利用Landsat 8 OLI影像反演氣溶膠光學厚度之成果論證臺中市交通流量對PM_(2.5)之影響

Using the Aerosol Optical Depth Data Retrieved from Landsat 8 OLI Imagery to Demonstrate the Influence of Traffic Flow on PM_(2.5) in Taichung

吳兆鴻(Chao-Hung Wu) ； 徐逸祥(Yi-Shiang Shiu) ； 張晏菁(Yen-Ching Chang)

《航測及遙測學刊》 24卷1期 (2019/03) Pp. 59-77

https://doi.org/10.6574/JPRS.201903_24(1).0005

摘要

細懸浮微粒PM_(2.5)為臺灣中南部地區最關切的環境議題之一，其成因目前仍眾說紛云，除臺中火力發電廠之外，汽機車排放的廢氣亦常被歸咎為主因。然PM_(2.5)和交通量的監測常因僅有點狀的監測數據而無法全面探討交通量是否對PM_(2.5)有直接影響。因此本研究利用Landsat 8 OLI影像及離散係數法反演氣溶膠光學厚度，結合臺中及鄰近縣市共25個空氣品質測站，推估原臺中市假日及非假日的PM_(2.5)；而交通量方面以Google地圖一般路況車流量的平均壅塞程度與實際車道的車道寬、路口停滯秒數等推算成估計的小客車當量來表達影像當下交通的壅塞程度。最後再考量空間自相關和不同空間單元的前提下，以空間迴歸模型決策模式找出最適合的迴歸模型，來解釋當地PM_(2.5)受本身及鄰近區域因子影響的情形。

關鍵字

氣溶膠光學厚度 ； 小客車當量 ； 空間迴歸模式

並列摘要

Fine particulate matter PM_(2.5) is one of the most concerned environmental issues in central and southern Taiwan. Except for Taichung Thermal Power Plant, the exhaust emissions from motor vehicle are often attributed to the main cause. The ground and point monitoring approaches of air quality and traffic flow are generally used while may not fully explore whether traffic flow has a significant impact on PM_(2.5). Therefore, this study used Landsat 8 OLI image and dispersion coefficient method to retrieve aerosol optical depth. We also included 25 air quality stations in Taichung and neighboring counties as the reference to estimate PM_(2.5) for holiday and normal day in Taichung City. The passenger car equivalent was estimated with average congestion level from Google Map general road traffic as well as actual road width and the seconds of stopped-time to estimate the congestion of the traffic. Finally, considering the spatial autocorrelation and different spatial units, the spatial regression model decision process is used to explore the most suitable regression model to explain the local PM_(2.5) affected by the local and neighboring factors.

並列關鍵字

aerosol optical depth ； passenger car equivalent ； spatial regression model