Ecosystem models are important tools for exploring the temporal and spatial patterns of ecosystem processes and their responses to climate change. However, the implications of uncertainty in model parameters are often overlooked, especially in regional ecosystem model simulations. Here, we use eddy-covariance observations to estimate parameters in a land surface model, and examine the effect on estimates of evapotranspiration (ET). Using the SWH model, which was developed based on the Shuttleworth-Wallace model (S-W model), we use Markov-Chain Monte Carlo techniques to optimize key model parameters using eddy covariance (EC) data from 163 FLUXNET sites . The SWH model was improved by 1) optimizing key parameters such as canopy conductance and soil resistance and 2) incorporating spatial information of key environmental variables (including meteorological and edaphic variables). The improved SWH model agreed well with the measurements with an increase in the coefficient of determination (R²) to 80% in the eight-day averaged ET estimation and a decrease in the root mean square error (RMSE) from 130.2 to 104.3 mm year^-1 compared with the original SWH model. The results showed that the SWH model performs better in herbaceous ecosystems than in woody ecosystems. The results of this work challenge the long-held definition of fixed parameter values and a universal and efficient approach for reducing modeling errors across a wide range of ecological applications is suggested.