Multi-model data fusion for hydrological forecasting using Knearest neighbor method

Hydrological forecasting is one the most important issues in water resources systems, which helps in dealing with the real time reservoir operation, flood and drought warning, and irrigation scheduling. While, many conceptual, empirical, and statistical methods have been proposed for hydrological forecasting, rarely a single model is found as the best forecasting model in all hydroclimatological conditions experienced in long-term operation of individual models to improve a forecast skill. As a matter of fact, data fusion provides better forecasts than could be achieved by the use of individual forecast models. This paper presents a comparative assessment of five different methods of data fusion by applying them in a real case study. Furthermore, a new statistical method based on the non-parametric K-nearest neighbor model is presented as an alternative method for data fusion. Results of data fusion approach is thoroughly analyzed and discussed. The results demonstrate that the use of data fusion could significantly improve forecasts in comparison with the use of singe models. Also, data fusion by K-NN method outperforms conventional methods by improving the accuracy and precision of forecasts.

Keywords:

Data Fusion , Nearest Neighbor Method , Streamflow Forecasting , Zayande-rud River