以渦度相關法驗證潛熱通量估算式並應用於AERMOD中改善包溫比的決定

Abstract

台灣目前使用的空氣品質模式AERMOD (AMS/EPA Regulatory Model)源於美國，模式當中包溫比是以地表條件及季節決定的，然而包溫比是會隨著氣象條件及地表條件等因素而改變，故以定值決定包溫比並不適當。包溫比是可感熱通量與潛熱通量的比值，其中潛熱通量是地表蒸發及植被蒸散的總和，但因涉及太多層面，故無法準確的模擬，因此便發展出許多估算式。 Penman-Monteith（PM）及Priestley and Taylor（PT）是目前最普遍被使用的估算式，但這兩個估算式並未經實測值驗證過，故無法確認其在臺灣的適用性。本研究以關渡地區為例以渦度相關法（Eddy Covariance）驗證Penman-Monteith及Priestley and Taylor估算式，使其估算結果能符合臺灣關渡地區現況。再利用估算式所求出的潛熱通量，進而求得逐時包溫比，取代模式中以定值決定包溫比的方式。 由研究結果顯示，Penman-Monteith及Priestley and Taylor均適用於關渡地區，其中Penman-Monteith式中的冠層阻力在關渡地區約為100（sm-1）。利用Penman-Monteith式所求得的包溫比，與實際包溫比的相關度較高；而Priestley and Taylor式所求的的包溫比，因其公式本身僅與溫度相關，故其包溫比僅與溫度的變化有關。將上述兩方法所求得的包溫比代入AERMOD中，並將渦度相關法所求得的視為實際值，經比較之後，發現預設值包溫比代入所求得的污染物濃度與實際的污染物濃度相關度較低，且在冬天污染物低估的情形較嚴重。本研究所求出的逐月包溫比與模式預設值差異極大，因預設值無法反映出地表狀況及氣象，因此在使用AERMOD時建議輸入逐月包溫比，可使模擬出的值更接近實際情形。

Many air quality models developed from the USA have been widely utilized in Taiwan. However, in these modeling systems, many of default surface parameters are set for the topographic and climatic conditions in the USA, which are quite different from those in Taiwan. In addition, the simulation of diffusion patterns of air pollutant is directly affected by the estimation of the atmospheric stability, which can be quantified from the sensible heat flux. In many studies, the utilization of Bowen ratio by using different approaches are commonly used to calculate sensible heat fluxes and further determine the atmospheric stability used in air quality models . However, very few studies discuss the comparison among these approaches by using local measurement. Therefore, the objective of this research is to determine the localized surface parameters suitable in AERMOD for the conditions in Taiwan, to improve the determination of atmospheric stability and the performance of air quality models. To achieve the abovementioned objective, in these researches, the widely used Penman-Monteith (PM) and Priestley –Taylor (EPA) formulas are being conducted to provide the quantitative basis for atmospheric stability. To verify the performance of these approaches, the measurement data collected by using eddy-covariance technique at Guandu grassland is applied in this study. In this study, Penman-Monteith and Priestley –Taylor formulas are applied to the study area, Guandu. The canopy resistance of Penman-Monteith formular is about 100 (sm-1). The Bowen ratio calculated by Penman-Monteith formula has a higher correlation with the Bowen ratio calculated by eddy covariance. The Bowen ratio calculated by Priestley –Taylor formula is only affected by temperature. Then, input two of the Bowen ratios above to the AERMOD and then take the number calculated by eddy covariance as the real value. We found out that the correlation with pollutant concentration using default Bowen Ratio and real pollutant concentration is irrelevant. Furthermore, the pollutant concentration calculated by default Bowen Ration underestimated much more seriously.