The need of weather forecasting is primary to support activities in various sectors so the effors of development for forecest methods to improve the precision and the accuracy of the weather information are very important. various weather forecasting models by engineering or stochastic model approach have been developed although each method has both weaknesses and strengths the efforts for developing techniques or methods to get the best model haue to be done. What is elaborated in this article represent the result of testing in three statistical methods to obtain the best weather foresesting models.three methods as mentioned before are the Autoregressive integrate moving average (ARIMA) Neural Network (NN) and Adiptif Splines Threshold Autoregression (ASTAR) to forescast the temperatur humudity and daily rainfall.The performance of these three methods s are evaluated by corelation values and root mean Square Errir(RMSE). The good performance characterized by a high corelation between actual and fprecastt valuest and also has a small RMSE. The results of this research indicate that ASTAR method produces better signet by a higher correlation lower RMSE values and the constant forecasting from the first day until the thirtieth. The corelation in ASTAR method for Tmax and RHmin respectivelyare 0.70 and 0 75 for ARIMA method are 0.12 and 0.47 for NN method are 0.02 and -0.06. The three methods have poor performance for Tmin RHmax and RRR.