ON THE CHOICE OF SOIL HYDRAULIC MODELS IN LAND-SURFACE SCHEMES YAPING SHAO and PARVIZ IRANNEJAD Centre for Advanced Numerical Computation in Engineering and Science, UNSW, Sydney, Australia (Received in final form 11 August 1998) Abstract. The uncertainties in soil hydraulic functions and soil hydraulic parameters affect the per- formance of land-surface schemes used in climate and weather prediction models. The soil hydraulic model of Clapp and Hornberger is most widely used in land-surface modelling, while other models favoured by soil physicists are hardly used for the purpose. In this study, we give a summary of four soil hydraulic models and examine the impact of these models on the performance of a land-surface scheme. It is found that inconsistency in soil hydraulic functions and parameters leads to different outcomes in land-surface modelling. We introduce a technique to match the soil hydraulic parameters for different models, so that the disagreement in the description of soil hydraulic properties among different models is reduced, while intrinsic differences in the soil hydraulic functions remain. The numerical tests also show that the land-surface model has a degree of tolerance to the uncertainties in soil hydraulic models, at least in the case of off-line simulations. The van Genuchten model performs well, but is numerically expensive. The Brooks–Corey and Clapp–Hornberger models are sufficiently accurate with numerical efficiency, and are therefore more suitable for land-surface schemes used in atmospheric models. Keywords: Climate modelling, Land-surface modelling, Soil hydraulic models, Soil moisture, Sur- face energy fluxes. 1. Introduction Soil moisture is an important quantity to be determined in land-surface schemes for atmospheric, hydrological and ecological modelling. Since the early studies of Budyko (1956) and Manabe (1969), many land-surface schemes have been developed with increased sophistication in representing the complex physical and biochemical land-surface processes in climate and weather prediction models (e.g., Dickinson et al., 1993). Most of the current land-surface schemes are one- dimensional, but techniques are being developed to allow such schemes to be ap- plicable to extended areas. Recent evaluations of land-surface schemes (e.g., Shao and Henderson-Sellers, 1996; Wood et al., 1998) have revealed many aspects that require improved treatment. As far as surface soil hydrology is concerned, there are four outstanding issues: (1) Spatial variation of soil hydraulic properties – soil hydraulic and other land- surface properties vary strongly in space. For large-scale applications, such as Boundary-Layer Meteorology 90: 83–115, 1999. © 1999 Kluwer Academic Publishers. Printed in the Netherlands.