請用此 Handle URI 來引用此文件: http://hdl.handle.net/11455/5439
標題: GTx,TPAQM和MM5模式污染物模擬之比較
Comparison of trajectory determinations by GTx,TPAQM and MM5
作者: 謝政育
Hsieh, Cheng Yu
關鍵字: GTx
高斯軌跡傳遞係數模式
TPAQM
MM5
pre-treatment meteorological part
光化學軌跡模式
PSU-NCAR第五代中尺度氣象模式
氣象前處理模組
出版社: 環境工程學系
摘要: 由於台灣已進入工業化的社會,在所處環境的空氣品質越來越惡劣下,為了有效提供環保署相關的空氣污染管制決策,環保署成立空氣品質模式支援中心,並針對於台灣的獨特地型和模擬成果,選取了高斯軌跡傳遞係數模式 (GTx)和光化學軌跡模式 (TPAQM) 為優選模式之一,因此研究中選取了GTx 和TPAQM和全球廣泛使用的PSU-NCAR第五代中尺度氣象模式 (MM5) 三模式,探討其模式所模擬結果之間的相關性。 在研究中選取草屯鎮北投國小和南投市中興新村兩次密集採樣時間,進行模式的相關模擬和比較。在風場方面,在TPAQM和MM5模式所模擬的風場值,與測站觀測值進行統計上的分析,成果如下:TPAQM在U值的RMSE、MAE、BIAS和CORREL分別為1.70、1.23、-0.33和0.57,而在MM5模擬方面分別為3.64、2.75、-1.71和0.42;另外TPAQM在V值的模擬數據表現上分別為1.62、1.06、0.13和0.60,而在MM5則分別為3.00、2.33、-0.88和0.38。由此可以發現TPAQM的表現上皆優於MM5所模擬的風場。 在逆軌跡方面,GTx和MM5所模擬的結果當中,在中部地區常常有山谷風環流效應或繞山機制的產生,但MM5模式對於台灣似乎沒有此機制的產生,因而可能會造成污染物傳輸路線的誤差。 在TPAQM和MM5氣象前處理模組加入GTx主程式中,三模式所模擬的濃度值表現都不佳,其中MM5在SO2的模擬有0.5的相關係數;TPAQM在NOx的模擬有0.44相關係數;GTx在PM10有0.42的正相關;但忽略觀測值的比較時,將發現三模式所模擬的汙染物濃度有不錯的相關性。 在研究中,MM5模式在風場或污染物模擬中常有偏高的模擬值出現,推測由於在邊界條件並無加入觀測值所造成;而實驗所觀測的數值太少,也是造成三模式模擬結果的誤差,因此希望能過有更多的觀測資料來加入三模式之中,進而探討出模式之間不同優劣的機制。
The air quality in Taiwan is more and more abominable in recent years. In order to offer the Environmental Protection Administration (EPA) air pollution control policies effectively, the Environmental Protection Administration establishes the 「Air Quality Modeling Center ,Taiwan」. Besides, for the complex topography in Taiwan, Gaussian Trajectory transfer-coefficient modeling system (GTx) and Trajectory Photochemical Air Quality Model (TPAQM) are considered as the better tool to investigate the issue about air quality. In this study, the correlations between simulations and observations were studied using GTx, TPAQM and the Fifth-Generation NCAR / Penn State Mesoscale Model (MM5). The experiments carried out at Beitou primary school and Jhongsing Village were choose as study cases. The analysis of the wind field are as follows: RMSE, MAE , BIAS and CORREL of U component in TPAQM are 1.70 , 1.23, -0.33 and 0.57 respectively, but 3.64 , 2.75, -1.71 and 0.42 respectively in MM5; besides, V component in TPAQM are 1.62 , 1.06 , 0.13 and 0.60 respectively , but 3.00 , 2.33, -0.88 and 0.38 respectively in MM5. Therefore, it can be found that the display of TPAQM is superior to the wind field of MM5 simulation. In the simulations of back-trajectory, results of GTx and MM5 simulation showed that the alternating drainage wind and upslope wind system are driven by the mountainous in Nan-Tou valley in center Taiwan, but MM5 seems that there is not formulation of this mechanism; therefore may causes the trajectory error. Joining to GTx main program with pre-treatment meteorological part of TPAQM and MM5, the simulations of the three models are not good comparing with observations. The correlation coefficient in SO2 of MM5 simulation is 0.5; the correlation coefficient in NOx of TPAQM simulation is 0.44; the correlation coefficient in PM10 of GTx simulation is 0.42; but while neglecting the comparison of observations, it would be found that concentration of pollutants simulation in the three models are of good mutuality with observations. In this study, the simulations in MM5 often overestimated the wind field or the pollutant simulation. The cause should be there are no observations in the boundary conditions and the insufficient observations. Therefore, more observations are required for complete simulations to compare the different mechanisms of models.
URI: http://hdl.handle.net/11455/5439
顯示於類別:環境工程學系所

文件中的檔案:
沒有與此文件相關的檔案。


在 DSpace 系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。