Font Size: a A A

Modeling patterns of interannual variability in sub-Saharan African rainfall to support food security

Posted on:2003-07-24Degree:Ph.DType:Thesis
University:University of California, Santa BarbaraCandidate:Funk, Christopher CharlesFull Text:PDF
GTID:2460390011981215Subject:Physical geography
Abstract/Summary:
This thesis contains three contributions to the study of rainfall in sub-Saharan Africa: (i) a diagnostic divergence equation that links the upper atmospheric Tropical Easterly Jet to changes in precipitation, (ii) a statistical analysis of the relative influence of large scale, regional and internal forcing factors and (iii) the creation of a 36 year daily/0.1° gridded rainfall dataset. The Tropical Easterly Jet is shown to create upper level divergence over the sub-tropics and convergence over the equator. Increases in the speed of the jet intensify these effects, leading to more rainfall in sub-tropics and less over the equator. The TEJ acts as an atmospheric bridge, connecting subequatorial Africa to changes in the eastern equatorial Pacific. Empirical analyses show that increases in TEJ jet speeds are strongly (R ∼ 0.7) linked to sub-equatorial rainfall in western, southern and Eastern Africa. Jet speeds and rainfall are also controlled by a tropical temperature signal that has an effect similar to ENSO and a strong decadal component. The TEJ/tropical temperature relationship is found to be strongest (R ∼ 0.9) over Western Africa, strongly suggesting that climate change in the Sahel is linked to global warming. In sub-equatorial regions the Indian ocean plays an important role, while the Atlantic ocean is more important for tropical Africa. The Zaire basin is shown to contain a strong auto-regressive component, with strong correlations to previous years (R ∼ 0.6), which suggests that vegetation feedbacks play a dominant role in this heavily vegetated region. The tendency of sub-tropical Africa to be dominated by upper-level winds, and the tendency of tropical Africa to co-vary with its own past should help improve our forecast models by adding new sources of information.
Keywords/Search Tags:Africa, Rainfall, Tropical
Related items