| 摘要: | 近年來，有許多研究透過不同的都市冠層模式與土地利用分類，改善氣象模式對於都市型態的描述，評估都市化對大氣環境的影響。而高解析度或準確的都市資訊(如土地使用、建物型態與分佈)並不容易取得，World Urban Database and Access Portal Tools (WUDAPT)資料庫透過衛星、光達及建物資料收集全球都市相關資訊，使用當地氣候區(Local Climate Zone, LCZ)作為土地利用分類依據，基於不同地表結構特性，提供十七種土地利用類別與參數，供都市冠層模式使用。
本研究使用Weather Research Forecast(WRF)耦合Building Energy Parameterization (BEP)及Building Energy Model (BEM)模擬臺灣地區氣象特性，探討都市冠層模式特徵與對氣象模擬的影響。此外，將臺灣地區LCZ十種都市分類資料納入都市模式中，並和使用原UCM模式識別的三種都市種類資料之模擬結果進行比較，評估較精細之都市土地利用分類與都市參數的模式表現。
模擬結果顯示，Base組將都市網格視為均值不透水區域，模擬的溫度與可感熱通量較高；BEP組中，因UCM考慮都市網格中不透水表面和植被覆蓋，相較於Base組造成的可感熱通量模擬較低，夜間都會區所模擬的溫度低於未使用UCM的模擬，模式低估且近地表呈現相對穩定的大氣結構；BEM組因進一步考慮人為熱量排放且其影響在夜間較顯著，進而改善都市地區夜間模擬，夜間地表大氣狀態相對不穩定，其人為熱排放影響主要局限於近地表300至400公尺左右。風速方面，UCM因建物阻力效應及粗糙度的增加，模擬風速較低，能有效的改善未使用UCM模擬中環境風場嚴重高估問題。
將LCZ都市分類結果納入BEP中，北部都會區日間以主要密集住宅區、商業區、住商混合與工業區的地方溫度量值較高，夜間緊湊型高層建物因建物密度與建物高度影響，顯示較高的溫度；風速方面，建築物密集程度越高且高度越高的都市分類，模擬風速值較低。高雄都會區的主要溫度較高的區域為工廠區域，而研究中顯示高雄都會區夜間開放型中層建物所模擬的溫度較高，異於北部都會區模擬特性，顯示不同都市區域建物結構對於氣象場模擬的差異。和使用原UCM模式所識別的三種都市種類模擬比較顯示，使用LCZ資料的可感熱通量模擬較低且日間差異較顯著，使得日間溫度模擬差異較顯著，都市占比較低的區域，多與使用LCZ資料所模擬地表溫度較低的區域相吻合，LCZ都市參數設定上的差異，皆為導致使用LCZ資料所模擬地表溫度較低的因素，整體模式使用LCZ都市資料溫度低估量值較大。風速方面，兩組都市土地利用資料對於風場模擬在市中心的差異不大，模擬風速值皆較低。
;In recent years, many studies have used different urban canopy models (UCM) and land use classifications to improve the description of urban types in meteorological models. High resolution or accurate urban information such as land use, building types and distribution isn′t easy to acquire. The World Urban Database and Access Portal Tools (WUDAPT) database collects global urban information through satellite, LiDAR and building data, using Local Climate Zone (LCZ) classification based on different surface properties to provide seventeen land use types and their corresponding parameters provided to urban canopy model for application.
To investigate the characteristics of urban canopy models and their effects on meteorological simulations, Weather Research Forecast (WRF) coupled with Building Energy Parameterization (BEP) and Building Energy Model (BEM) was applied to simulate the meteorological characteristics of Taiwan. Besides, to evaluate the refined urban land use classification and urban canopy parameters, ten urban classification data of LCZ in Taiwan were incorporated in the urban model and compared with the simulation of using the three urban types identified in the default UCM.
The BEP simulation showed lower temperature in urban areas than the simulation w/o UCM during nighttime. Due to consideration of the percentage of pervious regions and existence of urban vegetation for urban grids in UCM, less surface heat fluxes were simulated. The results show tendency to underestimate temperature and produce a relatively stable boundary layer. Further consideration of the anthropogenic heat release in BEM simulation, improved the nighttime simulations. Besides, BEM produce relatively unstable boundary layer and the impact of anthropogenic heat emission mainly limited around 300 to 400 meters near the surface in nighttime. UCMs simulated lower wind speed due to the blocking effect of buildings and the increase in roughness, which effectively improve the overestimation of wind field in the simulation without UCM.
Incorporating the LCZ urban classification results into the BEP model, the main dense residential, commercial and industrial area in northern metropolitan area show higher temperature in daytime. The compact high-rise buildings show higher temperature in nighttime due to its density and height of buildings. The urban classification with higher density and higher height of buildings, the lower wind speed was simulated due to the poor ventilation effect and stronger blocking effect. In Kaohsiung area, it shows higher temperature in industrial areas and higher temperature in open high-rise buildings in nighttime, which is different from north areas. The different regional building structures have important impact on meteorological simulation.
Compared with the simulation of the three urban types identified by the original UCM model, the sensible heat flux simulation using the LCZ data is lower and more significant in daytime, which makes the difference in daytime simulation more obvious. The area with lower surface temperature simulated by the LCZ data is consistent with the urban fraction differences. Besides, the different urban parameter settings of LCZ are factors causing the lower surface temperature compared with the simulation of the three types. The two different urban land use data show insignificant differences in the simulation of wind field in the city center. |
