Ensemble model to enhance robustness of flash flood forecasting using an Artificial Neural Network: case-study on the Gardon Basin (south-eastern France)
DOI:
https://doi.org/10.21701/bolgeomin.129.3.007Keywords:
cross validation, flash flood, forecasting, model selection, neural networkAbstract
During the last few decades neural networks have been increasingly used in hydrological modelling for their fundamental property of parsimony and of universal approximation of non-linear functions. For the purpose of flash flood forecasting, feed-forward and recurrent multi-layer perceptrons appear to be efficient tools. Nevertheless, their forecasting performances are sensitive to the initialization of the network parameters. We have studied the cross-validation efficiency to select initialization providing the best forecasts in real time situation. Sensitivity to initialization of feed-forward and recurrent models is compared for one-hour lead-time forecasts. This study shows that cross-validation is unable to select the best initialization. A more robust model has been designed using the median of several models outputs; in this context, this paper analyses the design of the ensemble model for both recurrent and feed-forward models.
Downloads
References
Abrahart, R. J. and See, L. M. 1997. Neural network model-ling of non-linear hydrological relationships,Hydrological Earth System Sciences, 11 (5), 1563-1579. https://doi.org/10.5194/hess-11-1563-2007
Anctil, F., Lauzon, N. and Filion, M. 2008. Added gains of soilmoisture content observations for streamflow predic-tions using neural networks, Journal of Hydrology,359(3-4), 225-234. https://doi.org/10.1016/j.jhydrol.2008.07.003
Artigue, G., Johannet, A., Borrell, V. and Pistre, S. 2012.Flash flood forecasting in poorly gauged basins usingneural networks: case study of the Gardon de Mialetbasin (southern France), Natural Hazards Earth SystemSciences, 12 (11), 3307-3324. https://doi.org/10.5194/nhess-12-3307-2012
Ayral, P. A. 2005. Contribution à la spatialisation du modèleopérationnel de prévision des crues éclair ALHTAÏR,Université de Provence Aix-Marseille.
Barron, A. R. 1993. Universal approximation bounds forsuperpositions of a sigmoidal function, IEEETransactions on Information Theory, 39 (3), 930-945. https://doi.org/10.1109/18.256500
Borga, M., Anagnostou, E. N., Blöschl, G. and Creutin, J. D.2011. Flash flood forecasting, warning and risk manage-ment: the HYDRATE project, Environmental SciencePolicy, 14 (7), 834-844. https://doi.org/10.1016/j.envsci.2011.05.017
Bornancin-Plantier, A., Johannet, A., Borrell Estupina, V.,Roussel-Ragot, P. and Dreyfus, G. 2011. Conception demodèles de prévision des crues éclair par apprentissageartificiel, in EGU2011-1794, 2011, vol. 13.C
osandey, C. and Robinson, M.: Hydrologie continentale, A.Colin., 2000.Dawson, C. W. and Wilby, R. L. 2001. Hydrological modellingusing artificial neural networks, Progress in PhysicalGeography, 25 (1), 80-108. https://doi.org/10.1191/030913301674775671
Diettrich, T. G. 2015. Ensemble Methods in MachineLearning, in Lecture Notes in Computer Science, p. 115,Springer Verlag, New-York. [online] Available from:http://www.eecs.wsu.edu/~holder/courses/CptS570/fall07/papers/Dietterich00.pdf (Accessed 16 June 2015).
Dreyfus, G. 2005. Neural Networks: Methodology andApplications, Softcover reprint of hardcover 1st ed. 2005edition., Springer, Berlin; New York.
Garambois, P. A., Larnier, K., Roux, H., Labat, D. and Dartus,D. 2014. Analysis of flash flood-triggering rainfall for aprocess-oriented hydrological model, AtmosphericResearch, 137, 14-24. https://doi.org/10.1016/j.atmosres.2013.09.016
Gaume, E., Bain, V., Bernardara, P., Newinger, O., Barbuc,M., Bateman, A., Blaškoviová, L., Blöschl, G., Borga, M.,Dumitrescu, A., Daliakopoulos, I., Garcia, J., Irimescu, A.,Kohnova, S., Koutroulis, A., Marchi, L., Matreata, S.,Medina, V., Preciso, E., Sempere-Torres, D., Stancalie, G.,Szolgay, J., Tsanis, I., Velasco, D. and Viglione, A. 2009. Acompilation of data on European flash floods, Journal ofHydrology, 367 (1-2), 70-78. https://doi.org/10.1016/j.jhydrol.2008.12.028
Geman, S., Bienenstock, E. and Doursat, R. 1992. NeuralNetworks and the Bias/Variance Dilemma, NeuralComputing, 4 (1), 1-58. https://doi.org/10.1162/neco.1992.4.1.1
Hornik, K., Stinchcombe, M. and White, H. 1989. Multilayerfeedforward networks are universal approximators,Neural Networks, 2 (5), 359-366. https://doi.org/10.1016/0893-6080(89)90020-8
Huet, P., Martin, X., Prime, J.-L., Foin, P., Laurain, C. andCannard, P. 2003. Retour d'expériences des crues deseptembre 2002 dans les départements du Gard, del'Hérault, du Vaucluse, des Bouches-du-Rhône, del'Ardèche et de la Drôme., Inspection générale del'Environnement, Paris, France. [online] Available from:http://cgedd.documentation.developpement-durable.gouv.fr/document.xsp?id=Cgpc-OUV00000419(Accessed 24 March 2015).
Kitanidis, P. K. and Bras, R. L. 1980. Real-time forecastingwith a conceptual hydrologic model: 2. Applications andresults, Water Resources Research, 16(6), 1034-1044. https://doi.org/10.1029/WR016i006p01034
Kong-A-Siou, L., Johannet, A., Borrell, V. and Pistre, S. 2011.Complexity selection of a neural network model forkarst flood forecasting: The case of the Lez Basin (south-ern France). Journal of Hydrology. 403 (3-4), 367-380. https://doi.org/10.1016/j.jhydrol.2011.04.015
Kong-A-Siou, L., Johannet, A., Valérie, B. E. and Pistre, S.2012. Optimization of the generalization capability forrainfall-runoff modeling by neural networks: the case ofthe Lez aquifer (southern France), Environmental EarthSciences, 65 (8), 2365-2375. https://doi.org/10.1007/s12665-011-1450-9
Kong-A-Siou, L., Fleury, P., Johannet, A., Borrell Estupina, V.,Pistre, S. and Dörfliger N. , 2014. Performance and com-plementarity of two systemic models (reservoir andneural networks) used to simulate spring discharge andpiezometry for a karst aquifer. Journal of Hydrology, 519(D), 3178-3192. https://doi.org/10.1016/j.jhydrol.2014.10.041
Le Lay, M. and Saulnier, G. M. 2007. Exploring the signatureof climate and landscape spatial variabilities in flashflood events: Case of the 8-9 September 2002 Cévennes-Vivarais catastrophic event, Geophysics Res.Letters, 34(13) [online] https://doi.org/10.1029/2007GL029746
Llasat, M. C., Llasat-Botija, M., Prat, M. A., Porcú, F., Price, C.,Mugnai, A., Lagouvardos, K., Kotroni, V., Katsanos, D.,Michaelides, S. and others 2010. High-impact floods andflash floods in Mediterranean countries: the FLASH pre-liminary database, Advances in Geosciences, 23 (23),47-55. https://doi.org/10.5194/adgeo-23-47-2010
Llasat, M. C., Marcos, R., Llasat-Botija, M., Gilabert, J.,Turco, M. and Quintana-Seguí, P. 2014. Flash flood evo-lution in North-Western Mediterranean, AtmosphericResearch, 149, 230-243. https://doi.org/10.1016/j.atmosres.2014.05.024
Marchandise, A. 2007. Modélisation hydrologique distribuéesur le Gardon d'Anduze; étude comparative de différentsmodèles pluie-débit, extrapolation de la normale à l'ex-trême et tests d'hypothèses sur les processushydrologiques, Université Montpellier II-Sciences etTechniques du Languedoc. [online] Available from:http://www.ohmcv.fr/Documents/theses/these_marchan-dise-old.pdf (Accessed 8 December 2014).
Montz, B. E. and Gruntfest, E. 2002. Flash flood mitigation:recommendations for research and applications, GlobalEnvironmental Change Part B Environmental Hazards, 4(1), 15-22. https://doi.org/10.1016/S1464-2867(02)00011-6
Moussa, R. 2010. When monstrosity can be beautiful whilenormality can be ugly: assessing the performance ofevent-based flood models, Hydrology Sciences Journal,55, 1074-1084. https://doi.org/10.1080/02626667.2010.505893
Moussa, R., Chahinian, N. and Bocquillon, C. 2007.Distributed hydrological modelling of a Mediterraneanmountainous catchment - Model construction andmulti-site validation, Journal of Hydrology, 337 (1-2),35-51 https://doi.org/10.1016/j.jhydrol.2007.01.028
Nash, Je. and Sutcliffe, J. V., 1970. River flow forecastingthrough conceptual models part I-A discussion of prin-ciples, Journal of Hydrology, 10 (3), 282-290. https://doi.org/10.1016/0022-1694(70)90255-6
Nikolopoulos, E. I., Anagnostou, E. N., Borga, M., Vivoni, E.R. and Papadopoulos, A. 2011. Sensitivity of a mountainbasin flash flood to initial wetness condition and rainfallvariability, Journal of Hydrology, 402 (3-4), 165-178 https://doi.org/10.1016/j.jhydrol.2010.12.020
Price, C., Yair, Y., Mugnai, A., Lagouvardos, K., Llasat, M. C.,Michaelides, S., Dayan, U., Dietrich, S., Galanti, E.,Garrote, L., Harats, N., Katsanos, D., Kohn, M., Kotroni,V., Llasat-Botija, M., Lynn, B., Mediero, L., Morin, E.,Nicolaides, K., Rozalis, S., Savvidou, K. and Ziv, B. 2011.The FLASH Project: using lightning data to better under-stand and predict flash floods, Environmental SciencePolicy, 14 (7), 898-911. https://doi.org/10.1016/j.envsci.2011.03.004
Schmidhuber, J., 2015. Deep Learning in Neural Networks:An Overview. Neural Networks,61, 85-117. https://doi.org/10.1016/j.neunet.2014.09.003 PMid:25462637
SIEE, 2004. Validation des relevés hydrométriques del'événement des 8 & 9 septembre 2002, DirectionDépartementale de l'Equipement du Gard.
Sjöberg, J., Zhang, Q., Ljung, L., Benveniste, A., Delyon, B.,Glorennec, P.-Y., Hjalmarsson, H. and Juditsky, A. 1995.Nonlinear black-box modeling in system identification: aunified overview, Automatica, 31 (12), 1691-1724. https://doi.org/10.1016/0005-1098(95)00120-8
Stone, M. 1974. Cross-Validatory Choice and Assessment ofStatistical Predictions, Journal of the Royal StatisticalSociety Series B Methodology, 36 (2), 111-147. https://doi.org/10.1111/j.2517-6161.1974.tb00994.x
Toukourou, M., Johannet, A., Dreyfus, G. and Ayral, P.-A.2011. Rainfall-runoff modeling of flash floods in theabsence of rainfall forecasts: the case of "Cévenol flashfloods," Applied Intelligence, 35 (2), 178-189. https://doi.org/10.1007/s10489-010-0210-y
Tramblay, Y., Bouvier, C., Martin, C., Didon-Lescot, J.-F.,Todorovik, D. and Domergue, J.-M. 2010. Assessment ofinitial soil moisture conditions for event-based rain-fall-runoff modelling, Journal of Hydrology, 387 (3-4),176-187 https://doi.org/10.1016/j.jhydrol.2010.04.006
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2024 Consejo Superior de Investigaciones Científicas (CSIC)
This work is licensed under a Creative Commons Attribution 4.0 International License.
© CSIC. Manuscripts published in both the print and online versions of this journal are the property of the Consejo Superior de Investigaciones Científicas, and quoting this source is a requirement for any partial or full reproduction.
All contents of this electronic edition, except where otherwise noted, are distributed under a Creative Commons Attribution 4.0 International (CC BY 4.0) licence. You may read the basic information and the legal text of the licence. The indication of the CC BY 4.0 licence must be expressly stated in this way when necessary.
Self-archiving in repositories, personal webpages or similar, of any version other than the final version of the work produced by the publisher, is not allowed.
