sign in sign up
<<
Home , Climate sensitivity, Radiative forcing

SENSITIVITY AND THE ROLE OF Stephen E. Schwartz

Upton, New York

Joint Meeting ARM and Properties Working Groups Lansdowne VA, November 11-13, 2008 OVERVIEW Earth’s energy balance Perturbations to Earth’s energy balance Radiative forcing by incremental greenhouse gases and aerosols Estimates of climate sensitivity Approaches to determining Earth’s climate sensitivity Concerns with evaluation of climate models over the twentieth century Ð uncertainty in aerosol forcing Implications of uncertainty in climate sensitivity GLOBAL ENERGY BALANCE Global and annual average energy fluxes in watts per square meter

Schwartz, 1996, modified from Ramanathan, 1987 RADIATIVE FORCING A change in a radiative flux term in Earth’s radiation budget, ∆F, W m-2. Working hypothesis: On a global basis radiative forcings are additive and fungible. ¥ This hypothesis is fundamental to the radiative forcing concept. ¥ This hypothesis underlies much of the assessment of over the industrial period. GLOBAL-MEAN RADIATIVE FORCINGS (RF) Pre-industrial to present (Intergovernmental Panel on Climate Change, 2007)

LOSU denotes level of scientific understanding. GLOBAL ENERGY BALANCE Global and annual average energy fluxes in watts per square meter

Schwartz, 1996, modified from Ramanathan, 1987 CLIMATE RESPONSE The change in global and annual mean temperature, ∆T, K, resulting from a given radiative forcing.

Working hypothesis: The change in global mean temperature is proportional to the forcing, but independent of its and spatial distribution.

∆T = S ∆F CLIMATE SENSITIVITY The change in global and annual mean temperature per unit forcing, S, K/(W m-2), S = ∆T/∆F. Climate sensitivity is not known and is the objective of much current research on climate change. Climate sensitivity is often expressed as the temperature for doubled CO2 concentration ∆T2×.

∆T2× = S∆F2× ∆F2× ≈ 3.7 W m-2 CLIMATE SENSITIVITY ESTIMATES THROUGH THE AGES Estimates of central value and uncertainty range from major national and international assessments

6 1.5 Charney Sensitivity, K/(Wm

, K 5 NRC Ð Ð Ð Ð Ð Ð IPCC Ð Ð Ð Ð Ð Ð 2

CO 4

× > 66% 1.0 3 "Likely" 1 σ 2 0.5 -2

1 ) Sensitivity to 2 to 2 Sensitivity 0 0.0 1975 1980 1985 1990 1995 2000 2005 2010 KEY APPROACHES TO DETERMINING CLIMATE SENSITIVITY ¥ Paleoclimate studies: Forcing and response over time scales from thousands to millions of years. CLIMATE SENSITIVITY ESTIMATES THROUGH THE AGES Estimates of central value and uncertainty range from specific approaches and major national and international assessments

6 1.5 Charney Sensitivity, K/(Wm

, K 5 NRC Ð Ð Ð Ð Ð Ð IPCC Ð Ð Ð Ð Ð Ð 2

CO 4

× > 66% 1.0 Paleo 3 "Likely" 1 σ 2 1 σ 0.5 -2

1 Hoffert ) Sensitivity to 2 to 2 Sensitivity Covey 0 0.0 1975 1980 1985 1990 1995 2000 2005 2010

Climate sensitivity from paleo climate has been a major contributor to present assessment of climate sensitivity. KEY APPROACHES TO DETERMINING CLIMATE SENSITIVITY ¥ Paleoclimate studies: Forcing and response over time scales from thousands to millions of years. ¥ Empirical: Forcing and response over the instrumental record. CLIMATE SENSITIVITY ESTIMATES THROUGH THE AGES Estimates of central value and uncertainty range from specific approaches and major national and international assessments

6 1.5 Charney Empirical Sensitivity, K / (W m

, K 5 2 NRC – – – – – – IPCC – – – – – –

CO 4 × > 66% 1.0 Paleo 3 "Likely" ~1 σ 2 1 σ 1 σ 0.5 Gregory -2

1 Hoffert ) Sensitivity to 2 Covey 0 0.0 1975 1980 1985 1990 1995 2000 2005 2010

Empirical approach does not greatly constrain sensitivity because of uncertainty in aerosol forcing over the period of instrumental record. KEY APPROACHES TO DETERMINING CLIMATE SENSITIVITY ¥ Paleoclimate studies: Forcing and response over time scales from thousands to millions of years. ¥ Empirical: Forcing and response over the instrumental record. ¥ Climate modeling: Understanding the processes that comprise Earth’s and representing them in large-scale numerical models. CLIMATE SENSITIVITY ESTIMATES FROM GLOBAL CLIMATE MODELS 18 Current global climate models Ð IPCC AR4, 2007

6 1.5 Sensitivity, K/(Wm

, K 5 2

CO 4 × 1.0 3

2 0.5 -2

1 ) Sensitivity to 2 to 2 Sensitivity 0 0.0

Range of model sensitivities is identical with range of current overall IPCC sensitivity estimate. TOO ROSY A PICTURE? Ensemble of 58 model runs with 14 global climate models

Simulations that incorporate anthropogenic forcings, including increasing “ concentrations and the effects of aerosols, and that also incorporate natural external forcings provide a consistent explanation of the observed temperature record. These simulations used models with different climate sensitivities, rates of “ ocean heat uptake and magnitudes and types of forcings. TOO ROSY A PICTURE? Ensemble of 58 model runs with 14 global climate models

Factor of 2

Factor of 4

Schwartz, Charlson & Rodhe, Nature Reports Ð Climate Change, 2007 Uncertainty in modeled temperature increase Ð less than a factor of 2, red Ð is well less than uncertainty in forcing Ð a factor of 4, green. NEWS OF THE WEEK 6 JULY 2007 VOL 317 SCIENCE www.sciencemag.org Published by AAAS

CLIMATE CHANGE –RICHARD A. KERR Another Global Warming Icon Comes Under Attack Climate scientists are used to skeptics taking rather well (see figure). A narrow range of gas changes are well known, they note, but potshots at their favorite line of evidence for simulated warmings (purple band) falls right not so the counteracting cooling of pollutant global warming. It comes with the territory. on the actual warming (black line) and dis- hazes, called aerosols. Aerosols cool the But now a group of mainstream atmospheric tinctly above simulations run under condi- by reflecting away and scientists is disputing a rising icon increasing the reflectivity of of global warming, and researchers 1.0 . Somehow, the three are giving some ground. Models using only natural forcings researchers say, modelers failed The challenge to one part of Models using both natural and to draw on all the uncertainty the latest climate assessment by anthropogenic forcings inherent in aerosols so that the 0.5 the Intergovernmental Panel on Observations 20th-century simulations look Climate Change (IPCC) “is not a more certain than they should. question of whether the Earth is Modeler Jeffrey Kiehl of the warming or whether it will con- 0.0 National Center for Atmospheric

tinue to warm” under human anomaly (°C) Temperature Research in Boulder, Colorado, influence, says atmospheric scien- reached the same conclusion by a tist Robert Charlson of the Uni- 1900 1950 2000 different route. In an unpublished versity of Washington, Seattle, Not so certain. The uncertainty range in the modeled warming (red bar) is but widely circulated analysis, he one of three authors of a commen- only half the uncertainty range (orange) of human influences. plotted the combined effect of tary published online last week in greenhouse gases and aerosols Nature Reports: Climate Change. used in each of 11 models versus Instead, he and his co-authors argue that tions free of human influence (blue band). how responsive each model was to a given the simulation by 14 different climate mod- But the group of three atmospheric scien- amount of greenhouse gases. The latter fac- els of the warming in the 20th century is not tists—Charlson; Stephen Schwartz of the tor, called climate sensitivity, varies from the reassuring success IPCC claims it to be. Brookhaven National Laboratory in Upton, model to model. He found that the more sen- Future warming could be much worse than New York; and Henning Rodhe of Stock- sitive a model was, the stronger the aerosol that modeling suggests, they say, or even holm University, Sweden—says the close cooling that drove the model. The net result more moderate. IPCC authors concede the match between models and the actual warm- of having greater sensitivity compensated by group has a point, but they say their report— ing is deceptive. The match “conveys a lot a greater aerosol effect was to narrow the if you look in the right places—reflects the more confidence [in the models] than can be apparent range of uncertainty, as Schwartz uncertainty the critics are pointing out. supported in actuality,” says Schwartz. and his colleagues note. Twentieth-century simulations would To prove their point, the commentary “I don’t want certain interests to claim seem like a straightforward test of climate authors note the range of the simulated warm- that modelers are dishonest,” says Kiehl. models. In the run-up to the IPCC climate ings, that is, the width of the purple band. The “That’s not what’s going on. Given the range science report released last February (Science, range is only half as large as they would of uncertainty, they are trying to get the best 9 February, p. 754), 14 groups ran their mod- expect it to be, they say, considering the large fit [to observations] with their model.” els under 20th-century conditions of rising range of uncertainty in the factors driving cli- That’s simply a useful step toward using a greenhouse gases. As a group, the models did mate change in the simulations. Greenhouse- model for predicting future warming. L TOO ROSY A PICTURE? Ensemble of 58 model runs with 14 global climate models

Factor of 2

Factor of 4

Schwartz, Charlson & Rodhe, Nature Reports Ð Climate Change, 2007 The models did not span the full range of the uncertainty and/or . . . The forcings used in the model runs were anticorrelated with the sensitivities of the models. CORRELATION OF AEROSOL FORCING, TOTAL FORCING, AND SENSITIVITY IN CLIMATE MODELS Eleven models used in 2007 IPCC analysis Sensitivity, K/(W m-2) 2.0 1.5 1.0 0.8 0.7 0.6 0.5 2.2 2.2

2.0 2.0

1.8 1.8 -2 -2

1.6 1.6

1.4 1.4

1.2 1.2 Total forcing, W m Total forcing, W m Total forcing, 1.0 1.0

0.8 0.8 Correlation coefficient r = 0.79 Line denotes slope = 1 0.6 0.6 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 -1.6 -1.2 -0.8 -0.4 Inverse Sensitivity, W m-2/K Aerosol forcing, W m-2 Modified from Kiehl, GRL, 2007 Climate models with lower sensitivity (higher inverse sensitivity) employed a greater total forcing. Greater total forcing is due to lower magnitude (less negative) aerosol forcing. IMPLICATIONS OF UNCERTAINTY IN CLIMATE SENSITIVITY Uncertainty in climate sensitivity translates directly into . . . ¥ Uncertainty in the amount of incremental atmospheric CO2 that would result in a given increase in global mean surface temperature. ¥ Uncertainty in the amount of fossil carbon that can be combusted consonant with a given climate effect. At present this uncertainty is about a factor of 3. This uncertainty is attributable largely to uncertainty in aerosol forcing. Reduction in uncertainty in aerosol forcing is essential to reducing uncertainty in climate sensitivity.