臭氧空間暴露推估模型之比較－以六輕工業區為例

Comparison of Geospatial-Temporal Modeling Approaches in Ozone Pollution Estimations

曾于庭(Yu-Ting Zeng) ； 吳治達(Chih-Da Wu) ； 陳裕政(Yu-Cheng Chen) ； 許金玉(Chin-Yu Hsu) ； 陳穆貞(Mu-Jean Chen)

《航測及遙測學刊》 24卷4期 (2019/12) Pp. 235-244

https://doi.org/10.6574/JPRS.201912_24(4).0003

摘要

隨著地理資訊系統以及遙感探測技術上的純熟，空氣汙染空間暴露推估模型的發展更具多樣性。在過去研究中，已有許多應用單一模型進行空氣汙染物時空分布推估的案例，但系統性的比較不同推估模型解釋能力之研究仍不多見。基於此，本研究以環保署於雲林及嘉義所設置之十個特殊性工業區測站，於2015至2018年之空汙觀測資料為材料，選擇多元逐步迴歸為基礎之土地利用迴歸、地理加權迴歸及地理時間加權迴歸等三種統計建模方法，綜合比較不同方法學於O_3汙染物分布推估能力之差異。研究結果顯示，O_3模型R^2值介於0.51－0.64，並且以時間地理加權迴歸之模型結果最佳。

關鍵字

臭氧 ； 土地利用迴歸 ； 地理加權迴歸 ； 時間地理加權迴歸

並列摘要

Recent advancements in the geographic information systems and remote sensing technology have supported the development of geospatial-temporal modeling approaches for air pollution. Previous studies estimated the spatial-temporal variability of air pollutants using a single model, but only a few studies considered exposure assessment using multiple models and compared model performance. In this study, O_3 data during 2015 to 2018 was collected from specific industrial monitoring stations provided by the Taiwan Environmental Protection Agency. Three geospatial-temporal modeling approaches including land-use regression (LUR), geographically weighted regression (GWR), and geographically and temporally weighted regression (GTWR) were used to predict O_3 for our comparison. The results showed that R^2 obtained from all models were 0.51 to 0.64. Furthermore, the GTWR model has the greatest performances compared to LUR, and GWR models.

並列關鍵字

Ozone ； Geographically and Temporally Weighted Regression (GTWR) ； Geographically Weighted Regression (GWR) ； Land-use Regression (LUR)