不同大氣穩定度下街道尺度風速與空氣污染物時空分布分析

Spatial-temporal Patterns of Street-level Wind Fields and Air Pollutants Under Different Atmospheric Stability

邵旻純(Min-Chun Shao)

指導教授 ： 莊振義

國立臺灣大學/理學院/地理環境資源學研究所/碩士(2020年)

https://doi.org/10.6342/NTU202000570

摘要

在都會區中行人高度(離地面兩公尺高範圍內)的空氣污染物對於人體健康會產生直接危害，而風場為影響空氣污染物空間分布的主要因素之一。過去研究多使用環境流體力學模式，利用推估或量測的風場，於單一街道或小範圍街廓模擬污染物濃度空間分布。然而，此方法受計算資源限制，無法及時呈現大範圍區域之模擬結果。而針對研究範圍大於街道尺度的文獻，大多在不考慮地表特性的情況下使用空間內插方法進行風場推估，並未考慮地表大氣動力特性與大氣穩定度的影響。本研究在考量平均地表粗糙度以及都市中的大氣穩定度下，使用改良的空間內插方法估算行人高度範圍的風速，並用來檢視與行人高度空氣污染物濃度的特徵。在都市當中，植被與建物會影響地表粗糙度並進而影響風場，本研究以風速剖面指數(wind profile power law) ，在中性大氣穩定度條件下，針對風速比較反距離權重法（均方根誤差為1.51 ms-1）以及普通克利金（均方根誤差為1.52 ms-1）兩種內插方法，發現兩者皆適合本研究使用，採取計算速度較快的前者。此外，本研究比較三種不同大氣穩定條件，包括極度不穩定(extremely unstable）、中性(neutral）以及中度穩定(moderately stable），在風速推估的結果發現極度不穩定的大氣狀況設定(冬季均方根誤差為1.11 ms-1；夏季均方根誤差為0.85 ms-1)在台北地區的風速估算較符合實際量測值。而對於行人高度的污染物，藉由設置於同棟建築物不同樓層的空氣盒子數據當中，發現擺放於行人高度的污染濃度平均高於高樓層的量測值（≈ 0~5 μgm-3），因此本研究僅採用現有位於2公尺高的空氣盒子與推估出的行人高度風速進行時間序列分析。取得風速與污染物之間的回歸關係，利用回歸係數計算的污染物估計值與實際量測值平均誤差在2 μgm-3以內。本研究另外透過網際網路地理資訊系統，提供使用者即時環境資料空間資訊。

關鍵字

都市建成環境 ； 地理資訊系統 ； 地表粗糙長度 ； 懸浮微粒

並列摘要

Wind field is a key factor to alter the spatial distribution of air pollutants in the urban area because the complex compositions of different land-use types and buildings strongly affect the physical properties and air flow in the built-up area. Street-level wind velocity is recognized to have significant impacts on air pollutant concentrations at the pedestrian level (~2m) which is becoming one of the major environmental issues for human health in many metropolitan areas around the world. Previous studies have discussed the distribution of air pollutants for different wind conditions with Computational Fluid Dynamics modelling in the small area of street canyon, and spatial interpolation approaches have been used to estimate the distribution of wind fields and air pollutant concentration in the relatively large areas as well. However, the former technique requires considerable computational resources, and the latter methods usually ignore the heterogeneity of surface characteristics and the role of atmospheric stability on wind speed estimation. In this study, surface roughness and atmospheric stability in urban area are taken into consideration, and a modified interpolation method is used to obtain wind velocity for the range of the pedestrian level. Besides, the patterns of air pollutant concentrations are inspected through the results of estimated wind speed. In urban area, vegetation and buildings strongly affect surface roughness and wind filed. The results show that under neutral stability condition, inverse distance weighting (IDW) (root mean square error (RMSE = 1.51 ms-1) performs almost the same accuracy as ordinary kriging OK (RMSE = 1.52 ms-1) for the wind speed estimation with wind profile power law. Because of low time cost, IDW is selected in this study. Furthermore, three kinds of atmospheric stability conditions, including extremely unstable, neutral, and moderately stable stability, are compared in this study. The results under unstable stability condition for wind speed estimation (RMSE is 1.11 ms-1 in winter; RMSE is 0.85 ms-1 in summer.) are more consistent with field measurements than the others in the study area. In this study, real-time spatial distribution of wind speed at pedestrian height are achieved with surface roughness, atmospheric stability, and the interpolation method. This model is verified by selecting several wind stations to ensure the performance of the proposed approach, and difference between estimated results and actual wind speed is almost within -1~1 ms-1. Additionally, by installing AirBox devices at different heights outside the wall of buildings, it is found that the mean concentrations at pedestrian height is greater than the highest floor (≈ 0~5 μgm-3). Therefore, this study only uses the AirBox devices at 2-meter-hight, and the relationship between these particulate matters and estimated wind speed for pedestrian height is carried out the time-series analysis. With the correlation coefficient of relationship, the average errors between actual and estimated values are within 2 μgm-3 in the study area. In this study, the real time environment spatial data is also provided for users through Web Geographic Information System (WebGIS).

並列關鍵字

built environment ； GIS ； PM2.5 ； PM10 ； surface roughness length

參考文獻

Abhijith, K. V., Kumar, P., Gallagher, J., McNabola, A., Baldauf, R., Pilla, F., . . . Pulvirenti, B. (2017). Air pollution abatement performances of green infrastructure in open road and built-up street canyon environments – A review. Atmospheric Environment, 162, 71-86. doi:https://doi.org/10.1016/j.atmosenv.2017.05.014

Akdağ, S. A., Ö, G., & Yağci, E. (2013, 20-23 Oct. 2013). Wind speed extrapolation methods and their effect on energy generation estimation. Paper presented at the 2013 International Conference on Renewable Energy Research and Applications (ICRERA).

Arain, M. A., Blair, R., Finkelstein, N., Brook, J. R., Sahsuvaroglu, T., Beckerman, B., . . . Jerrett, M. (2007). The use of wind fields in a land use regression model to predict air pollution concentrations for health exposure studies. Atmospheric Environment, 41(16), 3453-3464. doi:https://doi.org/10.1016/j.atmosenv.2006.11.063

Balczó, M., Gromke, C., & Ruck, B. (2009). Numerical modeling of flow and pollutant dispersion in street canyons with tree planting. Meteorologische Zeitschrift, 18(2), 197-206.

Barnes, M., Brade, T. K., Mackenzie, A. R., Whyatt, J., Carruthers, D., Stocker, J., . . . Hewitt, C. (2014). Spatially-varying surface roughness and ground-level air quality in an operational dispersion model. Environmental Pollution, 185, 44-51.