The paper proposes a statistical method for post-processing ensemble forecasts using Bayesian Model Averaging (BMA) to calibrate and sharpen the predictive distributions. BMA combines the individual forecasts from different models, weighted by their posterior probabilities, to produce a more accurate and calibrated forecast. The method is applied to 48-hour forecasts of surface temperature in the Pacific Northwest using the University of Washington's MM5 ensemble. The results show that BMA significantly improves the calibration of the forecasts, making the prediction intervals much narrower and more accurate compared to raw ensemble forecasts and sample climatology. The BMA deterministic forecast also outperforms the ensemble mean and individual forecasts in terms of root-mean-square errors. The method is further validated through simulations, demonstrating its effectiveness even when the ensemble is well-calibrated or overdispersed. The paper concludes by discussing the practical implications and potential applications of BMA in weather forecasting.The paper proposes a statistical method for post-processing ensemble forecasts using Bayesian Model Averaging (BMA) to calibrate and sharpen the predictive distributions. BMA combines the individual forecasts from different models, weighted by their posterior probabilities, to produce a more accurate and calibrated forecast. The method is applied to 48-hour forecasts of surface temperature in the Pacific Northwest using the University of Washington's MM5 ensemble. The results show that BMA significantly improves the calibration of the forecasts, making the prediction intervals much narrower and more accurate compared to raw ensemble forecasts and sample climatology. The BMA deterministic forecast also outperforms the ensemble mean and individual forecasts in terms of root-mean-square errors. The method is further validated through simulations, demonstrating its effectiveness even when the ensemble is well-calibrated or overdispersed. The paper concludes by discussing the practical implications and potential applications of BMA in weather forecasting.