The Search for a Human Fingerprint in the Changing Thermal Structure of the Atmosphere

Transcription

1 The Search for a Human Fingerprint in the Changing Thermal Structure of the Atmosphere Ben Santer Program for Climate Model Diagnosis and Intercomparison Lawrence Livermore National Laboratory Third Workshop on Understanding Climate Change from Data Northwestern University, Evanston, Illinois August 15, 2013 CMIP-5 models (ALL+8.5)! Observations (Santa Rosa)! Pressure (hpa)! Pressure (hpa)! Trend over 1979 to 2012 ( C/decade)!

2 My collaborators and co-authors Jeff Painter, Celine Bonfils, Peter Gleckler, Charles Doutriaux, Karl Taylor (Lawrence Livermore National Laboratory) Carl Mears, Frank Wentz (Remote Sensing Systems) Susan Solomon (M.I.T.) Tom Wigley (University of Adelaide) Gavin Schmidt (NASA/Goddard Institute for Space Studies) Nathan Gillett (Canadian Climate Centre) Peter Thorne (National Climatic Data Center) 2

3 Structure Identifying human influences on atmospheric temperature Æ Early fingerprint work (mid-1990s) Æ Update with latest satellite data and CMIP-5 simulations Comparing modeled and observed temperature variability Conclusions 3

4 Different factors that influence climate have different fingerprints Source: Global Climate Change Impacts in the United States (Karl et al., 2009; modified from Santer et al., 1996, 2006)

5 Different factors that influence climate have different fingerprints

6 Weather balloon estimates of atmospheric temperature change are consistent with human influence fingerprints Source: Allen and Sherwood, Nature Geoscience (2008) Trend ( C/decade; changes over 1970 to 2005)

7 Structure Identifying human influences on atmospheric temperature Æ Early fingerprint work (mid-1990s) Æ Update with latest satellite data and CMIP-5 simulations Comparing modeled and observed temperature variability Conclusions 7

8 Measuring atmospheric temperature from space Figure and text courtesy of Carl Mears, Remote Sensing Systems Higher temperatures = more microwave emissions from oxygen molecules By choosing different microwave frequencies, different layers in the atmosphere can be measured 8

9 Our fingerprint study uses zonal-mean changes in the temperature of broad atmospheric layers 9

10 The changing thermal structure of the atmosphere in the latest observations and model simulations Pressure (hpa)! CMIP-5 models (Human effects)! Pressure (hpa)! Observations (Santa Rosa)! Source: Santer et al., PNAS (2013b; in press) Trend ( C/decade over 1979 to 2012)! 10

11 Fingerprint detection explained pictorially. Observations 1979! 1980! 1981! 1982! 1983! 1984!!! 2010! 2011! Spatial covariance, ANT fingerprint and OBS! 0.5! 0.0! -0.5! 1980! 1985! 1990! 1995! 2000! 2005! 2010! Projection onto model fingerprint Projection time series Model ANTHRO fingerprint 11

12 Global-mean lower stratospheric temperature changes in CMIP-5 pre-industrial control runs (CTL) 12

13 Fingerprint detection explained pictorially. Control run 1! 2! 3! 4! 5! 6!!! 3999! 4000! Spatial covariance, ANT fingerprint and CTL! 0.5! 0.0! -0.5! 0! 1000! 2000! 3000! 4000! Projection onto model fingerprint Projection time series Model ANTHRO fingerprint 13

14 Estimating signal-to-noise ratios Trend length (years)! 12! 17! 22! 27! 32! Signal-to-noise ratio! 10! 8! 6! 4! 2! Santa Rosa observations Alabama observations 0! 1990! 1995! 2000! 2005! 2010! Last year of trend! 14

15 Global-mean lower stratospheric temperature changes in CMIP-5 simulations with solar and volcanic forcing (NAT) 15

16 Estimating signal-to-noise ratios Santa Rosa (CTL noise) U. Alabama (CTL noise) Santa Rosa (NAT noise) U. Alabama (NAT noise) 16

17 Global-mean lower stratospheric temperature changes in CMIP-5 Last Millennium simulations (P1000) 17

18 Estimating signal-to-noise ratios Santa Rosa (CTL noise) U. Alabama (CTL noise) Santa Rosa (NAT noise) U. Alabama (NAT noise) Santa Rosa (P1000 noise) U. Alabama (P1000 noise) 18

19 Why do we obtain such large S/N ratios? Pressure (hpa)! CMIP-5 models (Human effects)! Pressure (hpa)! Observations (Santa Rosa)! Source: Santer et al., PNAS (2013b; in press) Trend ( C/decade over 1979 to 2012)! 19

20 The dominant patterns of internal and total natural variability do not look like the searched-for fingerprint A CTL EOF1 (31.85%)! B CTL EOF2 (23.21%)! C CTL EOF3 (17.31%)! D NAT EOF1 (34.72%)! E NAT EOF2 (18.38%)! F NAT EOF3 (15.3%)! G P1000 EOF1 (48.84%)! H P1000 EOF2 (15.73%)! I P1000 EOF3 (12.07%)! EOF loading!

21 Structure Identifying human influences on atmospheric temperature Æ Early fingerprint work (mid-1990s) Æ Update with latest satellite data and CMIP-5 simulations Comparing modeled and observed temperature variability Conclusions 21

22 Comparing modeled and observed variability Slow (5 to 20 years) Fast (< 2 years) 22

23 Do CMIP-5 models underestimate the observed slow variability of tropospheric temperature? MODELS! Quadrant of doom! Slow temperature variability ( C)! OBSERVATIONS! Santa Rosa! Univ. Alabama! Fast temperature variability ( C)! Source: Santer et al., PNAS (2013a) 23

24 Do CMIP-5 models underestimate the observed slow variability of tropospheric temperature? MODELS! Slow temperature variability ( C)! OBSERVATIONS! Uncertainty! Santa Rosa! Univ. Alabama! Fast temperature variability ( C)! Source: Santer et al., PNAS (2013a) 24

25 Do CMIP-5 models underestimate the observed slow variability of tropospheric temperature? Slow temperature variability ( C)! MODELS! Model A! Model B! Model C! Model D! Model E! Model F! OBSERVATIONS! Santa Uncertainty! Rosa! Santa Rosa! Univ. Alabama! Fast temperature variability ( C)! Source: Santer et al., PNAS (2013a) 25

26 Do CMIP-5 models underestimate the observed slow variability of tropospheric temperature? Slow temperature variability ( C)! MODELS! Model A! Model B! Model C! Model D! Model E! Model F! Model G! Model H! Model I! Model I! Model J! OBSERVATIONS! Uncertainty! Santa Rosa! Univ. Alabama! Fast temperature variability ( C)! Source: Santer et al., PNAS (2013a) 26

27 Do CMIP-5 models underestimate the observed slow variability of tropospheric temperature? Slow temperature variability ( C)! MODELS! Model A! Model B! Model C! Model D! Model E! Model F! Model G! Model H! Model I! Model I! Model J! Model K! Model L! Model M! Model N! Model O! Model P! Model Q! Model R! Model R! OBSERVATIONS! Uncertainty! Santa Rosa! Univ. Alabama! Fast temperature variability ( C)! Source: Santer et al., PNAS (2013a) 27

28 Do CMIP-5 models underestimate the observed slow variability of tropospheric temperature? Slow temperature variability ( C)! MODELS! Model A! Model B! Model C! Model D! Model E! Model F! Model G! Model H! Model I! Model I! Model J! Model K! Model L! Model M! Model N! Model O! Model P! Model Q! Model R! Model R! Model average! OBSERVATIONS! Uncertainty! Santa Rosa! Univ. Alabama! Fast temperature variability ( C)! Source: Santer et al., PNAS (2013a) 28

29 Conclusions A human-caused latitude/altitude pattern of atmospheric temperature change is consistently identifiable in satellite observations This fingerprint of tropospheric warming and stratospheric cooling can be discriminated from the background noise of: Æ Internal climate variability (CTL) Æ Variability caused by natural changes in solar irradiance and volcanic aerosol loadings (NAT, P1000) Our significance testing framework is highly conservative Æ NAT and P1000 total noise estimates include volcanic eruptions and solar irradiance changes much larger than those observed over the satellite era Internal and total natural variability cannot produce sustained global-scale warming of the troposphere and cooling of the stratosphere 29