The use of a reduced form model to assess the sensitivity of a land surface model to biotic surface parameters

Land surface models (LSM) are designed to provide turbulent and radiative fluxes from the surface to the atmosphere that are in turn important in driving atmospheric models. The fluxes in land surface models are controlled by the surface properties and hence the correct parameterization of these properties and the processes that define them is vital in obtaining realistic fluxes. We investigate the sensitivity of turbulent fluxes predicted by the NCAR LSM to biotic surface parameters (roughness length, displacement height, leaf area index, rooting fraction with depth, albedo and minimum stomatal resistance). This is achieved using a multivariate reduced-form model that expresses the results of multiple realizations of the physical model as integrative response metrics such as summer ground, sensible and latent heat fluxes; average soil summer water content; and average soil temperature of the upper layer in each season. The sensitivity analysis shows that most response metrics were most sensitive to roughness length and displacement height. In summertime, leaf area index was important in determining summer ground heat flux, ground temperature and the timing of snow-free ground. This timing was also sensitive to albedo. Rooting fraction with depth was only important in determining summer soil water content. In general the NCAR LSM was more sensitive to a range of climate driven perturbations examined in a companion paper than to the range of biotic surface parameters chosen here.