CANONIC CORRELATION BETWEEN QUALITY PARAMETERS OF RIVER WATER LOCATED IN RIO GRANDE DO SUL SOUTH

The objective of this work was to evaluate the associations between water quality parameters in the spring / summer period and those established in the autumn / winter period, variables belonging to the Rio Tramandaí / RS. To this end, water quality secondary data provided by the State Environmental...

Full description

Bibliographic Details
Main Authors: Luana Nunes Centeno, Samanta Tolentino Cecconello, Nathan Belluzzo da Silveira, Rodrigo Valandro Mazzaro
Format: Article
Language:English
Published: Instituto Federal de Educação, Ciência e Tecnologia do Rio Grande do Norte 2020-02-01
Series:Holos
Subjects:
Online Access:http://www2.ifrn.edu.br/ojs/index.php/HOLOS/article/view/8858
Description
Summary:The objective of this work was to evaluate the associations between water quality parameters in the spring / summer period and those established in the autumn / winter period, variables belonging to the Rio Tramandaí / RS. To this end, water quality secondary data provided by the State Environmental Protection Foundation Henrique Luis Roessle were used in the period from 2005-1 to 2013-2. The analysis of canonical correlation was applied, considering the original data observed, where canonical charges were estimated. In addition, Wilks' multivariate significance test was used to evaluate the significance of the canonical roots together. Finally, the redundancy index was determined as the mean square canonical charge times the canonical R². It was observed through the applied technique that only the first canonical function was shown to be significant. In relation to the canonical loads, of the function: of the parameters of quality of the water related to the spring / summer period the parameters TC, and the TN presented higher values. Because for the parameters belonging to autumn / winter the TC. presented low canonical charges. The amount of shared variance explained by the dependent canonical statistical variable was, on average, 9.5%. It is observed that the TC had the highest percentage of variance explained by the dependent canonical statistical variable, that is, 0.874. These results demonstrate the possibility of using this multivariate technique to make inferences about quality.
ISSN:1807-1600