Simulation of Metal Contents through Correlated Optimal Monitoring Metals of Dagu River Sediments

Determination of pollutants for river sediments is essential and valuable. To allow a reduction in the number of monitoring metals analyzed in the future monitoring of watercourse, statistic analysis of heavy metals in Dagu River sediments was conducted and a simulation model was developed. Results show that all investigated metals except Zn and Ni in Dagu River sediments are positively correlated with three or more than three other metals. The optimal monitoring metals of Dagu River sediments are Pb (or Cr), Ni, Hg and Zn based on the correlations and their correlation coefficients of metals. To determine metal contents in Dagu River sediments in the future, only detection of the optimal monitoring metals in each sampling transect are required. The pollutant levels of other metals (As, Cd, Cr/Pb and Cu) can be simulated through well correlated metals using the simulation model developed from a nonlinear function of Sigmoid f(x) = 1/(1 + e-x), a BP manual neural network model.