Statistic Study of Relation between the Algal Blooms and Chemicals in Water Body

Abstract

The global trend of the algal blooms is rising fast without any questions. This paper is to investigate whether we can predict the number of the algae in water body by the content of some common chemicals to avoid the algal blooms and, at the same time, use less time and costs. Step regression will be used to filter the independent variables, then use principal component analysis to get rid of multilinearity in the model and use ordinary least squares regression to fit the linear relationship between the principal components chosen and the number of algae. The results show that there is some relationship between the variables studied but not very strong. The R-squared values calculated are 2.86 for the row data and 3.58 for the test data. And the major influential factors are mxPH, mnO2 and PO4. In conclusion, we can use the model to briefly predict the trend of the algal blooms but cannot get the number accurately. However, the results do reveal to us that maximum pH value, minimum value of oxygen and average of orthophosphate indeed have some significant impacts on this kind of algae.