Performance of chaos theory in meteorological forecasts (Case Study: Tehran-temperate climate)
Performance of chaos theory in meteorological forecasts (Case Study: Tehran-temperate climate)
mitra mesbahzadeh1 sohrab hajjam2 abdollah sedaghat kerdar3
1) Islamic Azad University, Science and Research, Tehran -
2) Islamic Azad University, Science and Research, Tehran -
3) Atmospheric Science & Meteorological Research Center, Tehran-
IOI Code
:
XYCZ-AAHHK
Publication :
International Conference on Engineering and Applied Science(iceasconf.com)
Abstract :
Accurate forecasts, for providing a suitable platform on water resource management and disaster management planning is important in meteorology and requirement of methods with high accuracy and updating of predictive models is a priority. In evaluation of hydrological and climate data, the investigation of precipitation parameter is in non-linear time series method. This research compare the performance of intelligent systems based on nonlinear, chaos theory and neural network in estimating monthly precipitation in temperate climate of Tehran. Results neural network system and method for predicting local model and nearest neighbor showed that methods based on chaos not only are sensitive to changes in the range of data but although length the data and attitude data on review process, depending on the conditions also effect on the results. Evaluation results show that the predictions of chaos are of acceptable accuracy and have very high validity. Chaos theory in comparison with neural networks system has provided the better results in temperate climate. Chaos theory by considering nature of data climate of survey and according to the procedures required in the prediction of meteorological parameters can provide excellent results. Using this theory, due to the sensitivity hydro-meteorological forecasts is very efficient and useful.
Keywords : chaos theory neural networks system anfis nearest neighbors meteorological parameters