It has long been suggested that degraded vegetation cover would result in decreased evapotranspiration, reduced precipitation and eventually further degraded cover, initiating an albedo-precipitation feedback cycle (Charney 1975). The lagged response of vegetation cover and soil moisture, which amplify low-frequency oceanic forcings (Giannini et al. 2003; Ward 1998) and buffer out high-frequency 'noise' appear required to closely simulate rainfall variations (Zeng et al. 1999). This conclusion should be no surprise owing to the unrivalled tropical landmass of northern Africa, but the transition from research on land-atmosphere modeling (Goutorbe et al. 1997; Dolman et al. 1997), causative mechanisms of climate change (Long and Entekhabi 2000; Xue et al. 2004) onto the operational implementation of dynamics land surface schemes in climate models remains incomplete in spite of the rapidly growing array of remote sensing observations.
Was this article helpful?