92nd American Meteorological Society Annual Meeting (January 22-26, 2012)

Thursday, 26 January 2012
Vegetation Dynamics and Climate Variability in West Africa
Hall E (New Orleans Convention Center )
Guoqiong Song, University of California, Los Angeles, CA; and Y. K. Xue

West Africa ecosystem region suffered from the most severe and longest drought in the world during the Twentieth Century since the later 1960s. Although numerous modeling studies have established the importance of land/atmosphere interaction in West African climate system, few studies systematically describe the response of vegetation dynamics to climate variability in the Sahelian region.

This study is to systematically understand dynamic vegetation processes in West Africa. We use the offline Simplified Simple Biosphere Version 4/ Top-down Representation of Interactive Foliage and Flora Including Dynamics Model (SSiB4/TRIFFID), which is a fully coupled biophysical-dynamic vegetation (DVM) model. A 100-year equilibrium run was conducted to generate the vegetation initial condition for TRIFFID first. A 59-year simulation from 1948 was then conducted using the reanalysis as meteorological forcing.

Compared with satellite data, the simulated vegetation conditions over Sahel region exhibits seasonal, inter-annual, consistent with West Africa monsoon variability; the simulated inter-decadal variability in vegetation conditions are consistent with the Sahel drought in the 1970s and the 1980s and partial recovery in the 1990s and the 2000s. The spatial distribution of simulated LAIs are well simulated compared with satellite products, but simulated LAIs in tropical region are larger than satellite LAIs, especially in winter, probably due to the residual spin up processes. In tropical rainforest area simulated LAI is larger than satellite products, and has less seasonality.

To investigate the mechanism of dynamic vegetation, water, carbon, and radiation interactions, further analyses are conducted to find relationships between simulated vegetation variability and environmental conditions. It is found that the vegetation characteristics simulated by SSiB4/TRIFFID responds primarily to five factors: air temperature, atmospheric carbon concentration, soil moisture, carbon assimilation rate, and absorbed photosynthetically active radiation. For instance, in temporal analysis, broadleaf trees and C4 plants generally negatively correlates with canopy temperature, and positively correlates with soil moisture. In spatial analysis, vegetation positively correlates with soil moisture but negatively correlates with short wave down; meanwhile, broadleaf tress/C4-plants have positive correlation with long wave down and positive/negative correlation with canopy temperature.

Supplementary URL: