Runoff Prediction by Support Vector Machine for Chalous River Basin of Iran

Abstract

Runoff is the result from the comprehensive action of climate conditions and drainage area underlying surface. Rainfall, evaporation, temperature, wind speed, solar radiation and relative humidity are the most important factor which effect on runoff. Prediction of runoff amounts is performed using Support Vector Machine (SVM). In this paper, the prediction of runoff for Chalous River basin along the Caspian Sea is investigated. A model based on SVM approach is proposed to runoff, predicated on a total of 8 years daily data sets, including field investigation records for the Chalous River Basin along the southern shoreline of Caspian Sea. This study addresses the question of whether Support Vector Machine (SVM) approach could be used to predict runoff. Results revealed that SVM provides an effective means of efficiently recognizing accurately predicting the runoff and the prediction of the future runoff evolution trend with this model will provide the basis for water regulation and water resources reasonable configuration.