Land surface processes and Sahel climate

This paper examines the question of land surface-atmosphere interactions in the West African Sahel and their role in the interannual variability of ra infall. In the Sahel, mean rainfall decreased by 25-40% between 1931-1960 a nd 1968-1997; every year in the 1950s was wet, and nearly every year since 1970 has been anomalously dry. Thus the intensity and multiyear persistence of drought conditions are unusual and perhaps unique features of Sahel cli mate. This article presents arguments for the role of land surface feedback in producing these features and reviews research relevant to land surface processes in the region, such as results from the 1992 Hydrologic Atmospher ic Pilot Experiment (HAPEX)-Sahel experiment and recent studies on aerosols and on the issue of desertification in the region, a factor implicated by some as a cause of the changes in rainfall. Included also is a summary of e vidence of feedback on meteorological processes, presented from both model results and observations. The reviewed studies demonstrate numerous ways in which the state of the land surface can influence interactions with the at mosphere. Surface hydrology essentially acts to delay and prolong the effec ts of meteorological drought. Each evaporative component of the surface wat er balance has its own timescale, with the presence of vegetation affecting the process both by delaying and prolonging the return of soil moisture to the atmosphere but at the same time accelerating the process through the e vaporation of canopy-intercepted water. Hence the vegetation structure, inc luding rooting depth, can modulate the land-atmosphere interaction. Such pr ocesses take on particular significance in the Sahel, where there is a high degree of recycling of atmospheric moisture and where the meteorological p rocesses from the scale of boundary layer development to mesoscale disturba nce generation are strongly influenced by moisture. Simple models of these feedback processes and their various timescales have demonstrated that the net feedback to the atmosphere is positive for both wet and dry surface ano malies. Hence the role of the surface is to reinforce meteorologically indu ced changes. Recovery from the dry state is slower than from the wet state, suggesting that dry conditions would tend to persist longer, as is actuall y observed in the Sahel. These simple models suggest that the surface hydro logy locks the system into a drought mode that persists for several years, until the system randomly slips into a persistent wet mode. The hypothesis that desertification in the Sahel might likewise be responsible for the per sistent drought is found to be untenable. Rather than a progressive encroac hment of the desert onto the savanna, the vegetation cover responds dramati cally to interannual fluctuations in rainfall. There is little evidence of large-scale denudation of soils, increase in surface albedo, or reduction o f the productivity of the land, although degradation has probably occurred in some areas. There has, however, been a steady buildup of dust in the reg ion over the last half a century. Significant radiative effects of the dust have been demonstrated; therefore the dust has probably influenced large-s cale climate. The buildup is probably mainly a result of changes in the lan d surface that accompanied the shift to drier conditions, but it may have b een exacerbated by anthropogenic factors. Complex general circulation model s nearly universally underscore the importance of feedback processes in the region. Although it has not been unequivocally demonstrated that the rainf all regime of the Sahel is modulated by surface processes, there is recent observational evidence that this is case.