The effect of mid-latitude SST on East African summer rains

Gulilat Tefera Diro, David Grimes and Emily Black

Africa group meeting- February 25, 2008

Outline

• Observational approach

– Relation between rainfall and global SST• Correlation and Composite analysis

– Possible mechanisms for the observed teleconnections between mid-latitude Pacific SST and rainfall

• Modelling approach

– Observed SST forced AGCM (to assess the model capabilities)• Composites of SST based on model rainfall anomaly

– Idealised SST experiment• Does the mid-latitude SST directly affect East African summer rain?

Observation: Relation between SST and rainfallCorrelation

Composites

The link between SST and rainfall via large scale features

Both excess (deficit) rains and warm (cold) SST are associated with a northward(southward) shift in the ITCZ.

The mid latitude SSTA during September is linked to rainfall over Ethiopia via the north/south displacement of ITCZ.

Low level Link

• Both excess/deficit rains and warm/cold SSTs are associated with

– Strong EALLJ

– Strong monsoon trough

– Westerly anomaly over Northeast Africa

Warm/cold Mid-latitude northwest Pacific SST

- north/south shift in ITCZ- Strong/weak EALLJ- Strong/weak westerly over

northeast Africa - Strong/weak Azores high

(NAO)

Excess/deficit rainfall

Modelling approach

Observed SST forced HadAM3

• HadAM3 (3.752.5, 19L) was used to investigate the observed link between SST and rainfall

Relation between model rainfall and observed SST

Question: does the model (HadISST forced HadAM3) rainfall anomaly also linked to the mid-latitude Pacific SST anomaly?

Does the mid latitude SST directly affect rainfall?

• Idealised SST experiments:

– The control run: the AGCM is forced with climatological SST

– The warm SSTA run: the model is forced with an idealised warm SSTA over the mid-latitude Pacific and climatological SST elsewhere

– The cold SSTA run: the model is forced with an idealised cold SSTA pattern over the northwest mid-latitude Pacific and climatological SST elsewhere

The SSTA imposed in June

Result from the idealised Experiment

• Impact on the total Kiremt (JJAS) rainfall

• The mid-latitude northwest Pacific has an impact on the total JJAS rainfall over NE Africa although it is a bit far north.

Result from the idealised Experiment

Cold SSTA over mid-latitude northwest Pacific causes a southward shift in the ITCZ over Africa

The impact of mid-latitude SSTA on the total rainfall

The impact on the sub-seasonal rainfall

• The mid-latitude northwest Pacific SST directly affect September rainfall

• This implies it affects the total rainfall by affecting early cessation of the rainy season

Summary• The mid-latitude northwest Pacific in September correlates

significantly with the eastern Ethiopian summer rainfall

• This has an implication that the mid-latitude SST can be used as forecasting the cessation of Kiremt rainfall.

• Observational study suggests that the SST and rainfall are linked by among other things via the ITCZ

• The modelling study suggests that the mid-latitude northwest Pacific affects the rainfall over Africa via the north-south shift in the ITCZ.

• The modelling study also suggest that mid-latitude northwest Pacific SST directly affect the eastern Ethiopian rainfall during September.

• This has an important implication that the mid-latitude SST can be used as a predictor to forecast the early cessation of the rainy season