Limits in Using Multiresolution Analysis to Forecast Turbidity by Neural Networks. Case Study on the Yport Basin, Normandie-France
Approximately, 25% of the world population drinking water depends on karst aquifers. Nevertheless, due to their poor filtration properties, karst aquifers are very sensitive to pollution and specifically to turbidity. As physical processes involved in transport of solid/suspended particles (advection, diffusion, deposit…) are complicated and badly known in underground conditions, a black-box modeling approach using neural networks is promising. Despite the well-known ability of universal approximation of multilayer perceptron, it appears difficult to efficiently take into account hydrological conditions of the basin. Indeed, these conditions depend both on the initial state of the basin (schematically wet or dry: long timescale component), and on the intensity of rainfall, usually associated to short timescale component. In this context, the present paper addresses the application of the multiresolution analysis to decompose the turbidity on several timescales in order to better consider various phenomena at various timescales (flow in thin or wide fissures for example). Because of “boundary effects”, usually neglected by authors, a specific adaptation was shown as necessary that diminishes the quality of results for real-time forecasting. Decomposing turbidity using multiresolution analysis adds thus questionable improvements.
KeywordsTurbidity Neural networks Forecasting Multiresolution analysis Yport
The authors would like to thank the CODAH for providing rainfall and turbidity data. The Normandie Region and Seine-Normandie Water Agency are thanked for the co-funding of the study. We are also very grateful to S. Lemarie and J. Ratiarson for the very helpful discussions they helped to organize. Our thanks are extended to D. Bertin in the design and implementation of the NN simulation tool. Finally, Marc Steinmann is thanked for careful reading of the present paper.
- Artigue, G., et al. (2012). Flash flood forecasting in poorly gauged basins using neural networks: case study of the Gardon de Mialet basin. NHESS, 12(11), 3307–3324.Google Scholar
- Darras, T., et al. (2014). Influence of the Initialization of Multilayer Perceptron for Flash Floods Forecasting: How Designing a Robust Model, (ITISE 2014), 687–698.Google Scholar
- Dreyfus, G. (2005). Neural networks: methodology and applications. Springer Science & Business Media, 497p.Google Scholar
- Gaillard, T., Hauchard, E., & Roux, J. C. (2012) Les Fontaines d’Yport (Seine-Maritime), émergences majeures du littoral normand crayeux: Exploitation et vulnérabilité de la ressource en eau. Ressources et gestion des aquifères littoraux. Cassis 2012. 235–244.Google Scholar
- Hole, J. P., Roux. J. C., 1978. Vulnérabilité aux pollutions du bassin hydrogéologique des sources d’Yport (Seine-Maritime), Rapport BRGM. 81.Google Scholar
- Hornik, K., Stinchcombe, M., & White, H. (1989). Multilayer feedforward networks are universal approximators. Neural networks, 2(5), 359–366.Google Scholar
- Houria, B., Mahdi, K., & Zohra, T. F. (2014). PSO-ANNs based suspended sediment concentration in Ksob basin, Algeria. Journal of Engineering and Technology Research, 6(8), 129–136.Google Scholar
- Johannet, A., et al. (2012). Prediction of spring discharge by neural networks using orthogonal wavelet decomposition. In IJCNN 2012. https://doi.org/10.1109/IJCNN.2012.6252620.
- Johannet, A., Vayssade, B., & Bertin, D. (2008). Neural networks: from black box towards transparent box. Application to evapotranspiration modeling. International Journal of Computational Intelligence, 4(3), 163–170.Google Scholar
- Kong-A-Siou, L., et al. (2011). Complexity selection of a neural network model for karst flood forecasting: The case of the Lez basin (southern France). Journal of Hydrology 403, 367–380.Google Scholar
- Nash, J. E., & Sutcliffe, J. V. (1970). River flow forecasting through conceptual models part I-A discussion of principles. Journal of hydrology, 10(3), 282–290.Google Scholar
- Nieto, P. G., et al. (2014). Hybrid PSO–SVM-based method for long-term forecasting of turbidity in the Nalón river basin: A case study in Northern Spain. Ec. Engineering, 73, 192–200.Google Scholar
- SAFEGE (2012) Étude du Bassin d’Alimentation du forage d’Yport “Phase 1 :Délimitation du bassin d’alimentation et analyse de la vulnérabilité intrinsèque du basin”. 112p.Google Scholar
- Savary M., et al. (2017) Operational Turbidity Forecast Using Both Recurrent and Feed-Forward Based Multilayer Perceptrons. In: Rojas I., Pomares H., Valenzuela O. (eds) Advances in Time Series Analysis and Forecasting. ITISE 2016.Google Scholar
- Savary, M., PhD Dissertation (2018), Prévision de turbidité par apprentissage statistique : application au captage AEP d'Yport (Normandie). Thèse de doctorat de l'Université de Normandie en Sciences de l'univers, Ecole doctorale BISE, soutenue le 12 juillet 2018.Google Scholar