Climate Change in the Sahel Michela Biasutti biasuttildeo

Climate Change in the Sahel Michela Biasutti biasutti@ldeo. columbia. edu in collaboration with : Alessandra Giannini, Adam Sobel, Isaac Held

Observed, Annual Mean, Continental Scale Rainfall Variability Also see, e. g. Nicholson 1986 (J. Clim. App. Met. )

The Sahel http: //www. pbs. org/wnet/africa/explore/sahel_overview. html

Sahel variability and relation to SST: OBSERVATIONS EOF 1: Sahel rainfall Giannini et al. , 2003, 2005 Associated SST

Sahel variability and relation to SST: ATMOSPHERIC GCMs Giannini et al.

Current questions and challenges: • Was the Sahel drought (and associated SST variations) of natural or anthropogenic origin? Are the 20 th century Sahel drought and SST anomalies consistent with internal climate variability? Are they reproduced in forced runs? In runs forced with natural forcings? In runs forced by anthropogenic forcings? • How will Sahel rainfall change in the greenhouse future?

IPCC Simulations PI Pre-Industrial Control (PI) NASA/GISS XX 20 th Century Simulation (XX) A 1 B Global Warming Scenario (A 1 B) GCMs

The forced component: Sahel XX-PI Rainfall Change Biasutti and Giannini, 2006

XX-PI Rainfall Change

XX-PI SST Change

60 XX Simulations Importance of Internal Noise 1950 -1985 Trend 1950 -1999 Trend 1930 -1999 Trend something missing?

The role of land surface feedback: OBS QTCM Zeng et al. 1999 atmos+ocean +land +vegetation and dust?

Internal Noise vs. Forced Signal

Effect of GHG 4 x(yrs 50: 70)-PI Mean Rainfall Change Robustness of Rainfall Change 20 Surface Temperature

Effect of Reflective Aerosols SULFATE AEROSOL FORCINGS (1850 -1997) Temp RESPONSE NASA/GISS Precip RESPONSE ROTSTAYN AND LOHMANN ‘ 02

Some Conclusions • 20 th Century drying of the Sahel is reproduced by almost all IPCC AR 4 models it is (partly) externally forced. (But natural, internal variability is substantial. ) • The forcing was anthropogenic, with the most robust(*) signal coming from the sulfate aerosol forcing. • The response to GHG increase alone is inconsistent across models, which implies an uncertain outlook for the Sahel.

IPCC Scenarios for the 21 st Century IPCC 2001

Precipitation Response in the Sahel GFDL

What are the possible causes of discrepancy? Is it SST? : s different SST anomalies? sdifferent sensitivity to same SST anomalies? Is it something else? s direct GHG influence?

Relationship of Sahel rainfall & SST (pre-industrial, not forced) Biasutti et al. , 2007

goodness of model PI (training run) Linear Multi-Regressive Model: from SST (Indo-Pacific & Atlantic Gradient) to Sahel Rainfall XX A 1 B interannual (=detrended)

2000 -2100 Trends Sahel Rainfall Indo-Pacific SST Atlantic SST Gradient

goodness of model PI XX Linear Multi-Regressive Model trained on (detrended) PI: from SST (Indo-Pacific & Atlantic Gradient) to Sahel Rainfall nb: same results if NTA & STA are used � (3 predictors). A 1 B interannual + trend

goodness of model XX Linear Multi-Regressive Model trained on (unfiltered) XX: from SST (Indo-Pacific & Atlantic Gradient) to Sahel Rainfall interannual A 1 B interannual + trend

Conclusions • ~30%(? ) of 20 th Century drying of the Sahel was externally forced. The forcing was anthropogenic, with the most robust signal coming from the sulfate aerosol forcing. • In the 21 st Century, when GHG are the dominant forcing, the Sahel response is inconsistent across models. • Global SST changes can explain the 20 th Century trend, but, in most models, not the 21 st Century one (at least not through the same mechanisms active in the past). • A model’s good representation of the past is no indication of a trustworthy prediction of the future. How can we reduce the uncertainty of our climate outlook?

Current Direction: Explore the role of land/sea contrast • Strengthened paleo monsoons follow orbitally-forced increases in seasonality & land/sea contrast. • GHGs force an enhanced contrast. Should we expect enhanced monsoons?

Model ‘Composite’ on Sahel response to GHG 1%to 4 x. CO 2 and A 1 B simulation ncar wet models miroc dry models gfdl A 1 B-XX A 1 B-PI 4 x(@XX)-PI 4 x(@4 x)-PI

Dry/Wet Composites: (2075 2100) - (1975 2000)

Sahara Low & Sahel Rainfall Sahara Low = = Tropical Z 850 - Sahara Z 850 interannual + trend

What controls the Sahara Low? • Is it a heat low? • Is it dynamically influenced from afar? • Is it an effect, instead of a cause, of enhanced Sahel rain?

extra slides

Heat Low? Baroclinic variability in Sahara Z 850 sfc 850 mb 500 mb 200 mb

Other Influences? Barotropic variability in Sahara Z 850 sfc 500 mb 200 mb

Possible Influences on Sahara Low

Sahel Giannini et al. , 2003

19 Coupled GCM : XX-PI SST Change OBSERVED NASA/GISS analysis of surface temp – linear trend 1950 -2000 Hansen et al. 1999 (J. Geophys. Res. ) XX(1975 -2000)-PI(1880 -1900)

Natural or Anthropogenic? natural & anthropogenic

Natural or Anthropogenic? Observed Tsfc anomalies due to the Pinatubo Eruption Robock, Rev. Geophys. , 2000 Composite of Volcanic Years (1975 -1999 period) in Models with Volcanism

10 -year Running Average of Aug-Oct NH Surface T and MDR SST

Best Fit Linear Combination of Global Warming and Aerosol Forcing (red) versus Tropical Atlantic SST (blue) MDR SST Global mean T + aerosol forcing Mann and Emanuel (2006)

Role of Indian and Pacific Lu and Delworth, 2005 back

Changes in Sahel/SST Regression Coefficients Interannual Only PI Interannual & Trend XX A 1 B

Dry/Wet Composites: (2075 2100) - (1975 2000)

CGCM experiments for IPCC AR 4 1% /yr to 4 times pre-industrial CO 2 (4 x) stabilization @ 720 ppm (A 1 B) 20 th Century (XX) Pre-Industrial (PI)

20 th Century Forcings in XX runs ukmo hadgem 1 ANTHRO INDIRECT ORGANIC C BLACK C ukmo hadcm 3 ANTHRO INDIRECT ncar pcm 1 NAT+ANTHRO ncar ccsm 3 NAT+ANTHRO mri cgcm 2 NAT+ANTHRO mpi echam 5 ANTHRO miroc 3. 2 medres NAT+ANTHRO LAND USE INDIRECT ORGANIC C BLACK C miroc 3. 2 hires NAT+ANTHRO LAND USE INDIRECT ORGANIC C BLACK C ipsl cm 4 ANTHRO inm cm 3 NAT+ANTHRO iap fgoals ANTHRO giss e r NAT+ANTHRO LAND USE INDIRECT ORGANIC C BLACK C giss e h NAT+ANTHRO LAND USE INDIRECT ORGANIC C BLACK C giss aom ANTHRO gfdl cm 2. 1 NAT+ANTHRO LAND USE ORGANIC C BLACK C gfdl cm 2. 0 NAT+ANTHRO LAND USE ORGANIC C BLAC C csiro mk 3 ANTHRO cnrm cm 3 ANTHRO cccma cgcm 3 ANTHRO BLACK C INDIRECT back