A new approach to predict daily pH in rivers based on the 'a trous' redundant wavelet transform algorithm

Rajaee, Taher and Ravansalar, Masoud and Adamowski, Jan F. and Deo, Ravinesh C. (2018) A new approach to predict daily pH in rivers based on the 'a trous' redundant wavelet transform algorithm. Water, Air and Soil Pollution, 229 (3). ISSN 0049-6979

Abstract

Prediction of pH is an important issue in managing water quality in surface waters (e.g., rivers, lakes) as well as drinking water. The capacity of artificial neural network (ANN), wavelet-artificial neural network (WANN), traditional multiple linear regression (MLR), and wavelet-multiple linear regression (WMLR) models to predict daily pH levels (1, 2, and 3 days ahead) at the Chattahoochee River gauging station (near Atlanta, GA, USA) was assessed. In the proposed WANN model, the original time series of pH and discharge (Q) were decomposed (after being split into training and testing series) into several sub-series by the the à trous (AT) wavelet transform algorithm. The wavelet coefficients were summed to obtain useful input time series for the ANN model to then develop the WANN model for pH prediction. The redundant à trous algorithm was used for data decomposition. Model implementation indicated the values of 1-day-ahead pH predicted by the WANN model closely matched the observed values (with a coefficient of determination, R2 = 0.956; Root Mean Square Error, RMSE = 0.019; and Mean Absolute Error, MAE = 0.015). It is therefore possible that the WANN model’s accuracy can be attributed to its better predictive ability (due to the use of the AT) to remove the noise caused by pH shifts (e.g., acid precipitation). Peak pH values predicted by the WANN model were also closer to observed values compared to the other machine learning models


Statistics for USQ ePrint 33861
Statistics for this ePrint Item
Item Type: Article (Commonwealth Reporting Category C)
Refereed: Yes
Item Status: Live Archive
Additional Information: Permanent restricted access to Published version, in accordance with the copyright policy of the publisher.
Faculty / Department / School: Current - Faculty of Health, Engineering and Sciences - School of Agricultural, Computational and Environmental Sciences
Date Deposited: 21 Mar 2018 06:20
Last Modified: 21 Mar 2018 06:20
Uncontrolled Keywords: artificial neural network; multiple linear regression; ‘a trous’ algorithm; river water; quality pH
Fields of Research : 08 Information and Computing Sciences > 0801 Artificial Intelligence and Image Processing > 080105 Expert Systems
08 Information and Computing Sciences > 0801 Artificial Intelligence and Image Processing > 080108 Neural, Evolutionary and Fuzzy Computation
08 Information and Computing Sciences > 0801 Artificial Intelligence and Image Processing > 080110 Simulation and Modelling
05 Environmental Sciences > 0501 Ecological Applications > 050101 Ecological Impacts of Climate Change
Socio-Economic Objective: D Environment > 96 Environment > 9603 Climate and Climate Change > 960303 Climate Change Models
Identification Number or DOI: 10.1007/s11270-018-3715-3
URI: http://eprints.usq.edu.au/id/eprint/33861

Actions (login required)

View Item Archive Repository Staff Only