Predictability of the African Drought. Part III: Evaluation of NCAR/CSM and UK Metoffice/PRECIS models Dr. Andre KAMGA Foamouhoue ICTP/ African Drought Conference- June 2008
Motivation Economic and social development in Africa rely quite significantly on rainfall ( eg. GDP loss in Mozambique due to the severity of tropical cyclones related floods in 2000, 10 Billions CFA aid for 2008 to the Agriculture community following very late start of the 2007 summer monsoon in Senegal) It is therefore crucial for Africa s sustainable development to capitalize on substantial investments made by the international community to monitor and predict weather phenomena and particularly rainfall
Objectives Highlight strengths and weaknesses of climate models over parts of Africa Suggest approach to handle these deficiencies when using models for forecasting and projections Discuss future research agenda for climate model development/interpretation over region.
2006 Some evidences of climate variability Rainfall anomalies for Niger (average of 59 stations between 1961 and 2006) Figure 1 : Anomalies moyennes annuelles de la pluviométrie (en mm) période 1961 2006 200,0 150,0 100,0 50,0 0,0-50,0-100,0-150,0-200,0 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Années Anomalies
Tmax anomalies Figure 2: Anomalies moyennes annuelles de la température maximale ( C) période 19612006 1980 1990 2000 1,50 1,00 0,50 0,00-0,50-1,00-1,50 Années
Tmin Anomalies Figure 3: Anomalies moyennes annuelles de la température minimale( C ) période 19612006 1980 1,50 1,00 0,50 0,00-0,50-1,00-1,50-2,00 Années
MOIS DE AVRIL 34,0 33,5 33,0 32,5 32,0 31,5 31,0 30,5 30,0 1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 ANNEES MOIS DE JUILLET 31,5 31,0 30,5 30,0 29,5 29,0 28,5 28,0 27,5 27,0 26,5 1950 1952 1954 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 Monthly Tmax for April and July for Libreville-Gabon ANNEES
Record Tmin mostly observed during the 80s and 90s Records des temperatures minimales depuis la creation des stations Stations/Localités Tillaberi Niamey Konni Zinder Tahoua Maradi Diffa Mainé Soroa Gouré N Guigmi Magaria Agadez Bilma 10,5 10,4 9,6 7,9 7,8 5 5 4,8 4,5 2,6 2 0,9 2,4- janv-98 janv-82 janv-81 déc-80 déc-80 déc-80 janv-83 déc-80 janv-93 janv-93 janv-82 févr-01 janv-95
Record Tmax mostly during the 90s Records des temperatures maximales depuis la création des stations au niger Zinder Dosso Maradi Tahoua Niamey Aero B. Konni Agade z Mainé Soroa Diffa Tillabery Magaria Bilma N'Guigmi 44,6 44,7 45,2 45,6 45,6 45,6 46 46,4 46,5 46,8 46,8 47,1 47,2 23/04/1991 23/04/1973 19/02/1968 06/04/1998 12/04/1998 16/04/1998 30/05/1998 18/04/1998 07/06/1989 13/04/1998 09/06/1969 24/05/1998 23/05/1998
Number of very hot days/365: More hot days since the 80s. 0,14 0,12 0,1 0,08 0,06 0,04 0,02 0 1950 1952 1960 1971 1978 1981 1984 1987 1989 1991 1993 1995 1998 2001 2003 2005 2007 N'Guigmi Niamey Aéro Tillabéry
6 5 4 3 2 1 0 Number of extreme rain events > 60 mm 1952 1955 1957 1997 1959 1961 1963 1965 1967 1971 1973 1975 1977 1979 1981 1985 1989 1991 1994 2000 2003 2006 Diffa Magaria Mainé Soroa
Changes on discharge annual cycle River Niger in Niamey ( number 3 in Africa) Le fleuve Niger à Niamey - débits journaliers débit (m3/s) 3000 2500 2000 1500 1000 1967/1968 1969/1970 1988/1989 1994/1995 1997/1998 1998/1999 500 0 34090 34120 34150 34180 34210 34240 34270 34300 34330 33995 34025 34055 34085
14000 12000 10000 8000 6000 4000 2000 0 Rainfall, malaria and meningitis in Niger J F_M A_M_J J_A_S O_N_D 350 300 250 200 150 100 50 0 35 30 25 20 15 10 5 0 Taux d'attaque m oyen du Paludism e pour 100000 Hbts Cum ul m oyen trim estriel de pluie (mm) Taux d'attaque m oyen de la M e n in g ite p our 100000 Hbts S1 S5 S9 S13 S17 S21 S25 S29 S33 S37 S41 S45 S49 TA_Trim pluvio_cumul 35 cum ul Pluie hebdom adaire m o y en ( m m ) TA_Hebdo pluvio_cumul 30 25 20 15 10 5 0
Expected global impacts (IPCC reports)
Verification of Climate projections Global Model ( NCAR/CSM) Regional Model ( HADRM3 or PRECIS)
NCAR/CSM v1.3 described by Blackmon et al.(2001) JJA sahelian observed precip anom(left) and CSM simulated anom(rigth). Base period 1951-1980
Same as previously but fortemperature anomalies. Quite frequent warm years since early 80s in the obs and mid 80s in CSM.
JJA observed (above) and simulated (below) precip rates ( mm/day)
Observed minus simulated precip rates. More rainfall in the CSM over the much of Sahel and the Highlands (Guinean highlands, Jos plateau,and Cameroon mountains, Ethiopian Highlands and its extension probably up to the Drakensberg in South Africa).
PRECIS under predicts extreme temperatures.
NCAR/CSM cooling up to 6 K
Comparaison PRECIS NCEP Analyses Zonal wind at 200hPa et 600hPa Simulated AEJ and TEJ stronger than observed
Annual cycle of temperature Observations( blue); simulations (red). Nov-Mars( Model warming); April-Oct ( Model cooling) Tmax in March instead of April in the Obs R=0.58 R=0.68
Annual cycle of Temperature Simulation and observations quite similar R=0.97
Systematic Cool bias in the NCAR/CSM
Precipitation climatology May and June
Same as above but for August and September
Annual cycle of precipitation. Early onset, peak and withdrawal on the moonson precipitation well captured R=0.89 R=0.68
Annual cycle for Coastal cities R=0.97 R=0.77
Less rain in the CSM over Sahel With early onset. Substantial underforecasting over the gulf of Guinea.
Annual cycle on a Hovmöller diagram latitude temps
Moonsson circulation Stronger in the Model 10
Simulated AEJ below Its normal latitude Position. «explain under prediction of rainfall over Sahel?»
Simulated TEJ core less intense
Warming trend. Observations may be even warmer if the cool bias observed in past climate persists. With bias corrections CSM Temp projections over Africa become closer to other IPCC models. Normal-above normal precip trend more likely if the dry biasis corrected
Interannual variability over Sahel R=0.33 2. R=0.76
Gulf of Guinea R=0.77 R=0.87
IPCC Models verification for better estimation of uncertainties in scenarios over Africa It is necessary to maintain (CMIP) and strengthen a comprehensive verification of IPCC models over all Africa. Such an experiment may help to better interpret models outputs and significantly improve estimation and communication of projections and related uncertainties for impacts and adaptation in Africa.
END