Evaluation of Riverflow Forecasting Models

Snowfall has an important role in the amount and distribution of river flows in mountainous basins. Therefore، it is crucial to assess the application of river flow forecasting models that employ an index of snow in their structure. In this paper، a range of forecasting models that have been in use since decades ago are presented and compared with the more recently developed ones. These include some conventional models based on simple statistical methods and models that use more sophisticated methods such as Artificial Neural Networks (ANN) and Geographic Information Systems (GIS) with remotely sensed snow covered area (SCA) data obtained through NOAA satellites. Moreover، it is shown that the use of test periods is important in reaching realistic conclusions about the models. In addition، a testing method is presented that through an iterative process، evaluates the whole period of available data. Data from the Dez Basin in Iran is used to carry out the calculations. Furthermore، the role of spring rainfalls and their simultaneous implication in the forecasting models is investigated.