Wikindx Resources

Journal Article: BibTeX citation key:  Diattaa
Diatta, S. & Fink, A. H. Statistical relationship between remote climate indices and West African monsoon variability. IN International Journal of Climatology, .
Added by: Andreas Fink 2014-02-04 15:07:00    Last Edited by: Andreas Fink 2014-02-04 15:16:39
 B  
Categories: Monsoon system and its variability
Keywords: Climate, Interannual variability, Precipitation, Sahel, Soudanian zone, Western Africa
Creators: Diatta, Fink
Collection: International Journal of Climatology

Peer reviewed
Number of views:  835
Popularity index:  44.18%

 
Abstract
This paper revisits the strength and stationarity of linear statistical relationships between indices of monsoon
rainfall derived from rain gauge data for three regions in tropical West Africa [West Sahel (WS), Central Sahel (CS), and
Guinea Coast (GC)] and various indices describing remote variations in the ocean–atmosphere climate systems from 1921
to 2009. The results reveal that both the Atlantic Multi-decadal Oscillation (AMO) and the Atlantic Meridional Mode
(AMM) show a positive linear correlation to Sahel rainfall. The percent variance explained (PVE) ranges from 10 to 25%.
The correlation stems from periods longer than 8 years for AMO, but AMM does show a correlation on interannual time
scales for WS that was absent in the 1970s and 1980s. The PVE by El Ni˜no Southern Oscillation (ENSO) indices is,
though statistically significant, on the order of 10% and found mainly on the interannual time scale. A strong and stable
correlation with PVEs larger than 50% is found between the sea surface temperatures (SSTs) in the Atlantic 3 region
(ATL3, 0◦–20◦W, 3◦S–3◦N) and the GC rainfall. On the other hand, the correlation between ATL3 and WS rainfall
changed from significantly negative to significantly positive after the 1970s. The Eastern Mediterranean SSTs are found to
be significantly related to CS rainfall, especially in recent years with PVEs between 36 and 47%. Multi-linear regression
analyses reveal that the relative importance of the Indian Ocean is 42% in the optimal regression model. For the CS, this
value is 37% for the Eastern Mediterranean SSTs and 70% in case of the GC using the ATL3 index. However, except for
the GC, non-stationarities in the correlation between the climate state indexes and West African monsoon (WAM) rainfall
suggest the need of the application of different regression models depending on the active ‘teleconnection regime’.
Added by: Andreas Fink    Last Edited by: Andreas Fink