Nonlinear Processes in Geophysics www.nonlin-processes-geophys.net 1023-5809 1607-7946 17 5 2010 10.5194/npg-17-395-2010 http://www.nonlin-processes-geophys.net/17/395/2010/ http://www.nonlin-processes-geophys.net/17/395/2010/npg-17-395-2010.html http://www.nonlin-processes-geophys.net/17/395/2010/npg-17-395-2010.pdf 395 404 2010-09-02 The use of artificial neural networks to analyze and predict alongshore sediment transport B. van Maanen b.vanmaanen@niwa.co.nz G. Coco K. R. Bryan B. G. Ruessink National Institute of Water and Atmospheric Research (NIWA), P.O. Box 11-115, Hamilton, New Zealand Department of Earth and Ocean Sciences, University of Waikato, Private Bag 3105, Hamilton, New Zealand Department of Physical Geography, Faculty of Geosciences, Institute for Marine and Atmospheric Research, Utrecht University, The Netherlands An artificial neural network (ANN) was developed to predict the depth-integrated alongshore suspended sediment transport rate using 4 input variables (water depth, wave height and period, and alongshore velocity). The ANN was trained and validated using a dataset obtained on the intertidal beach of Egmond aan Zee, the Netherlands. Root-mean-square deviation between observations and predictions was calculated to show that, for this specific dataset, the ANN (ε_rms=0.43) outperforms the commonly used Bailard (1981) formula (ε_rms=1.63), even when this formula is calibrated (ε_rms=0.66). Because of correlations between input variables, the predictive quality of the ANN can be improved further by considering only 3 out of the 4 available input variables (ε_rms=0.39). Finally, we use the partial derivatives method to "open and lighten" the generated ANNs with the purpose of showing that, although specific to the dataset in question, they are not "black-box" type models and can be used to analyze the physical processes associated with alongshore sediment transport. In this case, the alongshore component of the velocity, by itself or in combination with other input variables, has the largest explanatory power. Moreover, the behaviour of the ANN indicates that predictions can be unphysical and therefore unreliable when the input lies outside the parameter space over which the ANN has been developed. Our approach of combining the strong predictive power of ANNs with "lightening" the black box and testing its sensitivity, demonstrates that the use of an ANN approach can result in the development of generalized models of suspended sediment transport. Bailard, J. A.: An energetics total load sediment transport model for a plane sloping beach, J. Geophys. Res., 86, 10938–10954, 1981. Bayram, A., Larson, M., Miller, H. C., and Kraus, N. C.: Cross-shore distribution of longshore sediment transport: comparison between predictive formulas and field measurements, Coast. Eng., 44, 79–99, 2001. Benítez, J. M., Castro, J. L., and Requena, I.: Are artifical neural networks black boxes?, IEEE Transactions on Neural Networks, 8, 1156–1164, 1997. Browne, M., Castelle, B., Strauss, D., Tomlinson, R., Blumenstein, M., and Lane, C.: Near-shore swell estimation from a global wind-wave model: Spectral process, linear and artificial neural network models, Coast. Eng., 54, 445–460, 2007. Burnham, K. P. and Anderson, D. R.: Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach, Springer, New York, NY, 488 pp., 2002. Dayhoff, J. E. and DeLeo, J. M.: Artificial neural networks: Opening the black box, Cancer, 91, 1615–1635, 2001. Dimopoulos, I., Bourret, P., and Lek, S.: Use of some sensitivity criteria for choosing networks with good generalization ability, Neural Process. Lett., 2, 1–4, 1995. Dimopoulos, I., Chronopoulos, J., Chronopoulos-Sereli, A., and Lek, S.: Neural network models to study relationships between lead concentration in grasses and permanent urban descriptors in Athens city (Greece), Ecol. Model., 120, 157–165, 1999. Faraway, J. and Chatfield, C.: Time series forecasting with neural networks: a comparative study using the airline data, Appl. Statistics, 47, 231–250, 1998. Gardner, M. W. and Dorling, S. R.: Artificial neural networks (the multilayer perceptron) – a review of applications in the atmospheric sciences, Atmos. Environ., 32, 2627–2636, 1998. Gevrey, M., Dimopoulos, I., and Lek, S.: Review and comparison of methods to study the contribution of variables in artificial neural network models, Ecol. Model., 160, 249–264, 2003. Gevrey, M., Dimopoulos, I., and Lek, S.: Two-way interaction of input variables in the sensitivity analysis of neural network models, Ecol. Model., 195, 43–50, 2006. Green, M. O. and Boon, J. D.: The measurement of constituent concentration in nonhomogeneous sediment suspensions using optical backscatter sensors, Mar. Geol., 110, 73–81, 1993. Kingston, K. S., Ruessink, B. G., van Enckevort, I. M. J., and Davidson, M. A.: Artificial neural network correction of remotely sensed sandbar location, Mar. Geol., 169, 137–160, 2000. Kolen, J. F. and Pollack, J. B.: Back propagation is sensitive to initial conditions, Complex Systems, 4, 269–280, 1990. Krasnopolsky, V. M.: Neural network emulations for complex multidimensional geophysical mappings: Applications of neural network techniques to atmospheric and oceanic satellite retrievals and numerical modeling, Rev. Geophys., 45, RG3009, doi:10.1029/2006RG000200, 2007. Lin, B. and Namin, M. M.: Modelling suspended sediment transport using an integrated numerical and ANNs model, J. Hydraul. Res., 43, 302–310, 2005. McCann, D. W.: A neural network short-term forecast of significant thunderstorms, Forecasting Techniques, 7, 525–534, 1992. Nagy, H. M., Watanabe, K., and Hirano, M.: Prediction of sediment load concentration in rivers using artificial neural network model, J. Hydraulic Eng., 128, 588–595, 2002. Olden, J. D.: An artificial neural network approach for studying phytoplankton succession, Hydrobiologia, 436, 131–143, 2000. Olden, J. D. and Jackson, D. A.: Illuminating the "black box": a randomization approach for understanding contributions in artificial neural networks, Ecol. Model., 154, 135–150 2002. Olden, J. D., Joy, M. K., and Death, R. G.: An accurate comparison of methods for quantifying variable importance in artificial neural networks using simulated data, Ecol. Model., 178, 389–397, 2004. Pape, L., Ruessink, B. G., Wiering, M. A., and Turner, I. L.: Recurrent neural network modeling of nearshore sandbar behavior, Neural Networks, 20, 509–518, 2007. Puleo, J. A., Johnson, R. V., Butt, T., Kooney, T. N., and Holland, K. T.: The effect of air bubbles on optical backscatter sensors, Mar. Geol., 230, 86–96, 2006. Rumelhart, D. E., Hinton, G. E., and Williams, R. J.: Learning representations by back-propagating errors, Nature, 323, 533–536, 1986. Sztobryn, M.: Forecast of storm surge by means of artificial neural network, J. Sea Res., 49, 317–322, 2003. Tsai, C. P. and Lee, T.: Back-propagation neural network in tidal-level forecasting, Journal of Waterway, Port, Coastal and Ocean Engineering, 125, 195–202, 1999. Tsai, C. P., Hsu, J. R. C., and Pan, K. L.: Prediction of storm-built beach profile parameters using neural networks, in Proceedings of the 27th International Conference on Coastal Engineering, ASCE, 3048–3061, 2000. Van Maanen, B., de Ruiter, P. J., and Ruessink, B. G.: An evaluation of two alongshore transport equations with field measurements, Coast. Eng., 56, 313–319, 2009. Van Rijn, L. C.: Sediment transport, part II: suspended load transport, J. Hydraulic Engineering, 110, 1613–1641, 1984. Vaughn, M. L.: Interpretation and knowledge discovery from the multilayer perceptron network: opening the black box, Neural Comput. Appl., 4, 72–82, 1996. Werbos, P. J.: Beyond regression: new tools for prediction and analysis in the behavioral sciences, PhD thesis, Cambridge, (MA), Harvard Univ., 1974.