A GLOBAL VIEW OF LARGE-SCALE PRECIPITATION VARIABILITY

Observational studies and model experiments make abundantly clear the need for a global perspective in order to understand the nature and ca uses of persistent regional precipitation anomalies. Rainfall in the d eep tropics is particularly important as a forcing mechanism for the a tmosphere's large-scale circulation and climate. Analysis of systemati c space-based observations and surface marine data over the past three decades has vastly improved our understanding of tropical convective regimes and their relationship to surface conditions. The characterist ics of the annual cycle of tropical convection and its relationship to sea surface temperature field and the general circulation of the trop ics are reviewed. The hierarchal nature of tropical precipitation vari ability on time/space scales ranging from synoptic cloud clusters thro ugh the intraseasonal Madden-Julian Oscillation to multiyear El Nino-S outhern Oscillation cycle is discussed. Links between tropical convect ion and extratropical precipitation on time scales ranging from synopt ic to multiyear are examined, with emphasis on conditions over the Nor th Pacific-North American sector during winter. Precipitation variabil ity over a number of regions bordering the Atlantic basin are related to Atlantic sector modes of SST and circulation variability. Systemati c modes of Atlantic variability and their relationship to regional pre cipitation variability are described with emphasis on the tropics. Cha nges in landscape characteristics (vegetative cover, soil moisture, su rface roughness), whether natural or human induced, result in changes in the surface radiation balance and the fluxes of heat and moisture. Our current understanding of the role of land surface processes in sus taining or intensifying anomalous precipitation regimes is briefly dis cussed. Identification of an anthropogenic trend in the presence of de cadal-scale natural variations in precipitation is a formidable challe nge. Three examples of large-amplitude secular variations in regional precipitation regimes (Sahel, North American Great Plains, and India) are discussed in terms of possible forcing mechanisms. Continuous glob al monitoring of precipitation is a challenging task. Satellite-based observations, in conjunction with effective use of surface-based ''gro und truth'' data and further development of four-dimensional data assi milation methodology, offer the only realistic prospects for significa nt improvement in the monitoring and quantification of global precipit ation. Current methods for estimating precipitation from space-based o bservations are described and an overview of the upcoming Tropical Rai nfall Measurement Mission is included.