Application of artificial neural networks for river regime

M. D. Bui, D. Huber, A. M.F. Silva, P. Rutschmann

Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review


Predicting the geometric characteristics of a regime channel is of utmost importance in the context of river engineering and management, as regime channels require minimum protection and minimum expenses for their maintenance. There are numerous empirical and analytical methods to predict these geometric characteristics. This paper develops and tests an Artificial Neural Network (ANN) as a model to forecast the river regime characteristics. ANN performance is compared against the Thermodynamic Entropy Theory of Yalin and da Silva (2001) and the Stability Theory of Julien and Wargadalam (1995). An improvement in the results of the ANN model has been achieved by distinguishing the input variables into sand and gravel bed materials as well as different discharge groups.

Original languageEnglish
Title of host publicationRiver Flow 2016
Subtitle of host publicationIowa City, USA, July 11-14, 2016
PublisherCRC Press
Number of pages7
ISBN (Electronic)9781317289128
ISBN (Print)9781315644479
StatePublished - 22 Jun 2016


Dive into the research topics of 'Application of artificial neural networks for river regime'. Together they form a unique fingerprint.

Cite this