Denoising with wavelet analysis and its application to estimate water quality model parameters

The carbon biochemical oxygen demand, nitrogen biochemical oxygen demand and dissolved oxygen deficit concentrations were calculated and were regarded as three groups of original signals in a synthetic river water quality system. The denoising original water quality signals unpolluted and that polluted by random noise with the wavelet analysis method through the multi-resolution technique shows that the superior denoising effect of wavelet analysis method will not distort any water quality signals unpolluted or polluted which are used for the estimation of water quality model parameters. The signals denoised by the Fourier analysis method and the signals non-denoised were also done for comparison.