Contrasting the day-night variation in thermal landscape, the higher local variation and irregularity on urban surface temperature pattern during daytime was observed and identified by landscape metrics compared with those of nighttime pattern. The diversity and fragmentation metrics had revealed the influence of urban development on the overall urban landscape pattern. Along with urban development, daytime pattern of urban thermal landscape presented more fragmentation, less diversity and uneven texture distribution within daytime observations. / During the process of worldwide urbanization along with high rise and high density housing development in large cities, urban warming has received growing concern among many environmental issues related to urban landscape change. Due to the complicated interplay between urban environment and local climate, it is far from certain about the urban effects on local warming. In this literature, a systematic monitoring and analysis of the spatial dependency and heterogeneity of urban thermal landscape at city scale remains inadequate. The goal of this doctoral research is to develop a research framework incorporating geospatial statistics, thermal infrared remote sensing and landscape ecology to study the urban effect on local surface thermal landscape regarding both the pattern and process. / GWR analysis offered an in-depth investigation of local effect on surface temperature variation which had been proven to be spatially varying and influenced by local weather condition with local environmental setting quantified with the referred site specific environmental factors. The local dominant factor accounted for most to the site specific surface temperature variation which varied significantly in space and time and prevented a general delineation of the relative association among environmental factors to surface temperature disparities. The effective adaptive measures could be devised locally with reference to day-night needs in the identification of this feature. / In summary, global regression analysis confirmed the relationship between environmental factors and surface temperature and gave a general overview of urban effect on local surface warming. The distinctive mechanism of dominating day-night surface warming was uncovered by regression analysis. Vegetation played the most important role which could be referred as surface cooling in average to local surface temperature variation as compared with other measures of local environment during both daytime and nighttime. Besides the dominant role of local solar radiation on surface warming, building square footage demonstrated the second important influence on local surface temperature elevation during daytime. During nighttime, population density played a dominant role on nighttime surface warming among different parameters, with the second important contribution of nighttime surface warming coming from road density. While elevation and distance from coast demonstrated obvious cooling effect on surface temperature within most nighttime models. / Located in a subtropical region, Hong Kong's development with high rise and high density housing made it a suitable site for studying urban effect on local warming. This research chose Hong Kong as the case study which hopes to enrich our knowledge regarding urban local thermal performance and add to our understanding of urban microclimate in hot-humid weather area. / Xue, Yucai. / Adviser: Tung Fung. / Source: Dissertation Abstracts International, Volume: 71-01, Section: B, page: 0185. / Thesis submitted in: September 2008. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2009. / Includes bibliographical references (leaves 201-228). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. Ann Arbor, MI : ProQuest Information and Learning Company, [200-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstracts in English and Chinese.

近幾十年來，全球氣候變化──特別是城市氣候變化──影響城市環境及居民生活質素的程度已引起公眾廣泛的討論。然而，過去研究一般採用之低空間解析度並不足夠為城市規劃及設計提供完善的資訊，引致對於氣候變化缺乏充分的考慮。高密度的城市環境（如香港）需要高時間解析度的氣候數據以制定有效的適應和減緩策略來應對未來氣候的變化。 / 本研究採用線性迴歸技術，以預測未來香港市區和郊區的氣溫。本研究利用氣象站和統計延伸得出之基線氣溫數據來建立統計降尺度模型，以預測未來香港市區和郊區之平均氣溫、最高氣溫和最低氣溫。 / 根據結果顯示，統計降尺度模型能夠有效建立大氣氣象參數和本港氣溫兩者之間的關係，尤其春季、秋季和冬季之氣溫預測表現理想。另外，冬季氣溫的上升趨勢則出現較大的升幅。研究結果亦顯示夜間氣溫的上升趨勢一般比日間為高。在未來的日子，郊區的溫度上升亦將會比市區為高。隨著城市化的影響納入預測溫度因素之中，預計郊區的氣溫將超過城市核心（天文台總部之氣象站），而郊區暖化的速度亦比市區和近郊為高。 / 本研究發現統計降尺度方法能有助利用全球氣候模型（GCM）提供之數據，以預測未來氣候之變化。城市規劃與設計過程是需要大量的數據進行評估氣候變化對城市環境的影響之研究，儘管統計降尺度方法有一定程度的局限性，它仍然是一個低成本而有效的方法。根據未來預測之氣溫，本研究具體指出未來之氣候變化對於城市規劃和設計過程的影響，亦提出了一系列於不同規劃層面適用之適應和減緩措施的建議。 / The effects of global climate change on urban environment have been widely discussed in recent decades. In particular, changes in urban climate have received much attention as they affect the living quality of urban dwellers. However, the coarse spatial scales employed in recent climate change studies were found to be insufficient in the context of urban planning and design. It leads to the lack of information on the changing urban climate and insufficient consideration of climate change in urban planning and design processes. In high-density cities like Hong Kong, the complex urban environment requires climatic data at very fine temporal resolution in order to formulate effective adaptation and mitigation strategies for future climate change. / The present study employed regression techniques to establish empirical relationship between large-scale predictor variables and local predictands in order to obtain future air temperature of urban and rural areas of Hong Kong. 40-year baseline conditions of local air temperature were obtained from both the observational and statistically extended temperature record. Monthly means of daily mean, maximum, and minimum air temperatures for both daytime and night-time were calculated for establishing statistical downscaling (SD) models to project future air temperature of urban and rural areas of Hong Kong. / The results suggest that regression-based downscaling techniques are able to capture the relationship between large-scale atmospheric conditions and station-scale meteorological parameters. The SD models performed particularly well in winter and considerably satisfactory results were obtained in spring and autumn. Night-time temperature trends generally exhibited greater increases than daytime trends. Seasonal variations were present with greatest increases observed in winter. Rural areas would likely experience greater warming than the urban areas in the future. With urbanization effect incorporated into the projected temperature series, it was found that air temperature projected for suburban stations would exceed that for the urban core. Rural warming also exhibited a higher rate than those observed in suburban and urban stations. / The present study shows that statistical downscaling approach provides a method to obtain information about future climatic conditions at local scale by using GCM outputs which are widely accepted to be useful tools to assist climate change studies. Despite of the limitations that historical climate would persist in projected climatic series, it allows a low-cost but effective measure for climate impact assessments, particularly in the context of urban planning and design, which requires extensive data for a wide range of studies. Based on the projected air temperature, implications of future climate change on urban planning and design of potential development were discussed and recommendations on potential adaptation and mitigation measures at different planning levels were also presented. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Lau, Ka Lun. / Thesis (Ph.D.) Chinese University of Hong Kong, 2013. / Includes bibliographical references (leaves 159-173). / Abstracts also in Chinese.