Freeze Up Water Temperature Forecast for the Yellow River Using Adaptive-Networks-Based Fuzzy Inference System

In this paper, the adaptive-network-based fuzzy inference system (ANFIS) is adopted to forecast the freeze up water temperature of the Yellow River, which is the foundation for ice condition forecast. With its hybrid learning scheme, ANFIS is convenient for modelling the nonlinear multivariable process, for instance, in the case of modelling the hydrological information variety. Through analysing the limited hydrological field data, which is collected from the hydrometric stations in Inner Mongolia reach of the Yellow River, the forecast factors are determined as the model inputs. The related member functions and the predicted period are set up according to determination coefficient of forecast results. And the forecast results of water temperature are compared with the measured records in Shizuishan Hydrometric Station. Through such comparisons, it is discovered that the water temperature forecast results approximately agree with those of the field records. The ANFIS model is acceptable for simulating the freeze up water temperature forecast process and satisfactory results have been obtained.