مقايسه و پيش بيني تغييرات دماي سالانه ي تبريز با ناهنجاري هاي دمايي كره ي زمين با استفاده از روش هاي رگرسيون خطي و شبكه ي عصبي مصنوعي

Comparison and Prediction of the Temperature Variations of Tabriz with the Global Temperature Anomalies Through the Use of Regression and Artificial Neural Network Methods

In this survey the data relating to global temperature anomalies and annual mean temperature of Tabriz station have been used for 1951-2003 period. The main methodologies are being used in this research are the Pearson correlation coefficient method, analysis of trend component of time series, simple linear and polynomial regression as a semi linear model and Artificial Neural Networks methods. The results of applying Pearson analysis indicate significant positive and direct correlation between global temperature anomalies and annual temperature in Tabriz station. The analysis of long term trend components of time series shows that the annual mean temperature of Tabriz and annual global temperature anomalies has a increasing trend towards the length of the period. Also have simulated the relationships between annual temperature in Tabriz station and global temperature anomalies using artificial neural networks. The applying different methods demonstrated that artificial neural network is recognized as a better and more accurate simulation model compared to the other models applied in this research, i.e., simple regression model, and semi-linear polynomial regression with the power of 6 models.