Analysis of parameter sensitivity on surface heat exchange in the Noah land surface model at a temperate desert steppe site in China

The dominant parameters in the Noah land surface model (LSM) are identified, and the effects of parameter optimization on the surface heat exchange are investigated at a temperate desert steppe site during growing season in Inner Mongolia, China. The relative impacts of parameters on surface heat flux are examined by the distributed evaluation of local sensitivity analysis (DELSA), and the Noah LSM is calibrated by the global shuffled complex evolution (SCE) against the corresponding observations during May–September of 2008 and 2009. The differences in flux simulations are assessed between the Noah LSM calibrated by the SCE with 27 parameters and 12 dominant parameters. The systematic error, unsystematic error, root mean squared error, and mean squared error decompositions are used to evaluate the model performance. Compared to the control experiment, parameter optimization by the SCE using net radiation, sensible heat flux, latent heat flux, and ground heat flux as the objective criterion, respectively, can obviously reduce the errors of the Noah LSM. The calibrated Noah LSM is further validated against flux observations of growing season in 2010, and it is found that the calibrated Noah LSM can be applied in the longer term at this site. The Noah LSM with 12 dominant parameters calibrated performs similar to that with 27 parameters calibrated.