Detection and Attribution of African Rainfall Change Martin
Detection and Attribution of African Rainfall Change Martin Hoerling, NOAA 11 th Annual CCSM Workshop 21 June 2006 in collaboration with Jim Hurrell, Jon Eischeid, Adam Phillips
Undernourished Population (2000 - 2002) according to FAO. Courtesy of FAO 2005 .
70% of Africans Live by Farming 40% of Exports Are from Agricultural Products
1993 2000 The Warming of Africa
Change in Fresh Water Availability by 2025 Only Due to Population Change
What is the Fate of 21 st Century African Rainfall? GFDL AR 4 coupled models predict an extremely dry Sahel in the future Held et al. 2005
A Physical Explanation for the 20 th Century Drying May Provide Insight on 21 st Century Change
Trend Toward Increased Aridity in the 20 th Century from station data
Trend Toward Increased Aridity in the 20 th Century from station data
Trend Toward Increased Aridity in the 20 th Century from station data
“Detection is the process of demonstrating that an observed change is significantly different (in a statistical sense) than can be explained by natural variability. ” (IPCC, 2001) “Attribution is the process of establishing cause and effect with some defined level of confidence, including assessment of competing hypotheses. “ (IPCC, 2001)
Detection &Attribution Tools Ø Observed, Monthly 1950 -1999 Gridded pcpn & sfc. T from U. of Delaware, and also East Anglia (CRU). Ø Unforced Coupled Atmosphere-Ocean Simulations, 3 CGCMs each spanning 240 Years. (Detection) Ø Atmospheric Climate Simulations Forced by Observed, Global SSTs for 1950 -1999. 5 AGCMs, 80 member, multi-model ensemble. (Attribution of Ocean Forcing) Ø Forced Coupled Atmosphere-Ocean Simulations (AR 4). Observed GHGs, Aerosols and Natural External Forcings for 1950 -1999. (Attribution of External Forcing). Ø Time-slice experiments at ~2050 to understand differences in African climate projections by GFDL and NCAR models.
OBSERVATIONS 20 th Century African Drying Correlated with Warming Tropical Oceans
AGCM SIMULATIONS Time Evolution of African Drying Consistent with Atmospheric Response to Observed Global SST Forcing
Rainfall Trend Patterns for Northern African Wet Season Spatial Pattern of 20 th Century African Drying Consistent with Atmospheric Response to Observed Global SST Forcing
Rainfall Trend Patterns for Southern African Wet Season Spatial Pattern of 20 th Century African Drying Consistent with Atmospheric Response to Observed Global SST Forcing
Attribution Summary I Observed SST Forcing During 1950 -1999 Was the Immediate Attributable Cause for African Drying Trends ü ü Time series of African drying Spatial patterns of African drying Seasonal movement of African drying Reproducibility of the SST forced drying signal
Indian Ocean warming appears most relevant for explaining 20 th Century Southern African drying
Sahel Rainfall Interhemispheric SST Contrast r = 0. 59 NH Warm NH Cold
Atlantic Ocean SST forcing appears most relevant for explaining 20 th Century Sahelian drying
ATLANTIC-ONLY AGCM SIMULATIONS Time Evolution of Sahelian Drying Consistent with Atmospheric Response to Observed Atlantic-basin SST Forcing
Attribution Summary II
Attribution Summary II
Attribution Summary II ü Observed trend amplitudes exceed the GHG+Aerosol Signal by an order of magnitude ü No Projection of the Observed African Rainfall Trends on a GHG+Aerosol Signal ü GFDL CM 2 a most intriguing exception (outlier) for Sahelian Drying
AR 4 CGCM SIMULATIONS & PROJECTIONS
AR 4 CGCM SIMULATIONS & PROJECTIONS
A Tale of Two Models 1950 -1999 AR 4 Simulated Rainfall Trends July-September
A Tale of Two Models 2000 -2049 AR 4 Projected Rainfall Trends July-September
Projected (A 1 B) SST Change by 2049 CCSM 3 +1. 2°C Total Change +1. 0°C Total Change- Global Mean July-September GFDL-CM 2
AGCM and AR 4 Projected JAS Rainfall Changes by 2049
AGCM and AR 4 Projected JAS Rainfall Changes by 2049
Outstanding Questions Q: Is there an attributable cause for the post-1950 African drying trends…PTI? A: Yes. The regionality and seasonality of African precipitation trends are directly attributable to fluctuations in global SSTs. Q: Is there an attributable cause for the post-1950 African drying trends…PT 2? A: The rainfall changes appear beyond those expected from natural coupled ocean- atmosphere variations under modern climate conditions. Q: What role did anthropogenic forcing play in post-1950 African drying? A: The vast majority of AR 4 simulations of 1950 -99 African rainfall trends fail to project upon the observed trend patterns, suggesting GHG&aerosol forcing played little or no role. Q: Is African rainfall sensitive to A 1 B Scenario GHG forcing to 2049? A: Yes. The majority of AR 4 simulations exhibit a positive (wet) Sahel rainfall response, and a negative (dry) southern African rainfall response. Projected SST changes appear to be key, as is the AGCM sensitivity to those SST forcings.
Outstanding Questions Q: Is there an attributable cause for the post-1950 African drying trends…PTI? A: Yes. The regionality and seasonality of African precipitation trends are directly attributable to fluctuations in global SSTs. Q: Is there an attributable cause for the post-1950 African drying trends…PT 2? A: The rainfall changes appear beyond those expected from natural coupled ocean- atmosphere variations under modern climate conditions. Q: What role did anthropogenic forcing play in post-1950 African drying? A: The vast majority of AR 4 simulations of 1950 -99 African rainfall trends fail to project upon the observed trend patterns, suggesting GHG&aerosol forcing played little or no role. Q: Is African rainfall sensitive to A 1 B Scenario GHG forcing to 2049? A: Yes. The majority of AR 4 simulations exhibit a positive (wet) Sahel rainfall response, and a negative (dry) southern African rainfall response. Projected SST changes appear to be key, as is the AGCM sensitivity to those SST forcings.
Outstanding Questions Q: Is there an attributable cause for the post-1950 African drying trends…PTI? A: Yes. The regionality and seasonality of African precipitation trends are directly attributable to fluctuations in global SSTs. Q: Is there an attributable cause for the post-1950 African drying trends…PT 2? A: The rainfall changes appear beyond those expected from natural coupled ocean- atmosphere variations under modern climate conditions. Q: What role did anthropogenic forcing play in post-1950 African drying? A: The vast majority of AR 4 simulations of 1950 -99 African rainfall trends fail to project upon the observed trend patterns, suggesting GHG&aerosol forcing played little or no role. Q: Is African rainfall sensitive to A 1 B Scenario GHG forcing to 2049? A: Yes. The majority of AR 4 simulations exhibit a positive (wet) Sahel rainfall response, and a negative (dry) southern African rainfall response. Projected SST changes appear to be key, as is the AGCM sensitivity to those SST forcings.
Outstanding Questions Q: Is there an attributable cause for the post-1950 African drying trends…PTI? A: Yes. The regionality and seasonality of African precipitation trends are directly attributable to fluctuations in global SSTs. Q: Is there an attributable cause for the post-1950 African drying trends…PT 2? A: The rainfall changes appear beyond those expected from natural coupled ocean- atmosphere variations under modern climate conditions. Q: What role did anthropogenic forcing play in post-1950 African drying? A: The vast majority of AR 4 simulations of 1950 -99 African rainfall trends fail to project upon the observed trend patterns, suggesting GHG&aerosol forcing played little or no role. Q: Is African rainfall sensitive to A 1 B Scenario GHG forcing to 2049? A: Yes. The majority of AR 4 simulations exhibit a positive (wet) Sahel rainfall response, and a negative (dry) southern African rainfall response. Projected SST changes appear to be key, as is the AGCM sensitivity to those SST forcings.
- Slides: 40
