The performance of a coupled SVAT and DGVM model over six decades in West Africa

This study investigated how a combined soil-vegetation-atmosphere transfer (SVAT) model, the Simplified Simple Biosphere Version 4, and a digital global vegetation model (DGVM) model, the Top-down Representation of Interactive Foliage and Flora Including Dynamics Model (TRIFFID) were, when coupled, able to simulate vegetation and soil processes forced by climate variables alone. The study was undertaken in West Africa from 1988 to 2008. During the modeling period the region experienced a severe, multiyear drought, preceded by much wetter conditions and followed by a gradual increase in annual rainfall. The ability of the model to respond to these fluctuations was explored by comparison with; the cover of six different plant functional types (PFTs); the leaf area index (LAI); primary production; and aspects of the surface and soil water balance. The results showed that the simulated PFT's spatial distribution of total leaf area index (LAI) corresponds well to climate variability and were consistent with satellite derived vegetation conditions. The vegetation characteristics simulated by SSiB4/TRIFFID in this study responded primarily to air temperature and soil moisture. Correlation analysis between the responses of vegetation and soil processes to precipitation indicated a lag of 2-3 months. The simulated interannual and inter-decadal variability in vegetation conditions was consistent with the West Africa drought in the 1970s and 1980s and partial recovery in the 1990s and 2000s.