4.8 for summer and 7.8 A degrees C for winter. Statistical modeling based on extreme value theory is found to permit realistic modeling of these extreme events. Trends of extreme UHI intensity, frequency, and duration are introduced through changes in parameters of generalized Pareto, Poisson, and geometric distributions, respectively. During the 27-year study period, none of the quantities in winter analyzed in this study increased significantly. The annual mean summertime daily maximum UHI intensities, which are samples from a Gaussian distribution, show an increasing but nonsignificant linear trend. However, the intensity of extreme UHI events in summer is increasing significantly, which implies that the risk of mortality and heat-related diseases due to heat stress at night (when the daily maximum UHI occurs) in summer is also increasing. The warming climate has threatened and will continue to threaten inhabitants of this subtropical high-density city. Strategies for adaptation to and mitigation of climate change, such as adding greenery and planning a city with good natural ventilation, are needed.
All Author(s) ListWang WW, Zhou W, Ng EYY, Xu Y
Journal nameNatural Hazards
Detailed descriptionVol. 83,
Year2016
Month9
Day1
Volume Number83
Issue Number2
PublisherSPRINGER
Pages885 - 907
ISSN0921-030X
eISSN1573-0840
LanguagesEnglish-United Kingdom
KeywordsExtreme value theory; Generalized Pareto distribution; Parametric trend; Peaks-over-threshold model; Urban heat island
Web of Science Subject CategoriesGeology; Geosciences, Multidisciplinary; Meteorology & Atmospheric Sciences; Water Resources

Last updated on 2020-01-12 at 00:50