The 20 th century carbon budget simulated with CCCma first generation earth system model (CanESM1)

Transcription

1 1/21 The 2 th century carbon budget simulated with CCCma first generation earth system model (CanESM1) Vivek K Arora, George J Boer, Charles L Curry, James R Christian, Kos Zahariev, Kenneth L Denman, Gregory M Flato, John F Scinocca, William J Merryfield Canadian Centre for Climate Modelling and Analysis Environment Canada

2 2/21 CCCma Earth System Model (CanESM1) Physical atmosphere and land surface Physical Ocean CO 2 and non- CO 2 GHGs Terrestrial carbon cycle component Ocean carbon cycle component AGCM3 T47/L31 ~3.75º resolution CLASS land surface scheme (v2.7) OGCM3.5 CO 2 prognostic CH 4, N 2 O, CFC-11 and CFC-12 specified Canadian Terrestrial Ecosystem Model (CTEM) Canadian Model of Ocean Carbon (CMOC)

3 3/21 Canadian Terrestrial Ecosystem Model (CTEM) G R gl R ml Leaf litter, D L CTEM 1. Plant Functional Types Leaves, C L 1 Needleleaf Evergreen 2 Needleleaf Deciduous Stem litter, D S A stem 3 Broadleaf Evergreen 4 Broadleaf Cold Deciduous 5 Broadleaf Dry Deciduous Stem, C S R gs R ms 6 C 3 Crop 7 C 4 Crop 8 C 3 Grass R hh R hd A roots R gr R mr 9 C 4 Grass Litter Pool, C D C D H Soil Carbon Pool, C H Root mortality, D R Roots, C R

4 4/21 Canadian Model of Ocean Carbon (CMOC) CMOC incorporates an inorganic chemistry module (solubility pump) and an ecosystem model (organic and carbonate pumps). The inorganic module has dissolved inorganic carbon (DIC) and total alkalinity as prognostic variables. N Nutrients P Phytoplankton Z Zooplankton D - Detritus N 2 fixation 1 m N CO 2 uptake f np N uptake f dn remineralization f pd phytoplankton mortality F CaCO3 sinking inorganic C CMOC v.1 P f aggr phytoplankton Mortality due to aggregation D f zn zooplankton respiration Chl f pz, grazing F PON, sinking organic C g a f pz, grazing fraction utilized Z (1-g a )f pz grazing fraction wasted Base of euphotic zone f zd1 linear and quadratic f zd2 mortalities of zooplankton

5 5/ Gross primary productivity (GPP), Pg C/yr Simulated GPP CTEM driven with observed CO 2 and reanalysis repeatedly. Luo et al. [1996], GBC, 1(2), J c, Rubisco limited photosynthesis rate J e, Light limited photosynthesis rate I 1 I 2 = = 1 J c 1 J e dj dc c i dj dc e i Year Ambient CO 2 (ppm) CTEM I 1 I 2 Luo et al. [1996] I 1 I Average CTEM GPP response over period. 1 dg p G dc p i Intercellular CO 2 (c i )(ppm)

6 6/21 Gross primary productivity (GPP), Pg C/yr Net primary productivity (NPP), Pg C/yr Simulated GPP Fitted curve CTEM driven with observed CO 2 and reanalysis repeatedly b) Simulated NPP Fitted curve Year Year G p ( t) = G p 1 + γ g C( t) ln C N p ( t) = N p 1 + γ γ g =.9 =1. 23 γ n n C( t) ln C Ricciuto et al. [28], GBC, 22, 26GB298. Friedlingstein et al. [1995], GBC, 9(4), Norby et al. [25], PNAS, 12, γ n =.71 γ n =. 68 γ n =. 6 95% CI ( )

7 7/21 Tons of literature suggesting down-regulation of terrestrial photosynthesis due to nutrient limitation and source/sink imbalance Although there are experimental sites suggesting that if nitrogen availability keeps up then down-regulation does not occur. And, of course, N fixing species do not experience any down-regulation. But the globally-averaged effect of down-regulation remains uncertain. G G p d ( t) = G ( t) = G p p 1 + γ 1 + γ g gd C( t) ln C C( t) ln C ξ( C) = 1+ γ 1+ γ gd g ln ln ( C / C ) ( C / C ) G = ξ ( C) d G p

8 8/21 Use results from experimental studies that grow plants and ambient and elevated CO 2 to infer down-regulation. Study Study Type Ambient CO 2 (ppm) Elevated CO 2 (ppm) Estimated downregulation factor ξ Inferred value of γ gd 1 Ainsworth et al. (23) Experimental Ainsworth et al. (24) Experimental Adam et al. (24) Experimental Bigras and Bertrand (26) Experimental Medlyn et al. (1999) Meta-analysis McGuire et al. (1995) 1 Meta-analysis ~35 ~ Mean.42 What the model response should be? γ gd =.42 ξ( C) = 1+ γ 1+ γ gd g ln ln ( C / C ) ( C / C ) Down-regulation Model response to CO 2 Ambient and elevated CO 2

9 9/21 Down-regulation of terrestrial photosynthesis rates as a function of atmospheric CO ξ CO 2 (ppm) 1% down-regulation in year 2 with CO2 conc. ~37 ppm Compare that to vegetation type dependent 16% to 4% reduction in Common Land Model (CLM) of Community Climate System Model (CCSM) [Oleson et al., 28; Stöckli et al., 28]

10 1/21 G p ( t) = G p Before down-regulation 1 + γ g C( t) ln C N p ( t) = N p 1 + γ n C( t) ln C γ g =.9 =1. 23 After down-regulation γ g =. 51 =. 69 γ n γ n Ricciuto et al. [28], GBC, 22, 26GB298. Friedlingstein et al. [1995], GBC, 9(4), Norby et al. [25], PNAS, 12, γ n =.71 γ n =. 68 γ n =. 6 95% CI ( )

11 11/21 Can t simulate historical CO 2 without including anthropogenic land use change Leaves, C L Combustion - 15% Paper- 7% Stem, C S Litter Pool, C D Soil Carbon Pool, C H Furniture - 15% Roots, C R

12 12/21 Can t simulate historical CO2 without including anthropogenic land use change Historical changes in crop area. Changes in pasture area are not taken into account, yet.

13 13/21 Treatment of LUC emissions Interactive LUC Specified LUC emissions Terrestrial photosynthesis down-regulation No Simulation A Simulation B Yes Simulation C Simulation D

14 14/21 Observed Simulation A, no down-regulation + interactive LUC Simulation B, no down-regulation + Houghton s LUC emissions Simulation C, down-regulation + interactive LUC emissions Simulation D, down-regulation + Houghton s LUC emissions Land-atmosphere CO 2 flux (Pg C/yr) CO2 (ppm) Atmospheric CO 2 (ppm) 1 Compare ppm range.5 with ppm range from C 4 MIP models Ocean-atmosphere CO 2 flux (Pg C/yr) Year 1 IPCC AR4 estimate for 198s with error bars.5 IPCC AR4 estimate for 199s with error bars

15 15/21 Zonal distribution of land- and ocean-atmosphere fluxes (gc/m 2.year) averaged over the period. Simulation A, no down-regulation + interactive LUC Simulation B, no down-regulation + Houghton s LUC emissions Simulation C, down-regulation + interactive LUC emissions Simulation D, down-regulation + Houghton s LUC emissions Land-atmosphere CO 2 flux Ocean-atmosphere CO 2 flux 3 2 Sink of atm. CO Takahashi et al. (28) ocean CO 2 flux for year Sink of atm. CO Source of atm. CO Source of atm. CO Latitude Latitude Large tropical source not consistent with inversion studies when using specified LUC emissions.

16 16/21 Cumulative fluxes (Pg C) Simulation Observation-based estimate A B C D Fossil fuel, cement and biofuel emissions Change in atmospheric carbon burden Land uptake (remainder) Ocean uptake ±19 115±17 (excluding areas poleward of 65º) Land and ocean uptake

17 17/21 Zonal and temporal behaviour of observation-based and simulated CO 2 concentrations a) Observation-based b) CanESM CO 2 (ppm) CO 2 (ppm) Latitude 5 Latitude Year Year 2 5 Latitude Latitude Year Year

18 18/21 Annual CO 2 cycle and latitudinal structure (1991-2) Monthly CO 2 anomalies (ppm) CanESM1 (1991-2) Observation-based (1991-2) TransCom models Latitudinal CO 2 anomalies (ppm) Month Latitude

19 19/ Latitude 2 North-South CO 2 gradient 1998 Year years of control simulation Years of transient simulation Latitude CO2 (ppm) Latitudinal CO 2 anomalies (ppm) Latitudinal CO2 anomalies (ppm) 1 years of control run Latitude

20 2/21 North-South CO 2 gradient (difference between Mauna Loa and South Pole CO 2 concentrations) AR4, Chapter 7, Denman et al. [27] CO 2 N/S gradient =.5 Emissions -.81 CO 2 N/S gradient =.46 Emissions -.52 CO 2 north-south gradient (ppm) Fossil fuel emissions (Pg C/yr)

21 21/21 Summary Modellers usually don t quantify their net terrestrial CO 2 fertilization effect. Implementation of down-regulation inferred from experimental studies yields rate of increase of NPP with CO 2 that is consistent with other studies. Implementation of terrestrial photosynthesis down-regulation yields better agreement with observations. Annual CO 2 cycle and inter-hemispheric CO 2 gradient are simulated reasonably well. Latitudinal gradient as a function of emissions also compares well with observations. Ocean uptake on the lower side. Anomalous southern hemisphere ocean uptake and release.