Water Quality Indexes Study Based on Watershed Mechanism

• Main Authors: Ming-Yuan Lu, 呂銘淵
• Other Authors: 陳鴻烈
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/87629593426626789390

碩士 === 國立中興大學 === 水土保持學系所 === 96 === The major target in this study is to find out the best water quality index by quantifying water quality mechanism of watershed. This study could be divided to three parts. The first part is to analyze the mechanism of RPI by multivariate analysis. The results show that there was very serious pollution in Dansuie River watershed. Even the water quality in Feitsui Reservoir was also seriously polluted and cannot reach the standards for water supply. The results of mechanism analysis show that the most important parameters in RPI are BOD and ammonia. The best regression model built by these two parameters could best explain RPI index. The second part is to analyze the mechanism of WQI index by the same processes in RPI. The results show that the major mechanism in WQI was explained by conductivity, dissolved oxygen, nitrate, and Total Kjeldahl Nitrogen. The high R2 value equals 0.8496 in the best regression model built by these four parameters. The third part is to compare RPI and WQI based on the analysis results of index mechanism. In the qualitative part, we tried to compare two indexes in index structure, calculating procedures, and standards for quality judgment. All the results in qualitative comparison prove that WQI is better than RPI. In quantitative part, we compared all results in the mechanism analysis. In selecting the best regression model, the results show that the adjust R2 in WQI is 0.834 and it is 0.608 in RPI. So it obviously proves WQI is better. We got the same results in the analysis of the best principle component regression model. The adjust R2 is 0.829 in WQI and 0.492 in RPI. In the results of principal component regression analysis also show the WQI could well catch the true water quality situation. The adjust R2 in the best model is 0.8292. It is very close to the adjust R2 in the best regression model of WQI which is 0.834. After all the comparisons in quantitative data, we got many evidences to prove that WQI is better than RPI whatever in the index sensitivity, completeness, ability of explaining. In this study, we proved that WQI is really better than RPI depended on qualitative mechanism. Therefore, we suggest that the organizations which are related to water quality protection should consider executing the WQI index. The reason is this index is closer to the truth of water quality in watershed.