Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Egypte → Rainfall-Runof Modeling Using Artifical Neural Network Technique Case Study : Blue Nile Catchment

Cairo University (2001)

Rainfall-Runof Modeling Using Artifical Neural Network Technique Case Study : Blue Nile Catchment

Mamdouh Ahmed Antar Sayed

Titre : Rainfall-Runof Modeling Using Artifical Neural Network Technique Case Study : Blue Nile Catchment

Auteur : Mamdouh Ahmed Antar Sayed

Etablissement de soutenance : Cairo University

Grade : Doctor of Philosophy (PhD) 2001

Résumé
The thesis presents an artificial neural network (ANN) rainfall-runoff model. The ANN modeling approach overcomes most of the problems associated with the implementation of physical rainfall runoff models. Throughout the research, the Blue Nile catchment is divided based on the main tributaries and the digital elevation data using the GIS tool into seven subcatchments. The mean areal precipitation over those subcatchments are produced as a main input to the ANN model. The performance of ANN model in validation experiment is compared with a physical distributed rainfall-runoff model that apply hydraulic and hydrologic fundamental equations in a grid base. To evaluate and compare between different results, statistical measures are used throughout the study. By applying the proposed methodology, the results over the case study area indicates that, ANN technique showed a great potential in representing and describing the rainfall-runoff relationship based on the available level of details of data and in contrast with other models. The thesis shows that the rainfall-runoff relationship over the Blue Nile basin can be presented sufficiently using only three layers of ANN model.

Mots Clés : Rainfall, Runoff, Physical models, Remote sensing, Neural Network

Présentation

Page publiée le 5 avril 2020