降雨對河川RPI指標之影響－以花蓮溪與秀姑巒溪為例

The Influence of Rainfall on RPI Indicator-Case Study on Hualien River and Siouguluan River

江漢全(Hann-Chyuan Chiang) ； 戴文堅(Wen-Chien Dai) ； 許麗淑(Li-Shu Syu)

《國立宜蘭大學工程學刊》 6期 (2010/03) Pp. 106-117

https://doi.org/10.6176/BCE.2010.06.21

摘要

河川污染指數（River pollution index, RPI）為一廣泛應用之水質指標，其包含溶氧（DO）、生化需氧量（BOD）、懸浮固體（SS）及氨氮（NH3-N）等四項參數，經計算得RPI指數後，將水質分為未（稍）受污染、輕度污染、中度污染及嚴重污染等四個污染程度。四項參數之SS項，易受自然或人為的污染而有劣化情形，花蓮縣內花蓮溪及秀姑巒溪之水質監測站，擁有長期監測資料，本研究係利用上述資料，探討降雨量對花蓮縣河川RPI指標之影響。本研究利用2002－2008年兩河川之月水質及雨量資料，依據豐、枯水期分類，先藉由Spearman相關性分析，了解月雨量、SS及RPI三者間之相關性，再由蒙地卡羅模擬，決定機率分佈函數、繪製累積分佈函數曲線。研究結果顯示，降雨量增加與SS升高呈顯著性正相關，而SS之升高與RPI增加為極顯著性相關。此外，當SS高於100 mg/L時，RPI即達中度污染程度，花蓮溪及秀姑巒溪於枯水期，達此污染程度之機率各為14.90%及9.80%；豐水期則各為42.75%及55.38%。由上述結果可知，降雨量與河川水質變化有顯著關係。因此，建議於豐、枯水期分別訂定SS及RPI之警戒值，以作為稽查河川人為污染之依據，建立更有效的河川水質管理方式。

關鍵字

河川污染指數 ； 懸浮固體物 ； 降雨量 ； Spearman相關性分析 ； 蒙地卡羅模擬

並列摘要

The RPI (River pollution index) is extensively applied to the assessment on river water quality. The RPI involves four parameters: DO, BOD, SS and NH3-N. The overall index is then divided into four pollution levels - unpolluted, light polluted, medium polluted and heavy polluted. The SS is liable to variation because of the effect of natural or artificial pollutions. This study investigated the influence of monthly rainfall on the RPI indicator on Hualien River and Siouguluan River, Hualien County, Taiwan, where water quality monitoring stations have been set for many years, hence have the long-term RPI data.First, we divide the water quality and monthly rainfall data during 2002-2008 from the two rivers into wet period and dry period. Then we get the correlation between each item's monthly rainfall, the SS and the RPI by Spearman's rank correlation analysis. Finally, we find the probability distribution function and cumulative probability distribution curve of the rainfall, the SS and the RPI by Monte Carlo simulation analysis.The results showed that the increased of the rainfall is strongly correlated to the SS, and the increased of the SS is significantly correlated to the RPI. Furthermore, while the SS exceeds 100 mg/L, the RPI may reach medium polluted level, whose probabilities are 14.90 % and 9.80 % in dry period, and 42.75 % and 55.38 % in wet period, respectively. Based on the above analysis, we conclude that the influence of rainfall on river water quality in Hualien is obvious. Therefore, it is useful to manage river water quality by setting warning limits for the SS and the RPI in wet period and dry period.

並列關鍵字

RPI ； SS ； rainfall ； Spearman's rank correlation ； Monte Carlo simulation analysis