An Evaluation of À Trous-Based Record Extension Techniques for Water Quality Record Extension

Hydrological data in general and water quality (WQ) particular frequently suffer from missing records and/or short-gauged monitoring/sampling sites. Many statistical regression techniques are employed to substitute values or extend at sites, such as the Kendall-Theil robust line (KTRL), its modified version (KTRL2), ordinary least squares (OLS), four MOVE techniques, of organic correlation (RLOC). In this study, aspiring achieve better accuracy precision, À Trous-Haar wavelet transform (WT) was adopted a denoising preprocessing step prior applying record extension techniques. An empirical study performed using real WQ data, National monitoring network Nile Delta Egypt, evaluate performance these eight record-extension with without WT step. Evaluations included precision when used for restoration variable. The results indicated that values, KTRL WT-KTRL outperformed other However, variables, WT-KTRL2, WT-MOVE3, WT-MOVE4 showed more accurate precise compared both their counterparts WT.