隨著世界上各大都市的快速發展，不透水的柏油、混凝土等人造建築取了植被，同時也改變了土地的物理特性。這些改變造成了城市地區夜間冷卻效率降低等環境問題，導致了所謂的都市熱島效應。在人造熱源的排放加乘下，此效果又更為嚴重。由於台灣的台北市為盆地地形，在高密度的居住情況及地形效應下，臺北盆地有著相較於同等級都市來的更嚴重的熱島問題。為了能更好地掌握台北地區邊界層內的物理量變化情況，本研究利用WRF模式(3.5.1版本)對台北的熱島效應做模擬。並針對Urban Canopy Model (UCM)及台北地區更新的城市參數來分析並改善台北地區的模擬結果。 本研究在進行台北地區個案模擬前，針對UCM會用到的參數做敏感度測試，並發現urban fraction及人造熱源對於溫度的模擬結果會有最大的影響力。而在台灣的個案中，由於北台灣都市規模較小，在未加入詳細都市參數(Urban Fraction、Urban Category)前，UCM對於夜間都市熱島效應的掌握能力並不好。因此我們預期加入較精細的都市參數資料對於模擬結果會有正向的影響。在加入本研究根據Landsat 8衛星資料所製作出的都市參數後，溫度模擬最明顯差異的區域落在MODIS與Landsat 8資料較不吻合處。將模擬結果與大台北地區其中11個城市測站的觀測值來作比較，可以發現加入了較詳細的都市參數後，有效的拉近了觀測與地表溫度之模擬結果。而水平風及混合比的垂直變化在加入詳細城市參數前後則沒有太大的變化。 ;Following the rapid development of metropolitan area in Northern Taiwan, changes in surface physical characteristic lower the efficiency of nighttime cooling effect and cause the Urban Heat Island Effect (UHI). This effect is further enhanced by the emission of anthropogenic heat (AH). With the basin topography and about a quarter of the total population on the island, Taipei basin has more significant urban heat island issue comparing to other mega-city around the world with such high intensity residency and topography effect. In order to get a better understanding of physics process in boundary layer of Taipei, WRF is implemented in this study to simulate UHI effect in Taipei metropolitan area. Urban Canopy Model (UCM) and updated urban parameter are also used to improve the result of simulation. According to the sensitivity on the parameters used by UCM, urban fraction and AH have the largest impact on temperature simulation. In the case of Taipei, UCM fails to capture the nighttime UHI effect since the size of Taipei metropolitan area is not big enough. After adding the detailed urban parameter (urban fraction and urban category) extracted from Landsat 8 into simulation, the biggest temperature difference with simulation by using the original UCM parameter falls on the area where data of MODIS and Landsat 8 coincide the least. By comparing the temperature simulation result with observation, it can be seen that after adding the detailed urban parameter, simulation result is closer to observation, but there is not much change in the vertical profile of horizontal wind and mixing ratio.
