Simulating changes in the terrestrial biosphere over glacial/interglacial timescales
P. K¨ohler & H. Fischer
Alfred Wegener Institute for Polar and Marine Research
P.O. Box 12 01 61, D-27515 Bremerhaven, Germany, email: pkoehler@awi-bremerhaven.de
Abstract
The state of the terrestrial biosphere during the Holoce- ne and the Last Glacial Maximum (LGM) was estimated from data bases and steady state simulations in former studies. Here, we used these previous estimates and run a simple globally averaged box model of the terrestrial carbon stocks forced by various paleo records (tempe- rature, CO2, sea level) from the Last Glacial Maximum (LGM) over termination I to the Holocene to determi- ne which forcing factors might be appropriate to ex- plain observed changes in the biosphere. Former forcing strength of this type of model on recent climate changes were to large to explain glacial/interglacial variations.
The terrestrial carbon stock at LGM seemed to consist of about 1600 PgC, 600 PgC less than in preindustrial times. During the transition the oceanic release of car- bon during the last 20 ky seemed to be in phase with the atmospheric CO2record, but four times larger than the CO2 increase due to the build-up of the terrestri- al stocks. Calculated changes in the isotopic signature of oceanicδ13C correspond well with data and suggest not only a dominant role of the biosphere during the stable climate conditions such as the LGM or the Holo- cene, but also a relevant influence on atmosphericδ13C during the transition.
Introduction
Variations of the earth climate determined also the chan- ges in the carbon stocks of the terrestrial biosphere. It is currently discussed that the glacial/interglacial chan- ges in atmospheric pCO2andδ13C were mainly controlled by processes in the ocean, while the isotopic signature of CO2during the Holocene and the LGM might point to dominant variations in the terrestrial biosphere car- bon stocks (Inderm¨uhle et al. 1999; Broecker et al. 2001;
Fischer et al. 2003).
Here, we try to add to current discussions by a transi- ent modelling approach driven by paleoclimatic records.
This directly implies the use of a simple model, otherwise missing data constraints might bias model dynamics.
Time depending driving forces of the model. A: sea level changes derived from coral reef terraces (Fairbanks, 1990). B: atmospheric CO2 concentrations mea- sured in the Taylor Dome ice core (Smith et al., 1999). C: Isotopes records in ice cores as proxies for temperature changes. GISP2, Greenland (Grootes und Stuiver, 1997) and Vostok, Antarktis (Jouzel et al., 1987).
-120 -100 -80 -60 -40 -20
sealevel[m] A
coral reefs
200 220 240 260 280
pCO2[ppmv] B
pCO2 Taylor Dome
-7.0 -6.8 -6.6 -6.4 -6.2 -6.0
13C[o/oo]
13C
0 5 10 15 20 25 30
Time [kyr BP]
-46 -44 -42 -40 -38 -36 -34
GISP218O[o/oo]
C
GISP2
-480 -460 -440 -420 -400
VostokD[o/oo]
Vostok
Model and Data
The box model of the terrestrial biosphere used he- re was based on previous studies (Emanuel et al. 1984;
Kheshgi & Jain 2003). We finally parameterized our mo- del with data from literature which led to a representation of 2200 PgC (vegetation: 700 PgC; soils: 1500 PgC) in the terrestrial biosphere during preindustrial times.
The model was forced with various paleo records, which changed through CO2fertilization, metabolic changes in NPP and respiration, and available land area the amount of carbon bound in the terrestrial biosphere. While sea level change had only minor impacts on the carbon sto- rage at land (∼ 4%), both temperature and CO2 ef- fects were solely capable of explaining the observed glaci- al/interglacial increase in the terrestrial biosphere. Thus, we identified acceptable forcing functions by a intensive sensitivity study and by comparison with other models.
Structure of the model BICYCLE (Box model of the Isotopic Carbon cYCLE), whose terrestrial biosphere was used here. Compartments: C4 and C3 ground vegetation, non-woody (NW) and woddy (W) parts of trees, detritus (D), fast (FS) and slow (SS) decomposing soil. Arrows indicate the fluxes of carbon.
C3
FS SS
NW W D C4
Biosphere Atmosphere
Rock
Sediment
SEATL
SANT
TEI−P SEI−P
SNPAC SNATL
TEATL
DATL DANT DI−P
0.4 2.1
1.4 17.1
4.4 17.4 8.7 0.2 3.8 0.1
1.0 2.8 1.7 0.4
9.4 28.4 23.6 5.7 2.9 2.6
2.6 20.8 9.0 3.8 10.8 16.8 23.0 4.3 0.5 0.2
5.6 7.6 17.6 18.6
0.8 C3
FS SS
NW W D C4
Biosphere
Atmosphere
Box model of the Isotopic Carbon cYCLE
BICYCLE
Atlantic Indo−Pacific
Biosphere
SO
The impact of different climate amplitudes in CO2 and temperature on total biospheric carbon was tested in steady state simulations.
Case I: enhenced CO2 fertilization. Case II: enhenced temperature dependence.
-100 -50 0
(pCO2) [ppm]
-1000 -500 0 500
(Carbon)[PgC] A
-100 -50 0
(pCO2) [ppm]
0 2 4 6
C/pCO2[PgC/ppm]
C/ pCO2 C
-100 -50 0
(pCO2) [ppm]
-1000 -500 0 500
(Carbon)[PgC] B
-100 -50 0
(pCO2) [ppm]
0 2 4
C/pCO2[PgC/ppm]
C/ pCO2 C
-10 -8 -6 -4 -2 0 T [K]
-1000 -500 0 500
(Carbon)[PgC] C
-10 -8 -6 -4 -2 0 2 T [K]
0 30 60 90 C/T[PgC/K]
C/ T C
-10 -8 -6 -4 -2 0 T [K]
-1000 -500 0 500
(Carbon)[PgC] D
-10 -8 -6 -4 -2 0 2 T [K]
0 30 60 90 C/T[PgC/K]
C/ T C
Results
Forcings used in recent studies were always too strong to explain the glacial terrestrial carbon content. The average terrestrial C at the LGM based on 10000 different forcing combinations was stabil around 1600 PgC, but its uncer- tainties was largely reduced if various filter functions were additionly applied.
We identify two cases of about 15 simulations whose re- sults correspond well with former case studies. They differ in the strength of CO2fertilization and temperature ef- fects.
It was now possible to determine the resulting carbon fluxes of the ocean to the atmosphere/biosphere subsy- stem by mass balance calculations. The ocean released about 800 PgC during the transition, from which a fourth stayed in the atmosphere and 3/4 cummulated in the ter- restrial stocks. Oceanic inorganic carbon was becoming 0.4◦/◦◦heavier during the G/IG transition, which was in good agreement with data constraints. This implies that the terrestrial biosphere might also be the driving forces for the dynamics in atmosphericδ13C during the transiti- on.
Simulation results of terrestrial carbon at LGM for different sce- narios and different targets/filters. Yellow:
previous estimates.
Scenario 600
800 1000 1200 1400 1600 1800 2000 2200 2400
Carbon[PgC]
Terrestrial carbon at LGM Terrestrial carbon at LGM
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
SD/mean[-]
Scenario 600
800 1000 1200 1400 1600 1800 2000 2200 2400
Carbon[PgC]
Terrestrial carbon at LGM
KheshgiA KheshgiB Kicklighter Allvariations C fC4 NPP C+fC4 C+NPP NPP+fC4 C+fC4+NPP CaseI CaseII
Transient modeling results of the simulated total biospheric carbon. A: Case I with enhenced CO2 fertilization. B: Case II with enhenced temperature depen- dence. Averages (thick black line) and 1 SD (grey area) are shown. Cumulative carbon fluxes from atmosphere/biosphere to the ocean (C: total carbon. D: iso- topic signature). Simulations of case I (enhenced CO2 fertilization) and case II (enhenced temperature dependence) compared with Taylor Dome ice core data.
1500 1600 1700 1800 1900 2000 2100 2200
0.0 0.2 0.4 0.6 0.8 1.0
normalized[-]
sealevel T (S) T (N) pCO2 total C
A
1500 1600 1700 1800 1900 2000 2100 2200
TotalC[PgC]
1500 1600 1700 1800 1900 2000 2100 2200
TotalC[PgC]
0.0 0.2 0.4 0.6 0.8 1.0
normalized[-]
sealevel T (S) T (N) pCO2 total C
B
1500 1600 1700 1800 1900 2000 2100 2200
TotalC[PgC]
200 400 600 800
Co2ab[PgC]Cumulativeflux
case II case I data pCO2(atm)
C
450 500 550 600
pCO2(atm)[PgC]
0.0 0.1 0.2 0.3 0.4 0.5
13Co[o/oo] D
Oceanicincrease
0 5 10 15 20 25
Time [ky BP]
-7.0 -6.8 -6.6 -6.4 -6.2 -6.0
13Catm[o/oo] case II case I data13Catm
References
Broecker, W. S. et al. 2001: G3 2, 2001GC000177.
Emanuel, W. R. et al., 1984. Ecology 65:970–983.
Fairbanks, R. G. 1990. Paleoceanography 5:937–948.
Fischer, H. et al. 2003. Mem Natl. Inst. Polar Res., Spc. Issue 57: 121-138.
Grootes, P. M., M. Stuiver, 1997. J. Geophys Res 102, 26455-26470.
Inderm¨uhle, A. et al. 1999. Nature 398:121–126.
Jouzel, J. et al. 1987. Nature 329, 403-408.
Kheshgi, H. S., A. K. Jain, 2003. GBC 17, 1047, doi: 101029/2001GB001842.
Smith, H. J. et al. 1999. Nature 400:248–250.
