Impact of climatecarbon cycle feedbacks on emissions scenarios

Impact of climate-carbon cycle feedbacks on emissions scenarios to achieve stabilisation Chris Jones (1) Peter Cox (2), Chris Huntingford (3) 1. Hadley Centre, Met Office, Exeter 2. Centre for Ecology and Hydrology, Dorset 3. Centre for Ecology and Hydrology, Wallingford © Crown copyright 2004 Page 1

Outline § Climate-Carbon cycle feedbacks §Uncertainties/intercomparisons §Implications for stabilisation emissions § Results §GCM experiments §Simple “reduced form” model results § Discussion §Uncertainties – between and within models §Reducing uncertainty? Model validation §Defining “optimal” pathways to stabilisation? © Crown copyright 2004 Page 2

Climate Carbon Cycle feedbacks § Well known that climate-carbon cycle models predict a positive feedback § Climate change will reduce the carbon cycle’s ability to sequester CO 2 § Models have consensus on sign (+ve), but not magnitude of feedback (i. e. C 4 MIP) § Uncertainties in the feedback strength mean large uncertainty in: § Future CO 2 levels given an emissions scenario § Permissible emissions to stabilise CO 2 at a given level © Crown copyright 2004 Page 3

Climate Carbon Cycle feedbacks § If climate change weakens natural carbon sinks then we must reduce emissions by more than previously thought to stabilise atmospheric CO 2 § Passing mention in TAR but needs to be brought out more § TAR showed range of permissible emissions but didn’t stress impact of climate feedbacks in reducing these § Huge political implications § Plea to AR 4 authors – Needs to be given more prominence. § Instead of “managing the carbon cycle” this comes under “being managed by the carbon cycle” © Crown copyright 2004 Page 4

WRE scenarios § “WRE” is a family of scenarios of CO 2 level, stabilising at 450, 550, 650, 750 and 1000 ppm § Wigley, Richels and Edmonds. ‘Economic and environmental choices in the stabilisation of atmospheric CO 2 concentrations’. Nature, 1996 § We run the carbon cycle GCM with the prescribed 550 CO 2 scenario and infer the emissions required to achieve it § Results shown in detail for 550 ppm § Summary of results for all levels © Crown copyright 2004 Page 5

WRE 550 CO 2 emissions § Climate feedbacks imply reduced permissible emissions § Lower

WRE 550 cumulative emissions § Similar to previous experiments § Ocean continues to uptake

Reduced Form “simple” model § GCM prohibitively expensive! § Simple model has: § Global means § climate in terms of T § Responds instantly to CO 2 § Carbon cycle calibrated to follow GCM from transient run of Cox et al 2000. § Does good job at matching WRE 550 GCM run § Aim is to give broad idea of response – don’t trust exact details… © Crown copyright 2004 Page 8

WRE 550 CO 2 emissions – simple model ( WRE 550 ) No feedbacks

Permissible Emissions Total emissions, WRE 2000 -2300 without feedbacks with feedbacks Stabilisation at 550 ppm 1355 Gt. C 1010 Gt. C 1393 Gt. C § Without feedbacks, we get close to the WRE result § Climate-Carbon cycle feedbacks significantly reduce the permissible emissions for stabilisation § This is true for stabilisation at any level © Crown copyright 2004 Page 10

Other stabilisation levels Greater reductions at higher stabilisation levels Not surprising given greater level

Uncertainties § Large uncertainties undermine political impact of results § Do we understand them?

Sources of uncertainty § The impact of carbon-cycle feedbacks on permissible emissions will depend on: § “Political” uncertainties: § Chosen level of stabilisation (and hence climate change) § Scientific uncertainties: § Climate sensitivity: Greater sensitivity will mean stronger feedbacks for given CO 2 level § carbon-cycle parameters § vegetation sensitivity to warming/CO 2 § Soil sensitivity § Ocean response to climate/circulation changes § All climate-carbon cycle studies to date show future weakening of the natural carbon sink in response to climate change § But significant uncertainty in strength of feedback © Crown copyright 2004 Page 13

Other models Without feedbacks With feedbacks § UVic model – courtesy of Damon Matthews (in press at GRL) § Stabilisation at 1000 ppm § Significant reduction in allowed emissions © Crown copyright 2004 Page 14

C 4 MIP models Stabilise at 1000 ppm by 2350 Cumulative Emissions Reductions (Gt.

C 4 MIP models Stabilise at 1000 ppm by 2350 C 4 MIP-min (g=0.

Range over C 4 MIP models Stabilise at 1000 ppm by 2350 Cumulative Emissions Reductions (Gt. C) C 4 MIP-mean* (g=0. 14) UVic * = C 4 MIP results estimated from gain factors derived from C 4 MIP transient expts © Crown copyright 2004 Page 18

Implications of uncertainty § 2 main implications of the C 4 MIP uncertainty § Uncertainty does not span zero § All models agree on positive feedback and hence some degree of reduction in permissible emissions § Required emissions vary greatly § Reductions due to climate feedbacks uncertain by almost an order of magnitude © Crown copyright 2004 Page 19

Reducing that uncertainty? § To what extent does the historical record constrain future behaviour? § Climate sensitivity? § No – can’t be well constrained observationally § Causes large spread in future climate and hence in future feedback strength © Crown copyright 2004 Page 20

Climate sensitivity § Uncertainty in historical forcing – especially from aerosols – means large uncertainty in climate sensitivity § TAR shows GCM range from 1. 5 -4. 5, but values up to 8 -10 K can’t be ruled out completely from observations. © Crown copyright 2004 Andreae et al, Nature, 2005 Page 21

Reducing that uncertainty? § To what extent does the historical record constrain future behaviour? § Climate sensitivity? § No – can’t be well constrained observationally § Causes large spread in future climate and hence in future feedback strength § Carbon cycle parameters? § Not directly from observations – CO 2 record can’t distinguish strong fertilisation/strong respiration from weak fertilisation/weak respiration. § But give different future behaviour © Crown copyright 2004 Page 22

Single parameter perturbations WRE 550 WRE 450 CO 2 fert’n Soil resp NPP(T) ∆T 2 x, 1. 5 -4. 5 ∆T 2 x, 1. 5 -10 § Large ensemble of simple model runs with perturbed parameters § In these runs, NPP sensitivity to climate is most important carbon-cycle parameter § More sensitivity than CO 2 fertilisation strength or soil respiration sensitivity to temperature § Similar conclusion to Matthews et al. , GRL, 2005. § Climate sensitivity outweighs carbon cycle uncertainty © Crown copyright 2004 Page 23

Multiple parameter perturbations Low climate sensitivity High climate sensitivity § Varying all these parameters, but still fitting historical emissions, gives only very weak constraint on future permissible emissions § High climate sensitivities lead to requirement for significant NEGATIVE emissions © Crown copyright 2004 Page 24

Reducing that uncertainty? § To what extent does the historical record constrain future behaviour? § Climate sensitivity? § No – can’t be well constrained observationally § Causes large spread in future climate and hence in future feedback strength § Carbon cycle parameters? § Not directly from observations – CO 2 record can’t distinguish strong fertilisation/strong respiration from weak fertilisation/weak respiration. § But give different future behaviour § Model validation? § Maybe – recreating observed behaviour is necessary but not sufficient test of a model § C 4 MIP phase 1 is essential step! © Crown copyright 2004 Page 25

C 4 MIP phase 1 - validation § Atmosphere only model with observed 20 th century SSTs § Just simulate terrestrial carbon cycle § Validate against range of obs: § Site-specific from flux towers § Regional estimates from inversion studies § Interannual variability – e. g. ENSO § Validation is important if we are to know which C 4 MIP models to trust § But, ability to get these right doesn’t constrain future feedback size – merely gives us clues about how to interpret the models § See Jones & Warnier report on Had. CM 3 LC at: § http: //www. metoffice. com/research/hadleycentre/pubs/HCTN/index. html © Crown copyright 2004 Page 26

C 4 MIP phase 1 - validation § Flux tower validation from Carbo. Europe

C 4 MIP phase 1 - validation § Comparison with Trans. Com inversions study (Gurney et al, Nature, 2002) § Regional carbon flux estimates from 1992 -96 § black = transcom § pink = Hadley C 4 MIP experiment § Agrees pretty well in most places © Crown copyright 2004 Page 28

Other potential issues § How important is time to stabilisation? § Emit soon and reduce strongly? Or more gradual? § Can we define an “optimal” pathway? § Sensitivity studies for stabilisation at 550 ppm at different rates: § Idealised profiles with asymptotic approach to stabilisation: § CO 2 = a 0 + a 1 * tanh (a 2 + a 3. τ) § Match CO 2 level and rate of change at 2000 § τ =time to (95%) stabilisation. Range from 20 -150 years. § Not attempted to quantify likelihood – more illustrative § How do climate-carbon cycle feedbacks affect resulting emissions profiles? © Crown copyright 2004 Page 29

‘Optimal’ pathways to stabilisation § “fast” (τ=30) and “slow” (τ=80) emissions profiles to 550

‘Optimal’ pathways to stabilisation § Total 21 st century emissions (higher may be seen

‘Optimal’ pathways to stabilisation § Max rate of required emissions reductions (higher may be

‘Optimal’ pathways to stabilisation § Open Questions: § Can we convert this into “desirability” somehow? § E. g. Linearly combine “total emissions” and “max rate of reduction” § deliberately simplistic – clearly many more factors to consider § “desirability” varies with timescale to stabilisation “worse” § How do climatecarbon cycle feedbacks affect our choice of “optimal”? Shifted optimum? “better” © Crown copyright 2004 Page 33

Conclusions § Climate feedbacks on the carbon cycle will reduce future natural carbon uptake § Hence, to stabilise CO 2, significantly greater emissions reductions may be required § This is true regardless of: § Stabilisation level § But higher levels see greater reduction § Model § But large spread of feedback strength between models § Timescale to stabilise § Strength of feedback may alter “optimal” shape of trajectory as well as magnitude © Crown copyright 2004 Page 34

Conclusions § Large uncertainties between/within models § Observational record directly offers only weak constraint on future behaviour § Validation of complex carbon cycle models against all available data is lacking § Will prove vital to reducing uncertainty © Crown copyright 2004 Page 35