Abstract
The prediction of discharge and its variability in a river is an essential component of surface-water planning. For that purpose, a functional relationship between stage and discharge is established with the help of field measurement and the relationship is expressed as a rating curve. Whereas direct discharge measurement in rivers is time consuming and expensive, and sometimes it is impossible, for this purpose the relation between discharge and stage as rating curve is estimated by using of measured data. The Qara Su River in Ardabil province, Iran, makes damages with its floods every year. Informing of flood discharge's value in upstream for management activity in downstream is necessary. In this research for estimate stage-discharge relation in Dost Bayglu hydrometry station over Qara Su River, using MATLAB software, Artificial Neural Network with different state and regression methods studied and finally compared together. As a result of Artificial Neural Network with a 1-2-3-2-1 structure, compared with a simple regression, has a better answer.
Key words: Artificial neural network, stage-discharge relation, Qara Su River, Dost Bayglu hydrometry station, regression.