合成函數在評估雙變量環境風險上之應用研究

Applying Copula Functions in the Evaluation of Bivariate Environmental Risks

潘浙楠(Jen-Nan Pan) ； 紀孟嘉(Meng-Chia Chi)

《品質學報》 15卷1期 (2008/02) Pp. 7-22

摘要

目前我國針對空氣污染罰款的部分是以行政院環保署所制定的「空氣污染防制法」作爲執法的依據。爲促進空氣品質的改善，在環保署所制定之空氣污染防制法中對於固定污染源防制費是採取累進方式收費，即針對不同空氣污染源所產生的污染量係採不同的累進方式計費。針對氮氧化物(NO(下標 X))及硫氧化物(SO(下標 X))兩種不同的空氣污染物，本研究依據現行環保法規先建立適合不同污染物的階梯損失函數(step loss function)再利用合成函數求算此兩種不同污染物同時發生之機率。使用合成函數的優點除了可建立呈高度相關之兩指標污染物不同邊際分配函數的聯合機率函數，使在考慮兩者同時發生之機率後所求得總期望損失較爲精確外，亦容易推展至多變量環境風險之評估。最後，本研究以高雄縣某人造樹脂公司所提供之氮氧化物及硫氧化物連續自動監測記錄資料爲例比較以階梯損失函數、田口單邊二次損失函數及INLF(inverted normal loss function)等方式估算兩個污染物同時發生對工廠所造成潛在風險及損失。

關鍵字

階梯損失函數 ； 空氣污染防制法 ； 合成函數copula

並列摘要

In view of the serious ecological problems caused by an pollution and other contaminations, almost all the countries around the world including the Environmental Protection Administration (EPA) of Taiwan have imposed regulations or laws, such as air pollution control act to reduce and control air pollution problems. In this research, we first establish a suitable step loss function based on various penalties stipulated in the Air Pollution Control Act of Taiwan. The differences among step loss function, Taguchi quadratic and inverted normal loss functions are compared too. Secondly, the risk and potential losses of the two different pollutants occur at the same times are estimated by using the copula and loss functions. The advantage of using copula function is that the joint probability function for different marginal distributions and, the expected total losses can be calculated in a more precise way. Finally, the application of copula function in estimating the risk and potential loss is demonstrated through a set of realistic and data provided by a chemical company located at Kaohsiung county of Taiwan.

並列關鍵字

step loss function ； air pollution control act ； copula function

參考文獻

Chiou, S. C.(2005).Econometrics and Statistics Colloquium.University of Chicago.

Cherbini, U.,Luciano, E.(2002).Bivariate option pricing with copula.Applied Mathematical Finance.8,69-85.

Frechet, M.(1957).Les tableaux de correlations dont les marges sont donnees.Annales de l`Universite de Lyon.20,13-31.

Frees, E. W.,Valdez, E. A.(1998).Understanding relationships using copulas.North American Actuarial Journal.2,1-25.

Sklar, A.(1959).Fonctions de r´epartition `a n dimensions et leurs marges.Publications de l`Institut de Statistique de L`Université de Paris.8,229-231.