Seasonal-to-Interannual Variability of Ethiopia/Horn of Africa Monsoon. Part I: Associations of Wavelet-Filtered Large-Scale Atmospheric Circulation and Global Sea Surface Temperature

Abstract

Horn of Africa rainfall varies on multiple time scales, but the underlying climate system controls on this variability have not been examined comprehensively. This study therefore investigates the linkages between June–September Horn of Africa (especially Ethiopian) rainfall and regional atmospheric circulation and global sea surface temperature (SST) variations on several key time scales. Wavelet analysis of 5-day average or monthly total rainfall for 1970–99 identifies the dominant coherent modes of rainfall variability. Several regional atmospheric variables and global SST are then identically wavelet filtered, based on the rainfall frequency bands. Regression, correlation, and composite analyses are subsequently used to identify the most important rainfall–climate system time-scale relationships. The results show that Ethiopian monsoon rainfall variation is largely linked with annual time-scale atmospheric circulation patterns involving variability in the major components of the monsoon system. Although variability on the seasonal (75–210 days), quasi-biennial (QB; 1.42–3.04 yr), and El Niño–Southern Oscillation (ENSO; 3.04–4.60 yr) time scales accounts for much less variance than the annual mode (210 days–1.42 yr), they significantly affect Ethiopian rainfall by preferentially modulating the major regional monsoon components and remote teleconnection linkages.The seasonal time scale largely acts in phase with the annual mode, by enhancing or reducing the lower-tropospheric southwesterlies from the equatorial Atlantic during wet or dry periods. The wet QB phase strengthens the Azores and Saharan high and the tropical easterly jet (TEJ) over the Arabian Sea, while the wet ENSO phase enhances the Mascarene high, the TEJ, and the monsoon trough more locally. The effects of tropical SST on Ethiopian rainfall also are prominent on the QB and ENSO time scales. While rainfall–SST correlations for both the QB and ENSO modes are strongly positive (negative) over the equatorial western (eastern) Pacific, only ENSO exhibits widespread strong negative correlations over the Indian Ocean. Opposite QB and ENSO associations tend to characterize dry Ethiopian conditions. The relationships identified on individual time scales now are being used to develop and validate statistical prediction models for Ethiopia.