http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
박석순,이상희,박배경,이상호 ( Seok Soon Park,Sang Hee Lee,Bae Kyung Park,Sang Ho Lee ) 한국하천호수학회 1995 생태와 환경 Vol.28 No.3
The paper presents an application of nonparametric statistical method to comparison of stream water qualities between two different watersheds : the Seuth and North Branches of the Han River where significant differences have been reported regarding watershed characteristics as well as aquatic biota. Water quality data were obtained from three synoptic field surveys conducted from March to July 1992. The water quality parameters included 5-day biochemical oxygen demand, chemical oxygen demand, pH, suspended solids, anionic surfactants, and conductivity. The normality of tributary water qualities in both watersheds were examined using the boxplot which is a concise graphical summary of the data distribution. The plots showed that most data were typically non-normally distributed as expected. Therefore, nonparametric statistical method called Wilcoxen Rank-Sum Test was adopted in this study. The result indicated that discernable differences of water qualities existed between tributaries of two branches. Highly significant differences were computed in pH and conductivity during all attempted flow conditions. The other parameters, surh as BOD, COD, SS, and AS were much higher in tributaries of the South Branch than in these of the North Branches during low and mid flow periods. During high flow period, however, no statistical differences were computed between two groups of data. Nonparametric aspects of water quality data were also discussed in this paper.
이상호,박석순,박배경 한국물환경학회 1994 한국물환경학회지 Vol.10 No.1
Multiple linear regression analysis was performed to compare stream water qualities relating to land-use types in the Han River watershed. Explanatory variables were ratios of each land use area per total watershed and response variables were water qualities of receiving streams. The computed regression coefficients of each explanatory variables were used as the degrees of water quality impacts by land use types. The watershed was classified into four (4) different land use types : urban and industrial area, agricultural area, forest area, and green belt area (which has been preserved from development). Ninety four (94) tributaries were sampled three times during different flow conditions from March to July, 1992. Water quality parameters included in this study were Biochemical Oxygen Demand, Total Suspended Solids, Anionic Surfactant, and Conductivity. Regression coefficients were computed for each water quality and significance of coefficient values were estimated using the t- test. Linear relationship between explanatory variables and response variable was also measured by the F-test. From this study, it was concluded that there were significant linear relationships between receiving stream water qualities and land use types such that regression analysis could be used as a good comparison tool. The computed values of regression coefficients could be used as quantitative measurements of land use impacts.