ROLE OF THE INDIAN AND ATLANTIC OCEANS ON THE CLIMATE VARIABILITY OF EASTERN AFRICA Charles C. Mutai Kenya Meteorological Department P. O. Box 30259,

ROLE OF THE INDIAN AND ATLANTIC OCEANS ON THE CLIMATE VARIABILITY OF

EASTERN AFRICA

Charles C. Mutai

Kenya Meteorological Department

P. O. Box 30259, Nairobi Kenya

CURRENT UNDERSTANDING OF EAST AFRICAN CLIMATE

• Rainfall is the climatic factor of maximum significance• Seasonal rainfall patterns related to migration of ITCZ:

long rains (MAM), short rains (S/OND)• Interannual variability of seasonal rains results from

complex interactions of forced and free atmospheric variations (OND variability stronger than MAM):SST forcing in Pacific, Atlantic and Indian (Ogallo et al. 1988; Nyenzi 1992; Ininda 1994; Mutai et al. 1988; among others)

OND rainfall positively related to ENSO Large/synoptic/meso - scale circulations; Anticyclones cyclones, easterly/westerly perturbations (Nakamura 1969; Anyamba 1984; Hastenrath et al. 1993; Okoola 1999; Mutai and Ward 2000; etc).

CURRENT UNDERSTANDING OF EAST AFRICAN CLIMATE (2)

• Predictive schemes based solely on ENSO showed skill (Farmer 1988; Hutchinson 1988; KMD/DMCN; etc.)

• Anomalous SST in the Atlantic and Indian Oceans - associated with rainfall fluctuations (Reverdin et al. 1986; Flohn 1987; Nicholson and Entekhabi 1987; Webster et al. 1999; Saji et al. 1999; Black et al. 2002; among others)

• Two SSTs (of the three) predictors used by Mutai et al. (1998) largely independent ENSO ~ SST variability in the northwestern tropical Pacific - Indian Ocean and in the equatorial - tropical South Atlantic.

• GCM simulations (Goddard and Graham 1999; Latif et al. 1999) indicated strong role for the Indian Ocean SSTs in East African rainfall variability

• SST gradients in the Atlantic and Indian Ocean are currently used in monthly / seasonal climate monitoring and forecasting (KMD and DMCN)

MEAN MONTHLY RAINFALL (mm) OVER EAST AFRICA, 1931-1985

“Long Rains”

“Short Rains”

0-25 25-50 50-100 100-200 > 200 mm

Nyenzi (1992)

REGIONALIZATION OF E.A. RAINFALL

Indeje et al. (2000)Indeje et al. (2000)

OND SST / OLR TELECONNECTIONS

X -1

MAM SST/OLR TELECONNECTIONS

1951-96

1974-96

OND RAINFALL-WIND TELECONNECTIONS

1974-96

OND RAINFALL-WIND TELECONNECTIONS

ENSO - variance - removed

CONTRIBUTION OF ATLANTIC OCEAN IN CLIMATE VARIABILITY OF EAST AFRICA

Mutai et al. (1998) developed an empirical prediction for East Africa OND rainfall based on three precursor global SST REOFs:

• SST REOF4 closely reflect the ENSO signal

• SST REOF5 - SST variance in northwestern tropical Pacific - Indian Ocean

• SST REOF2 - variability in the equatorial and tropical South Atlantic

ROTATED SST PREDICTORS (Mutai et al. 1998)

SST REOF5

SST REOF4

SST REOF2

OND TELECONNECTIONS WITH SEPTEMBER SST REOF2, 1974-

96

PENTAD 850-hPa WIND TELECONNECTIONS

Oct-Nov 1979-96

PENTAD 200-hPa WIND TELECONNECTIONS

Oct-Nov 1979-96

CONTRIBUTION OF INDIAN OCEAN IN CLIMATE VARIABILITY

OF EAST AFRICAEquatorial Indian Ocean variability can be simply be computed as a multivariate measure of key variables over the west (W), East (E) and central (C) Indian Ocean basin:

IOD = (TW - TE) - (PW - PE) - UC

where T, P and U represents the SST, sea-level pressure (SLP) and near-surface zonal wind index, respectively.

PENTAD COMPOSITE 850-hPa WINDS

Warm Phase ~

W. Indian Ocean

Cold Phase ~

W. Indian Ocean

SURFACE WIND FIELDS - Reverdin et al. 1986

NOV 1961

Climatology

Cf. 1997

• Goddard and Graham (1999) used observed SSTs 1970-92

• Latif et al. (1999) considered case of heavy rains of Dec-Jan 1997/98

• Isolated the primary control of SST anomalies in the Indian Ocean only, then in the Pacific and then in both the Indian and Pacific Oceans

• Role of Indian Ocean in climate variability over East Africa

GCM CLIMATE SIMULATIONS

GCM CLIMATE SIMULATIONS - Goddard and Graham 1999

1970-92

• Understanding mechanisms for Indian / Atlantic Ocean dynamics is important for East African climate

• Observational network in Indian Ocean?• Utility of satellite-derived and reanalysis

data

SUMMARY