6533b833fe1ef96bd129b3f3
RESEARCH PRODUCT
Prévisibilité des précipitations ouest africaines (échelles intrasaisonnières et saisonnières) au travers des simulations type DEMETER et des observations
Lotfi Boualisubject
[SDU.STU.CL] Sciences of the Universe [physics]/Earth Sciences/ClimatologyPrévisibilité saisonnière[SDU.STU.CL]Sciences of the Universe [physics]/Earth Sciences/Climatologymousson africaineClimatologieprécipitations[ SDU.STU.CL ] Sciences of the Universe [physics]/Earth Sciences/Climatologydescription
Following the recommendations proposed by the AMMA community three main objectives were set for this research. The first was to analyze the direct outputs of general circulation models of the atmosphere used in the DEMETER project. The second was to propose a new method for predicting seasonal rainfall in the Sahel using the direct outputs of GCMs the best simulated, that is to say, those related to atmospheric dynamics which explains the penetration of the West African monsoon in the continent, and therefore, the total rainfall during the monsoon season of boreal summer. The third goal finally was to analyze and compare probabilistic forecasts from the forum and PRESAO project ENSEMBLES. To meet these objectives, we relied on 7 databases: 4 of them are used as references (CRU, CMAP, reanalyses American-R1 and R2 and SST), simulations (precipitation and atmospheric dynamics) European projects ENSEMBLES and DEMETER and finally digitized maps from the PRESAOforum .It is demonstrated that :• The statistical-dynamic approach can significantly improve the scores obtained directly from wet raw GCMs DEMETER since reached 40% of the total variance explained (as against 10% for Demeter), performance is best when the dynamics simulated air is used.• The score of the model mean (MMEM) is very close to what may be the best GCM used individually, that whatever the method used (index or field). This shows that a multi-ensemble approach is more efficient than a multi-runs approach on a single model and that the overall approach allows to find in the GCMs teleconnections responsible for a good prediction of rainfall in the Sahel .
year | journal | country | edition | language |
---|---|---|---|---|
2009-05-25 |