期刊名称：International Research Journal of Environmental Sciences
电子版ISSN：2319-1414
出版年度：2016
卷号：5
期号：12
页码：7-13
语种：English
出版社：International Science Community Association
摘要：The hydrologic behaviour of any basin/watershed is studied by the hydrologic modelling. Principal component analysis (PCA) is a statistically method which uses an orthogonal transformation to convert a set of observations of possibly correlated variables into a set of values the of linearly uncorrelated variables called principal components. The PCA has been applied for 13 dimensionless geomorphic parameters of 11 selected watershed in upper and middle Godavari sub basin, Maharashtra (India) in order to group the parameters in different components on the basis of their significant correlations. Results of the PCA shows that first two PC are strongly correlated with some geomorphic parameters. However, the third PC is not found strongly correlated with any of the parameters but is moderately correlated with length width ratio (Lb/Lw). The result clearly shows that due to poor correlation with other, the hypsometric integral and main stream channel slope could not be grouped with any of the component. The principal component loadings matrix obtained using correlation matrix of finally selected eleven parameters reveals that first three components together account for 94.283 % of the total explained variance. The results shows that the PCA is good tool for screening out the insignificant parameters in the study of watersheds hydrologic behaviour like runoff and sediment yield modelling. Therefore, principal component lading matrix is applied in order to get better correlations and clearly grouped the parameters into physically significant components.
