Understanding Eastern Africa Rainfall Variability and Change (Towards Improved Prediction of Seasonal Precipitation) Brant Liebmann University of Colorado, Boulder, Colorado, USA Chris Funk U.S. Geological Survey, Sioux Falls, South Dakota, USA Martin Hoerling, Randall Dole NOAA, Boulder, Colorado, USA Ileana Bladé University of Barcelona, Barcelona, Spain
Model: ECHAM5 Atmospheric Model (Roeckner et al. 2006) ~ 0.75-degree resolution (T159) 40-member ensemble Run with Specified Sea Surface Temperatures
Period of Study:
“Change” = Trend (mm/yr) * length of record (34 years)
Seasonal Horn Precipitation Change
October - December Trend removed
Sign of Eastern Pacific SSTs correctly predicts Horn precipitation in 76.5% of years Ensemble Average
March - May
Gradient in SST produces Low-level convergence over Indonesia Enhanced convection over Indonesia Upper-level outflow (westward shift of ‘Walker’ circulation) Enhanced subsidence over East Africa
Percent correct prediction of observed Horn precipitation based on sign of model Indonesia precipitation Ensemble Average
40-member ensemble of ECHAM 5 atmospheric Model run with specified SST s The interannual anomaly of October-December Horn precipitation is well-simulated by the model ensemble-average, although knowing SSTs in the east Pacific gives almost as good a result The ensemble-average correctly predicts the sign of precipitation anomaly in March-May in two-thirds of years (mostly from precipitation over Indonesia) Model simulates observed change of Horn precipitation for both March-May (decrease) and October-December (increase)