建構都市街谷微氣候逐時溫度預測模型與驗證

Abstract

過去許多微氣候模擬的研究大多著重於數值分析，但數值模擬研究之結果僅能提供理論上的情況，必須將理論研究應用到實際環境狀況中，才能真正展現其實際舒緩成果，故本研究希望完成最冷日與最熱日的各街谷型態微氣候預測式，並完成預測特定街谷型態的全年逐時微氣候推估模式，提供給未來研究可參考使用。 本研究將會使用數值模擬並實行實地測量進行微氣候預測模型的驗證，提出一個完備的微氣候成果。本研究以高寬比、街道座向、高寬比與路寬交互作用三個因子為主要影響街道微氣候的顯著因子，將以此結果製作出微氣候預測式。後續對台北市溫羅汀街區進行實測，也使用ENVI-met建構溫羅汀街區的街谷模型，最後將實測、模擬和預測三個結果進行分析，應證預測式之準確度。 本研究之研究結果完成對微氣候預測式的製作與驗證。10/01四個測點之預測與實測變異係數CV分析結果皆在2.5%以內，整體小於5%，2/18四個測點之預測與實測變異係數CV分析結果整體在5%以內，總結來說預測與實測CV分析結果在5%以內，證明其預測結果優良。同樣對預測與模擬進行變異係數CV分析，分析結果依然在5%以內，代表預測式預測合理。 最後本研究將微氣候預測式實際應用在溫羅汀街區以及台北市8m以上道路，模擬出社區微氣候及市區大範圍氣候情況。溫羅汀街區8/01種樹預測結果顯示，在南北向與風向垂直的街區，種樹前與種樹後溫度修正量平均降低了0.1K，最大溫度修正量降低了0.137K，相比原來的降低了約10%，表示種植行道樹效益佳。根據台北市主要道路種樹前後結果顯示，種樹前後平均溫度修正量差異僅為0.052K，降溫幅度較低，推測原因為大型主要幹道其建築間距較遠(大於26m)，街道高寬比多在最低臨界值0.5，導致降溫效益將會變得較不顯著。

Previous research on microclimate simulations has primarily focused on numerical analysis, providing theoretical scenarios. However, to demonstrate practical mitigation outcomes, it is essential to apply theoretical research to real environmental conditions. Therefore, this study aims to develop prediction models for microclimates during the coldest and hottest days in various street canyon configurations. Additionally, a yearly hourly microclimate estimation model for specific street canyon patterns will be established to serve as a reference for future research. The proposed microclimate prediction model will be validated using numerical simulations and on-site measurements to provide comprehensive results. The study identifies three significant factors: aspect ratio, street orientation, and the interaction between aspect ratio and road width, as the main influencers of street microclimates. These findings will be used to create the microclimate prediction model. The study will perform field measurements in the Wunhua area of Taipei City and construct a street canyon model using ENVI-met for validation. Subsequently, the results of measurements, simulations, and predictions will be analyzed to verify the accuracy of the prediction model. The research results confirm the development and validation of the microclimate prediction model. The coefficient of variation (CV) analysis for four measurement points on 10/01 indicates that the prediction and measured values have a CV within 2.5%, overall below 5%. Similarly, on 2/18, the CV for four measurement points remains below 5%. In summary, the CV analysis for predictions and measurements is within 5%, demonstrating the excellent performance of the prediction model. The analysis of the CV for predictions and simulations also stays within 5%, indicating the reasonability of the prediction model. Finally, the microclimate prediction model is applied to the Wunhua area and roads over 8 meters wide in Taipei City, simulating community microclimates and citywide climatic conditions. The results of tree planting prediction in the Wunhua area on 8/01 show that in the north-south and wind-parallel streets, the average temperature reduction after tree planting is 0.1K, with a maximum temperature reduction of 0.137K, representing an approximately 10% improvement compared to the original conditions, highlighting the benefits of tree planting. However, for major roads in Taipei City, the average temperature reduction after tree planting is only 0.052K, indicating a lower cooling effect. This might be attributed to the larger spacing between buildings (over 26m) on major arterial roads, resulting in lower significance of cooling benefits due to their aspect ratios mostly being at the lowest critical value of 0.5.