Multiple linear regression analysis (MLR) applied for modeling a new WQI equation for monitoring the water quality of Mirim Lagoon, in the state of Rio Grande do Sul—Brazil

Abstract Accurate assessment of the type and extent water pollution is a difficult complicated task. Therefore, use Water Quality Index (WQI) proves to be useful tool, as this index has advantage resulting in single number that easy communicate understand. One statistical methods can used develop new WQI equation for given body Multiple Linear Regression (MLR). work aims Mirim Lagoon through MLR evaluate suitability equation. For this, 154 samples collected at 7 monitoring points over three years were analyzed, from which original results obtained. With these results, was applied, generating with only variables: phosphorus, dissolved oxygen (DO) thermotolerant coliforms. This model obtained an R 2 coefficient equal 0.728, demonstrating 72.8% data variability explained by generated model. Through paired t test it possible demonstrate did not differ significantly WQI, obtaining significance values greater than 0.05 both absolute quality ranges ( p value 0.886 0.467, respectively), attesting adequacy The reduction variables several advantages, financial ones, makes less costly, even reducing eclipse effect. concluded great importance generate fewer waters Lagoon.