Incorporation of TOPMODEL into land surface model SSiB and numerically testing the effects of the corporation at basin scale

In order to examine and analyze the effects of integration of land surface models with TOPMODEL and different approaches for the integration on the model simulation of water and energy balances, the coupled models have been developed, which in- corporate TOPMODEL into the Simplified Biosphere Model （SSiB） with different approaches（one divides a basin into a num- ber of zones according to the distribution of topographic index, and the other only divides the basin into saturated and unsatu- rated zones）. The coupled models are able to （but SSiB is not able to） take into account the impacts of topography variation and vertical heterogeneity of soil saturated hydraulic conductivity on horizontal distribution of soil moisture and in turn on water and energy balances within the basin（or a grid cell）. By using the coupled models and SSiB model itself, the daily hy- drological components such as runoffs are simulated and final results are analyzed carefully. Simulated daily results of hydro- logical components produced by both SSiB and coupled models show that （i） There is significant difference between results of soil wetness, its vertical distribution and seasonal variation, water and energy balance, and daily runoff in the basin predicted by SSiB and by the coupled models. The land surface model currently used such as SSiB is likely to misrepresent real feature of water and energy balances in the basin. （ii） Compared with the results for the basin predicted by SSiB, the coupled models predict more strong vertical and seasonal changes in soil wetness, higher evaporation and lower runoff, and improve the base flow simulation obviously. （iii） Comparing the results for the basin predicted by two coupled models with different integration approach and SSiB one by one, the results of daily runoffs and soil wetness predicted by the two coupled models are quite sim- ilar. It means, for the coupled models, the approach by dividing a region being considered into more subzones may have lim- ited effects on improving runoff simulation results. The scheme only to divide the region into saturated and unsaturated zones may be a convenient and effective scheme. But then, if the results from the two coupled models for the basin are carefully compared, the simulated results by the coupled model with dividing the basin into more subzones will show higher evaporation and surface runoff but lower subsurface flow, lower total runoff, and lower ground water level averaged for five yea