In order to improve land surface modeling predictions, the land surface models are generally calibrated against measurements. The study addressed the parameter sensitivity analysis, model calibration, the realistic quantification of parameter uncertainty and its effect on the results of Noah land surface model. The LH-OAT method was applied in the sensitivity analysis for the Noah LSM model parameters. Based on the eight important parameters effect on the land surface upward longwave radiation, the shuffled complex evolution metropolis (SCEM-UA) global optimization algorithms is used to automatically infer the posterior distribution of the model parameters. To overcome the computational burden, the optimization has been implemented using parallel computing. The Noah model prediction using the optimal parameters shows that the simulated upward longwave radiation matched measurements fairly well with an R2 value of 0.9842 and Root Mean Squared Error (RMSE) of 5.42W/m2. Results demonstrate that the SCEM-UA algorithm can efficiently evolve the posterior distribution of the parameters for the complex land surface model.