Introducing a New Long-Lead Hydrologic Forecasting System for Improving Reservoir Operation

Long-lead streamflow forecasting plays an important role in water resources planning and management. In this paper a new forecasting system named FARDA (Forecasting and Related Decision Analysis) is introduced. The results of the application of this system to two great river basins of Iran, namely Karkheh, and Karun are presented, briefly. Three data-driven models including K-Nearest Neighbor Regression (KNN), Artificial Neural Network (ANN), and monthly Rainfall-Runoff (R-R) models are performed within the system as individual forecasting models (IFMs). The fusion of all IFMs best outputs resulted from ordered series of model outputs is applied within the system to report the most reliable forecasts. All the forecasting models are presented in the model base of FARDA. Furthermore, the model base of the system consists of reservoir operation models which get benefit from the outputs of forecasting models to provide the best operating policies for the system of reservoirs. The inputs of those models are supported by the data base of the system which consists of different types of local and global hydro-climatological data as well as dams’ data and information. This paper presents some characteristics of the system such as its conceptual model, the framework of its data and model base, and its specific graphical user interface. Some of the results of the application of the system in recent years is also presented.