The use of an Artificial Neural Network tool (ANN) to predict wave overtopping discharges is particularly recommended for structures with complicated geometry that do not fall into the categories of relatively simple dike-type structures, rubble mound slopes or vertical walls. An improved Artificial Neural Network has been developed by the University of Bologna in cooperation with the co-authors of the Manual. This Neural Network is the tool officially adopted by EurOtop (2016).
The Overtopping Neural Network allows you to predict the hydraulic performance of coastal and harbour structures in terms of wave overtopping discharge (q), wave transmission coefficient (Kt) and wave reflection coefficient (Kr). The use of the Neural Network is web-based and is free upon registration. The Neural Network has been based on the CLASH-database that has been extended to more than 13,000 tests on wave overtopping over all kind of structures, now called the EurOtop-database.
Questions on the use of the Neural Network may be directed to the developers at University of Bologna. Questions that take less than 5 minutes will be answered for free. Questions on use and on results that take more than 5 minutes to answer may be charged at an hourly fee.
EurOtop ANN and EurOtop-database
www.unibo.it/overtopping-neuralnetwork/
Above is the link to the Neural Network as well as the EurOtop-database.
CLASH ANN and database
The Neural Network that has been developed in CLASH is still available and is also web-based. Note that this ANN has been built on a smaller number of tests than the EurOtop ANN. A main difference between both ANN’s may be that the CLASH ANN cannot predict very small overtopping discharges.
Link to Deltares website and ANN:
www.deltares.nl/en/software/overtopping-neural-network/
Direct download of CLASH database:
Database_20050101.xls 4.3 MB