In order to effectively integrate the prediction results of various forecasting methods, the monthly mean temperatures of the three climatic zones in Shaanxi from July 1999 to June 2010 are employed in the experiments. Based on the historical precision derived from products of the National Climate Centre and three kinds of prediction methods, the two schemes involving six objectiveintegrated dynamic forecasting models are established. The results show that: (1) The yearly average score is higher than each member in the range of the whole ensemble forecasting models. Average score of six ensemble forecasting models is 4.6 higher than other four members, and is 5.1 higher than operational publishing products provided by Shaanxi Climate Center. The yearly anomaly sign consistency rate of each ensemble forecasting model is higher than every value resulting from at least three prediction methods among all. The average value of six ensemble forecasting models with the anomaly sign consistency rate is 4.7% higher than other four members, and is 5.1% higher than operational publishing products predicted by Shaanxi Climate Center. (2) Yulin received the best forecasting result among three climatic zones. Average score of six ensemble forecasting models is 5.7 higher than other four members, and is 5.7 higher than operational publishing products provided by Shaanxi Climate Center. The average value of ensemble forecasting models with the anomaly sign consistency rate is 6.9% higher than four members, and is 7.3% higher than operational publishing products predicted by Shaanxi Climate Center. (3) The second scheme of ensemble forecasting models is better than the first one. Especially, the average score of Z23 is 5.1 higher than other four members, and is 5.7 higher than operational publishing products. The average value of ensemble forecasting models with the anomaly sign consistency rate is 6.0% higher than four members, and is 7.6% higher than operational publishing products score. Thus, it is recommended to be used in climate prediction.