ANN Modeling for Estimation of Surface and Subsurface Flows Based on Watershed Geomorphology

	ANN Modeling for Estimation of Surface and Subsurface Flows Based on Watershed Geomorphology
Journal of Agricultural Science and Technology
Article 5, Volume 9, Issue 4, 2007, Pages 303-316 PDF (348.77 K)
Authors
M. R. Najafi^* ¹; K. T. Lee²; S. M. Hosseini¹
¹Department of Irrigation, Faculty of Soil and Water Engineering, Campus of Agriculture and Natural Resources, University of Tehran, Karaj, Box: 31587-11167, Islamic Republic of Iran.
²Department of River and Harbor Engineering, National Taiwan Ocean University, 2 Bee-Ning Road, Keelung, Taiwan 202, ROC.
Abstract
In recent years, artificial neural networks (ANNs) have been widely used for flood esti-mation. In this study, an ANN model based on the geomorphologic characteristics of a watershed such as the number of possible paths and their probabilities is developed (GANN model). Nodes in the input layer are allocated to the surface flows, subsurface flows, excess-rainfall and infiltrated rain. The number of nodes related to excess rainfall is predetermined according to the time of concentration of the watershed. The dependability of the infiltrated rain and surface and subsurface flows on previous time steps are calcu-lated by assigning a different number of nodes to each component. The results of the study showed that the simulated hydrographs by the proposed ANN model have good agreement with the hydrographs observed.
Statistics Article View: 101 PDF Download: 72