期刊名称：Current Journal of Applied Science and Technology
印刷版ISSN：2457-1024
出版年度：2015
卷号：12
期号：2
页码：1-12
语种：English
出版社：Sciencedomain International
摘要：River Beas is an important river of India, originating in the Himalayas and merging into river Sutlej in Punjab, at Harike, a Ramsar wetland. Twenty two soil characteristics including eight heavy metals were studied at four sampling sites in the vicinity of the river for pre-monsoon, post-monsoon and winter seasons, over a stretch of 63 km from Beas to Harike towns in Punjab, India. Cluster analysis showed that soil characteristics of each site for all the seasons were different. First three principal components in principal component analysis explained 100% of the total variance for pre-monsoon and winter seasons, whereas 99.99% of the total variance for post-monsoon season. In factor analysis, factor-1 accounted for 36% of the total variance and had strong loadings on pH, conductivity, hydrogen, nitrogen and phosphorus. Textural characteristics explained the factor-2 which accounted for 21% of the variance, whereas water holding capacity (WHC), carbon and nitrogen were explained by the factor-3 with 20% of the total variance. In multiple linear regression analysis, reflectance values from bands 2(green), 3(red) and 4(near infra-red) of Landsat (TM) digital data were regressed on pH, phosphorus, potassium, iron, cobalt and manganese. Cobalt and manganese contributed negatively to the reflectance data, whereas pH, phosphorus, potassium and iron enhanced the reflectance data. Artificial neural network models were fitted to the data. Correlations between the target and output values of bands 2(green), 3(red) and 4(near infra-red) were highly significant.
