Ecology, Environment and Conservation Paper

SENSITIVITY ANALYSIS ON VARIOUS METEOROLOGICAL PARAMETERS RIVER FLOW FORECASTING MODEL USING SUPPORT VECTOR MACHINE

Nuratiah Zaini, M.A. Malek and Marina Yusoff

Abstract

The application of Support Vector Machine in river flow forecasting can further improve the management of water resources and flood mitigation. Other factors that influence river flow other than rainfall are various meteorological parameters namely maximum and minimum temperature, relative humidity, evaporation and mean wind speed. This study focuses on the sensitivity of various meteorological parameters used in river flow forecasting. Among the various input parameters designs available, this study uses Morris one-factor-at-a-time method. The sensitivity of the parameter used is determined by the performance of the developed forecasting model to forecast river flow in terms of various statistical test. Result obtained in this study proved that maximum temperature is found to be the most sensitive parameter in river flow forecasting for training dataset while evaporation is found to be the most sensitive in river flow forecasting for testing dataset.