DONATE

IAHR Document Library


« Back to Library Homepage « Proceedings of the 31st IAHR World Congress (Seoul, 2005)

Short-Term Prediction of Beach Morphology Using Artificial Neural Networks

Author(s): Bunchingiv Bazartseren; K. Peter Holz

Linked Author(s): Klaus-Peter Holz

Keywords: Morphodynamics; Beach morphology changes; Neural networks; Prediction

Abstract: This paper reports results of using Artificial Neural Networks (ANN) for simulation of beach morphology changes. The authors investigated a possibility to derive the morphology changes on individual points of a beach profile by using information of surrounding bathymetry points at preceding measurement instances. Since the time intervals between the bathymetry measurements are different in this case study, based on the data from the Kiel Bay, Germany, it was attempted to take the temporal scales into account by weighting. The results are then compared to those obtained by using elevations of neighboring points of the preceding measurement instances directly. Up to three preceding measurements are used for deriving development tendency. For more complex profile geometry with several sandbars, introducing weighted values appears to improve the model performance.

DOI:

Year: 2005

Copyright © 2024 International Association for Hydro-Environment Engineering and Research. All rights reserved. | Terms and Conditions