ABSTRACT
The concept of random variables has been widely used for over a century in the field of hydrology and water resources analysis and modeling. Great improvements have been made in recent years in the following fields:
• Understanding the stochastic nature of hydrologic variables such as streamflows and rainfall
• Modeling stochastic hydrological procedures • Developing new statistical models • Improving the parameter estimation techniques • Proposing new model evaluation and fitness tests • Quantifying uncertainty and imprecision
Forecasting the future state of the resources is necessary for application of operating policies of water resources systems in real-time decision making. For example, in a reservoir, which supplies water for different purposes, the amount of scheduled releases depends on the probable range of inflow to the reservoir. Due to lack of enough knowledge about physical processes in the hydrologic cycle, the application of statistical models in forecasting and generating synthetic data is highly expanded. Furthermore, generation of synthetic data helps to incorporate the uncertainties and probable extreme events in hydrological analyses.
