Conference Agenda

Forecasting river flow is crucial for providing early warnings and preparing for sudden floods. River flow is mainly affected by rainfall within the basin, leading to the development of various streamflow prediction techniques based on rainfall data. However, predicting total runoff directly from rainfall often results in inaccuracies, as rainfall during dry soil conditions contributes minimally to streamflow. To address this limitation, an improved method is proposed to calculate direct runoff by focusing on effective rainfall. By combining this with base flow contributions, the total streamflow can be accurately estimated, enhancing the reliability of flood forecasting models. In this study, the correlation between effective rainfall and direct runoff, as well as the correlation between total rainfall and total runoff, is analysed and compared. Based on the results of the correlation analysis in this study, Machine Learning (ML) techniques will be applied to achieve the final goal of predicting streamflow in future studies. It is expected that this approach will enhance the efficiency of flood early warning systems, providing more reliable tools for flood preparedness and management.