Climate Lab Logo

Publications Portal

START flash tag cloud requires Flash Player 9 or better.

Tag Cloud created by Roy Tanck


Empirical Analysis of Intra-seasonal Climate Variability over the Greater Horn of Africa

Jared Bowden & Fredrick Semazzi
Journal of Climate,

This study examines the intra-seasonal climate variability over the Greater Horn of Africa (GHA) during the rainy season of October, November, December (OND). The investigation is primarily based on empirical orthogonal function (EOF) analysis of the pentad Climate Prediction Center Merged Analysis of Precipitation (CMAP) data for the period, 1979-2001. The EOF analysis reveals two dominant modes of intra-seasonal variability for the OND season: mixed El Nino-Southern Oscillation / Indian Ocean Dipole (ENSO-IOD) and a decadal mode.

The leading mode is associated with ENSO-IOD co-variability. Case study of several intra-seasonal ENSO-IOD events within the recent decades indicate that during the warm (positive) events pentad rainfall is consistently above normal during the entire season despite fluctuations between pentads. However, case study analyses of negative ENSO-IOD events show that the negative cases are not mirror images of the warm events. The negative events exhibit pronounced wet and dry spells superimposed on the consistent dry anomaly background conditions. There is no large signal regarding the onset for either case, but the withdrawal of the positive (negative) events is anomalously wet (dry).

The second mode of variability is associated with decadal shift in the rainfall with northern (southern) GHA becoming wetter (drier) in the recent decade. The decadal changes in individual pentads can be quite different across the season and has a tendency to manifest itself through extreme events. The analysis underscores the need to exercise caution when applying seasonal average based statistics to infer the long term behavior on sub-seasonal time scales. Additional analyses further confirm the decadal rainfall shift using four different rainfall datasets. Averaging the datasets to help aid in removing bias of individual datasets shows that on average northern (southern) portions of GHA are 29% (19%) wetter (drier) in the recent decade.

(Download Publication)