- Email: [email protected]
Was the Antarctic glaciation delayed by a high degassing rate during the early Cenozoic?
Earth and Planetary Science Letters 371-372 (2013) 203–211
Contents lists available at SciVerse ScienceDirect
Earth and Planetary Science Letters journal homepage: www.elsevier.com/locate/epsl
Was the Antarctic glaciation delayed by a high degassing rate during the early Cenozoic? Vincent Lefebvre a,b,n, Yannick Donnadieu a, Yves Goddéris b, Frédéric Fluteau c, Lucie Hubert-Théou d a
Laboratoire des Sciences du Climat et de l’Environnement, CNRS-CEA, CEA Saclay, Orme des Merisiers, F-91190 Gif sur Yvette, France Géosciences-Environnement Toulouse, Observatoire Midi-Pyrénées CNRS, 14 Avenue Edouard Belin, F-31400 Toulouse, France c Institut de Physique du Globe de Paris, UMR 7154, Sorbonne Paris Cité, 1 rue Jussieu, F-75238 Paris, France d Department of Earth and Planetary Sciences, McGill University, Montreal, Canada b
art ic l e i nf o
a b s t r a c t
Article history: Received 3 August 2012 Received in revised form 25 March 2013 Accepted 31 March 2013 Editor: J. Lynch-Stieglitz Available online 24 April 2013
The Cenozoic is a period of major climatic changes marked by the formation of the Antarctic ice sheet at the Eocene/Oligocene (E/O) boundary. The opening of the southern ocean seaways and the decrease in atmospheric CO2 are two processes generally evoked to explain this E/O cooling. The debate is still ongoing but modeling studies tend to demonstrate that the decrease in atmospheric CO2 is the main driver of the cooling. However, uncertainties persist on what drove the decrease in atmospheric CO2 during the Cenozoic. In this study, we investigate the impact of continental drift, lithology distribution and volcanic degassing rates on the atmospheric carbon dioxide concentration over the Cenozoic within a coupled climate-carbon model (GEOCLIM). In the model, the continental drift results in driving low atmospheric CO2 levels during the Eocene and the Oligocene. The dispersed configuration and the location of a large continental area (North Africa, northern South America) within the Inter Tropical Convergence Zone (ITCZ) promote CO2 consumption by weathering, forcing CO2 to remain low. Icehouse conditions are also promoted by the drifting of India and the weathering of the Deccan basalts in the ITCZ during the Eocene, and by the weathering of the Ethiopian traps during the Oligocene. To prevent the building up of the Antarctic ice sheet at the Eocene, the model needs enhanced solid Earth degassing flux by 50% so that atmospheric CO2 levels stay above the glacial threshold (750 ppm). We find that the decrease in atmospheric CO2 from the Eocene to the Oligocene is probably due to a reduction in the source of volcanic CO2 rather than an increase in silicate weathering. The model results furthermore suggest that during the Miocene period, the northward drifting of both the African plate and India (including the Deccan traps) might have decreased the continental surface exposed to chemical weathering, therefore generating higher CO2 values. Finally, the uplift of the Tibetan plateau from the Miocene to the present-day induces in the model an increase in silicate weathering through the intensification of the South-Eastern Asian monsoon, causing atmospheric CO2 to come back to a pre-industrial value. & 2013 Elsevier B.V. All rights reserved.
Keywords: Cenozoic cooling Antarctic glaciations carbon cycle modeling Cenozoic CO2 silicate weathering
1. Introduction It is well documented that the Cenozoic is a period of transition from the warm Cretaceous climate toward the cooler present-day (Zachos et al., 2001). The two most salient features of the Cenozoic are the growth of the ice sheet over Antarctica at the Eocene– Oligocene boundary, and the onset of the Northern hemisphere glaciation about 5 million yr ago. For both of these events, a threshold of atmospheric CO2 to initiate glaciation has been estimated to be 750 ppm for Antarctica (DeConto and Pollard, 2003), and about n Corresponding author at: Laboratoire des Sciences du Climat et de l’Environnement, CNRS-CEA, CEA Saclay, Orme des Merisiers, F-91190 Gif sur Yvette, France. Tel.: +33 169 088 798. E-mail address: [email protected] (V. Lefebvre).
0012-821X/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.epsl.2013.03.049
280 ppm, a much lower level, for the Northern hemisphere (DeConto et al., 2008). Any reconstructions of the long-term CO2 evolution during the Cenozoic must account for these two threshold values and respective timing. Until now, the most commonly invoked mechanism for the Cenozoic CO2 drawdown has been the reduction of CO2 by weathering of silicate rocks promoted by the Himalayan uplift (Raymo and Ruddiman, 1992). However the weathering of silicate rocks exposed in the Himalayan orogen is alone not large enough to account for the observed drawdown in atmospheric CO2 (France-Lanord and Derry, 1997; Galy and France-Lanord, 1999). The associated burial of organic carbon in the Bengal fan might have helped to contribute in drawing CO2 down (Galy et al., 2007). This organic carbon sink is quite efficient today and throughout the Neogene, (Derry and FranceLanord, 1996). Other orogens could
204
V. Lefebvre et al. / Earth and Planetary Science Letters 371-372 (2013) 203–211
have been additional important contributors to the Cenozoic CO2 drawdown. Moquet et al. (2011) showed that silicate weathering in the Andes substantially contributes to the CO2 consumption today. Furthermore, weathering may take place in floodplains at the foot of active orogens (Bouchez et al., 2012). Despite years of debate, we have to acknowledge that, up to now, no quantitative model describing the integrated effects of mountain uplift on the carbon cycle evolution has been proposed. The reason for this is rooted in the complexity of the processes at play. In particular, a process-based description of the time evolution of the mechanical erosion is still lacking. Another mechanism for the Cenozoic cooling is the general decrease in the solid Earth degassing rates, as shown by Gaffin (1987) and Engebretson et al. (1992), who suggested that ocean crust production has halved since the Cretaceous. On the contrary, Larson (1991) and Rowley (2002) suggested that the overall degassing rate showed little variation over the last 100 Ma. According to these reconstructions, the past solid Earth degassing rate is still largely unconstrained, despite its central role in the evolution of global climate during the Cenozoic A third mechanism for the atmospheric CO2 evolution is better constrained over the Cenozoic and refers to the continental drift. When moving either in latitude or in longitude, the continents go across various climatic belts. These displacements strongly influence the chemical weathering of rocks. Marshall et al. (1988) were the first to propose the existence of a causal link between the continental configuration, the silicate rock weathering and the atmospheric CO2 level, using a simple zonally averaged model of the land distribution. Bluth and Kump (1991) and Gibbs et al. (1999) showed with a GCM, and a realistic continental drift history, that the end Mesozoic–early Cenozoic periods have been characterized by an efficient silicate weathering promoted by intense rainfall. More recently, the continental drift effect on atmospheric CO2 and global climate has been explored with the GEOCLIM model for the whole Phanerozoic (Donnadieu et al., 2006; Goddéris et al., 2008, 2012; Nardin et al., 2011). Superimposed on the continental drift, several large igneous provinces (LIPs) erupted during the Cenozoic or slightly before (Deccan trapps, Ethiopian trapps; Courtillot and Renne, 2003). Together with older basaltic provinces, and because they weather much faster than continental shield rocks (Dessert et al., 2003), LIPs constitute potential hot spots for CO2 consumption when carried by continental drift into warm and humid areas (Kent and Muttoni, 2008; Lefebvre et al., 2010). In this contribution, we use the coupled carbon-climate model GEOCLIM to quantify the impact of the Cenozoic drift of the continents and the LIPs on the time evolution of the atmospheric CO2.
2. Model description 2.1. General framework of the GEOCLIM model We use the most recent version of the GEOCLIM model described by Donnadieu et al. (2009). The model couples the 3-D climate model FOAM, the dynamic global vegetation model LPJ (Sitch et al., 2003) and the box model for the geological carbon and alkalinity global cycle COMBINE (Goddéris and Joachimski, 2004). The atmospheric component of FOAM is a parallelized version of NCAR's Community Climate Model 2 (CCM2) with the upgraded radiative and hydrologic physics incorporated in CCM3 v. 3.2. The atmosphere runs at R15 spectral resolution (4.51 7.51) with 18 levels. For this study, we use FOAM in mixed-layer mode where the atmosphere is coupled to a 50-m mixed-layer ocean, which parameterizes heat transport through diffusion, in order to save
computation time (one GEOCLIM simulation requires up to 12 GCM simulations). The vegetation model LPJ is a Dynamic Global Vegetation Model of vegetation biogeography and vegetation/soil biogeochemistry. Starting from the climate, soil and atmospheric information as input, it dynamically computes the transient vegetation composition in terms of plant functional groups and their associated carbon and water budgets. The calculated vegetation distribution directly impacts on (1) the albedo of each grid cell, (2) its thermal properties, (3) its roughness, and (4) its potential evaporation. In this contribution, we aim to calculate the steady-state pCO2 associated with five continental configurations spanning the whole Cenozoic. On a geological timescale, the global carbon and alkalinity cycles are balanced: F vol ¼ F silw þ ðF od −F ow Þ
ð1Þ
where Fvol is the total solid Earth CO2 degassing flux and Fsilw is the total CO2 consumed by continental silicate rocks (Walker et al., 1981). Fod is the burial of organic carbon and Fow the oxidation of sedimentary reduced carbon exposed at the continental surfaces. The difference, Fod−Fow, fluctuated over the Cenozoic (Goddéris and François, 1996). However, because in the present study we focus on estimating the role of continental drift and its impact on silicate weathering on the Cenozoic climate evolution, we deliberately set the Fod−Fow term to 0. This assumes steady state of the organic carbon sub-cycle. Under this assumption, Eq. (1) simplifies as follows: F vol ¼ F silw
ð2Þ
Fsilw depends on climate (runoff and air temperature), amongst others, and climate depends on atmospheric CO2 pressure. Accounting for these dependencies, Eq. (1) is readily solved for the steadystate atmospheric CO2. The advantage of GEOCLIM when compared to previously published models such as GEOCARB (Berner, 2004) is to resolve the continental climate spatially, so that climatic parameters such as runoff, air temperature and consumption by continental silicate weathering, are calculated at each grid cell. We can rewrite Eq. (1) including the impact of the continental configuration: ncont
F vol ¼ ∑ F jsilw ðT j ; runof f j Þ j¼1
ð3Þ
where ncont is the number of continental grid elements, Tj and runoffj the air temperature and runoff calculated by the GCM for the grid element j respectively. FOAM is run for each paleogeographical configuration at various CO2 concentrations (from 160 to 2400 ppm) accounting for the secular trend in the solar constant. Each simulation is run for 30 yr (until steady state is reached). Assuming that the modeled climate system behaves quasi-linearly, we can estimate the climatic variables (temperature and runoff) for any pCO2 and continental configuration to calculate the F jsilw term on each continental grid cell. Then, if Fvol is known, we can estimate the atmospheric CO2 level necessary for equilibrating Fvol and Fsilw (Eq. (2)). 2.2. Model representation of the lithology The geochemical model differentiates weathering of basaltic and granitic lithologies. For the granitic lithologies, total CO2 consumption through weathering is based on the work by Oliva et al. (2003), who compiled data for about 100 small granitic catchments (Eq. (4)). They showed that the cation flux released by granitic weathering is proportional to the runoff and is a function of temperature, assuming an apparent activation energy of 48.7 kJ/mol. The CO2 consumption is proportional to the cation
V. Lefebvre et al. / Earth and Planetary Science Letters 371-372 (2013) 203–211
flux, because in freshwater the cationic charge is mostly balanced by HCO3− anions. CO2 consumption is then proportional to the runoff and is a function of temperature with the same activation energy as the total cation flux. graw npixel E 1 1 F graw ¼ ∑ kgraw ⋅runof f j ⋅areaj ⋅exp a − ð4Þ Tj T0 R j¼1 where Fgraw is the total atmospheric CO2 consumption via granitic weathering, R is the perfect gas constant, Eagraw is the apparent activation energy of 48.7 kJ/mol, areaj is the surface of the continental grid cells. Tj and runoffj are the mean annual ground air temperature and the mean annual runoff at the grid cell j respectively. Weathering of basaltic lithologies is calculated according to Dessert et al. (2001): " # npixel Ebasw 1 1 a F basw ¼ ∑ kbasw ⋅runof f j ⋅areaj ⋅exp ð5Þ − Tj T0 R j¼1 where Fbasw is the total atmospheric CO2 consumption flux through basaltic weathering. Ebasw is fixed at 42.3 kJ/mol. kbasw and kgraw are a calibrated on present day climatic conditions provided by a FOAM simulation in a mixed-layer configuration. By assuming a total CO2 consumption by silicate weathering (Fgraw+Fbasw) equal to 13.6 1012 mol/yr (Dessert et al., 2003; Gaillardet et al., 1999) with a 30% contribution of basaltic lithologies and by using the present day climatic fields from FOAM, kbasw (6.25 109 with runoff in cm/ yr and area in 106 km2) and kgraw (1.23 109 with runoff in cm/yr and area in 106 km2) are determined.
we calculate the steady state CO2 for each time period assuming a constant degassing rate (6.8 1012 mol/yr). A uniform distribution of the basaltic rocks at each terrestrial grid cell is prescribed and calibrated in order for the basalt weathering to reach a 30% contribution of the total silicate weathering flux taken at present day (Dessert et al., 2003). These simulations aim to show the effect of the continental drift. In the second set of simulations (REG simulation), the consumption of CO2 by continental silicate weathering is calculated based on the spatial distribution of the basaltic provinces. Fig. 1 shows prescribed distribution of basaltic provinces plotted in the atmospheric resolution of the model FOAM. The size of each basaltic province is listed in Table 1. In the REG simulation, solid Earth degassing rate is fixed to present day value. In the third set of simulations (VOL), igneous provinces are explicitly accounted for as in the REG simulation, but we now consider a time-varying degassing rate throughout the Cenozoic. This is deduced from the reconstruction of the subduction rate provided by Engebretson et al. (1992) (see discussion below). These simulations are summarized in Table 2.
3. Results 3.1. Role of the continental drift The most striking result comes from the UNI simulations where atmospheric CO2 stays relatively constant oscillating around 340 ppmv over the whole Cenozoic (Fig. 2; Table 3). This is not
2.3. Model configurations We use five geographic reconstructions of the Cenozoic including Early Paleocene (65 Ma), Early Eocene (52 Ma), the middle of the Oligocene (30 Ma), Middle Miocene (15 Ma) and Present-Day (PD). The Miocene geography comes from Herold et al. (2008). The land– ocean configurations for 65 Ma, 52 Ma and 30 Ma are built from a synthesis of paleomagnetic data, hot spot tracks and geologic constraints (Besse and Courtillot, 2002; Dercourt et al., 1993). Earth's orbital parameters are set to present-day values for the five configurations. Solar radiation is assumed to evolve through time according to the stellar evolution models (from 0.65% reduction in the Paleocene to 0.15% in the Miocene, Gough, 1981). We use the estimated CO2 thresholds below which ice is allowed to accumulate on Antarctica and Greenland at 750 ppm (DeConto and Pollard, 2003) and 280 ppm (DeConto et al., 2008; Lunt et al., 2008) respectively. Based on these estimations, and because the FOAM GCM does not include an ice sheet model allowing the onset of an ice sheet, (1) we assume the absence of ice sheets in all geographies when CO2 is above 750 ppm, (2) we prescribe the existence of both ice sheets (North and South) for CO2 levels lower or equal to 280 ppm, and (3) we prescribe the Antarctic ice sheet for CO2 levels between 280 and 750 ppm. The ice sheet relief and albedo values for Antarctica and Greenland are taken from their modern states. 2.4. Setup of the atmospheric CO2 simulations Over the Cenozoic, mechanical erosion appears to increase worldwide in response to the erosion of the major Cenozoic orogens (Kump and Arthur, 1997; Wold and Hay, 1990). In order to account for a fluctuating role of physical erosion in our simulations, potentially promoting silicate weathering (West et al., 2005), we simply multiply our granitic silicate weathering rate by a fR factor which varies from 0.75 at 65 Ma to 1 at 15 Ma (method and values are from Berner (2004)). We then run three sets of simulations. In the first one (UNI),
205
Col.
Is.
Sib. Kam.
Tim. Dec.
Cent. Am. present at :
Par.
Emei.
Eth.
Indo.
Kar.
65, 52, 30, 15, PD Pat. 52, 30, 15, PD 30, 15, PD 15, PD PD
Tas.
Fig. 1. Present-day localization of the main basaltic provinces and their age of formation. Comparing the area of an igneous province with the area of the model grid cell, we define the fraction of the cell covered by the igneous province (the rest being covered by shield rocks).
Table 1 Ages and areas of the main basaltic provinces considered in this study (Chen et al., 2008; Corgne et al., 2001; Courtillot and Renne, 2003; Dessert et al., 2003; Honthass et al., 1995; Kuznetsov et al., 2007; Meffre et al., 2004; Wignall, 2001). Basaltic provinces
Age (Ma)
Area (106 km2)
Kamchakta (Kam.) Patagonia (Pat.) Columbia River (Col.) Ethiopia (Eth.) Island/Greenland (Isl.) Deccan (Dec.) Central America (Cent. Am.) Indonesia/SE Asia (Indo.) Parana (Par.) Karoo (Kar.) Siberia (Sib.) Emeishan (Emei.) Timan/NW Russia (Tim.) Tasmania (Tas.)
5–12 10 16 30 60 65 90 90–250 133 184 250 259 550 580
0.17 0.21 0.2 0.8 0.1 0.5 0.31 0.54 0.57 0.22 1.5 0.5 0.18 0.33
Total
6.13
206
V. Lefebvre et al. / Earth and Planetary Science Letters 371-372 (2013) 203–211
Table 2 Description of the parameters used in the simulations. Simulation
Basalt repartition
Mechanical erosion (65–15 Ma)
Degassing rate
Aims
UNI REG
Uniform Explicit with the modern geographical distribution (Fig. 1) Explicit with the modern geographical distribution (Fig. 1)
0.75–1 0.75–1
Constant (6.8 1012 mol/yr) Constant (6.8 1012 mol/yr)
Continental drift Continental drift+basalt
0.75–1
Deduced from Engebretson et al. (1992) (Fig. S1)
Degassing rate
VOL
because the carbon cycle is insensitive to the change in continental configuration, but rather the result of efficient weathering during the Cenozoic damping CO2 down to low levels. Indeed, using the same model, Donnadieu et al. (2009) found higher CO2 levels for pre-Cenozoic continental configurations. In addition, subtle changes in the paleogeography of our model such as the northward drift of Pangea between the Carnian and the Rhaetian time periods (i.e. 220 and 200 Ma) can trigger an atmospheric CO2 fluctuation of 2000 ppm (Goddéris et al., 2008). To illustrate the effect of paleogeography in our model, we compare global weathering rates from climatic simulations with identical atmospheric CO2 level (280 ppmv, i.e. the pre-industrial value) (Fig. 3). In these series of simulations, the basaltic weathering rates are substantially larger in the 65, 52 and 30 Ma configurations and lower in the 15 Ma configuration than the modern value (Fig. 3b). For the 65, 52 and 30 Ma paleogeographies, the trend in granitic weathering rates is slightly different because of the erosion corrective factor that implicitly draws down the weathering rates of the granitic surfaces by 25% (Fig. 3a). As a result, the global weathering rate, the sum contribution of basaltic and granitic lithologies, stays below (but close to) the present-day level (Fig. 3c). With lower global weathering rates, atmospheric CO2 levels stay above the pre-industrial value in the model (Fig. 2). In this set of experiments, we did not aim to quantify the effect of the physical erosion of active orogens on the global carbon cycle. However our results suggest that the weathering feedback remains quite efficient during the early Cenozoic even in the absence of intense physical erosion. Modeled CO2 levels for the 65 Ma and 15 Ma time periods are well within the range of estimates, while modeled CO2 levels of the Oligocene and Eocene are at the lower end estimates (Fig. 2). However, all CO2 levels from the UNI simulations are lower than the threshold for initiating an extensive glaciation in Antarctica. In these simulations, the transition to an icehouse world would be triggered from the beginning of the Cenozoic. Because the main mechanism in the UNI simulation is the continental drift forcing the global climate to be cold well before the effective start of the glaciation, we conclude that another mechanism is required to explain the evolution of the long-term CO2 trend (and climatic trend) throughout the Cenozoic.
and 65 Ma-old Deccan traps are still located in subtropical regions where runoff is low, inducing smaller global silicate weathering rates (4.87 1012 mol of CO2/yr for the Paleocene vs. 6.8 1012 mol of CO2/yr for the modern, see Figs. 3 and 4). The lower basaltic weathering rate at the Paleocene is compensated in the model by higher atmospheric CO2 concentration in order to keep global silicate weathering and total solid Earth degassing in balance (Eq. (1)). It is only during the Eocene that the Deccan traps finally leave the subtropical latitudes to enter the ITCZ increasing basaltic and silicate weathering (Fig. 4). An additional contribution at the Oligocene to the global silicate weathering comes from the weathering of the Ethiopian traps below the ITCZ (Figs. 3 and 4). During the Miocene, the northward drift of Africa and India reduced the basaltic weathering over the Ethiopian traps and the Deccan traps respectively, resulting in higher CO2 concentration while reaching steady state in the model (Figs. 3 and 4). Finally, from the MidMiocene to the present day, the silicate weathering flux increases in response to the strengthening of precipitation and runoff in southeastern China and Indonesia due to the Tibetan plateau uplift and subsequent intensification of the South East Asian monsoon (Fig. 4). Observations show a more intense monsoon in Asia during the last 8–9 Ma (Zhisheng et al., 2001). The displacement of the Emeishan traps and the Indonesian basaltic archipelago in the vicinity of the ITCZ emphasizes this pattern of increased weathering over the last 15 Ma. The modeled evolution in the weathering of the main basaltic provinces, calculated for each time slice of the REG simulation, confirms this scenario (Fig. 5). The model shows that weathering specifically of the Deccan traps reaches a maximum at 52 Ma, followed by a decrease during the Oligocene to reach relatively low rates from the Miocene to today. The Ethiopian traps are rapidly weathered during the Oligocene, directly after their onset, and consume less CO2 during the Miocene and the presentday. Finally, our model simulations show that CO2 concentration levels kept decreasing in the last 15 Ma because of a strengthening of the East Asian monsoon that resulted in stronger weathering of Emeishan and Indonesia basaltic provinces. The explicit calculation of the effect of the basaltic provinces leads to a general increase in CO2 from the UNI to the REG simulations. Nevertheless, atmospheric CO2 remains too low, i.e. below the Antarctic glaciation threshold, during the Paleocene (680 ppm) and the Eocene (360 ppm).
3.2. Role of the weathering of the LIPs 3.3. Role of the degassing rate The REG simulation (where basaltic provinces are spatially distributed) shows larger fluctuations in the atmospheric CO2 levels than the UNI simulation. In REG, atmospheric CO2 concentration stays relatively high during the Paleocene (680 ppm) and the Miocene (500 ppm), and drops to about 340 ppm at the Eocene and Oligocene (Fig. 2; Table 3). To separate the different causes in the REG CO2 trend, we calculate the silicate weathering flux for each reconstruction set with 280 ppm of CO2 (Figs. 3 and 4). The model results show that the surface of basaltic provinces experiencing a warmer and wetter climate is smaller during the Paleocene than during the rest of the Cenozoic. This is because the Ethiopian traps have not been emplaced at that time, and the 258 Ma-old Emeishan
The evolution of the global degassing rate throughout the Cenozoic, and more generally throughout the Earth history, remains poorly constrained. There are (1) no direct measurements of the Cenozoic fluctuations of the solid Earth degassing and (2) no clear consensus between the various reconstructions of the degassing flux through the geological past. For instance, recent re-evaluations suggest that because the ridge production is estimated to have stayed roughly constant for the last 180 Ma, the degassing rate should have had little fluctuations since the middle Jurassic (Rowley, 2002). However, as already noticed by Kerrick and Caldeira (1999), processes other than ridge production can influence
V. Lefebvre et al. / Earth and Planetary Science Letters 371-372 (2013) 203–211 5000 4000
UNI REG VOL
3000
Antarctic ice sheet N. Hemisphere ice sheet
2000
pCO2 (ppmv)
207
1000
Antarctic ice sheet threshold
500 400 300 N.H. threshold
Ma.
Paleo.
Eocene
Oligo.
Pli.
Cenozoic
Meso.
70
Miocene
Ple.
200
60
50
40
30
20
10
0
Time (Myr) Fig. 2. Atmospheric CO2 levels simulated by GEOCLIM. Black squares are the pCO2 data compiled by Beerling and Royer (2011). Occurrence of Antarctic and Northern Hemisphere ice sheets are taken from Zachos et al. (2001). Because our knowledge of the paleogeography is tied to the time resolution of the paleomagnetic data, we assume that the steady-state CO2 levels are valid within a 5 Ma time window and give the multi-million years baseline of the CO2 evolution, missing potential rapid shifts.
the Earth degassing rate. During the Paleocene and the Eocene, Kerrick and Caldeira (1999) suggested that the Earth degassing rate may have increased in response to regional metamorphism of the Tethysian orogeny, following the subduction of carbonates accumulated on the seafloor of the European and Asian margins. Using this information along with Engebretson et al.'s (1992) work on subduction history, we estimate that the global degassing rate was higher between 130 and 50 Myr than today (Fig. A1). We investigate here the potential impact of this degassing rate history on modeled atmospheric CO2. Increasing the degassing rate by 50% results in our model an increase of atmospheric CO2 to about 1310 ppmv at the Paleocene and 930 ppmv at the Eocene (Fig. 2; Table 3). These new CO2 levels are then well above the Antarctic glaciation threshold (750 ppm) and consequently more consistent with the Cenozoic climate history. Including the timing estimates of degassing rate by Engebretson et al. (1992) did not change our steady state CO2 levels at the Oligocene and Miocene calculated in the REG simulation. Details of the continental temperature, runoff and precipitation minus evaporation can be found in the Appendix (Figs. A2–A8).
4. Discussions The long-term CO2 calculated in the VOL simulation is in good agreement with the evolution of the climate during the Cenozoic. The intensity of weathering during the Paleocene and Eocene requires an enhanced CO2 source (Engebretson et al., 1992) to allow a greenhouse climate that prevents the Antarctic glaciation at that time. However, in our modeling study, we assumed that the organic carbon sub-cycle (i.e. the balance between the organic carbon burial and the sedimentary organic carbon oxidation over continents) is at steady state. Instead of a high solid Earth degassing rate, a low organic carbon burial (or strong kerogen oxidation) could alternatively lead to higher CO2 levels during the Early Cenozoic. This is a possibility based on the carbon isotope budget, which suggests that organic carbon burial has increased over the course of the Cenozoic (Derry and FranceLanord, 1996). We use the compilation of δ18O provided by Prokoph et al. (2008) to calculate the surface seawater temperature in midlatitudes (30–601N and S) using the paleothermometer equation of Anderson and Arthur (1983): T ¼ 16−4:14ðδ18 O−δ18 Osw Þ þ 0:13ðδ18 O−δ18 Osw Þ2
ð6Þ
Table 3 Steady-state atmospheric CO2 level calculated for each simulation. Period
Paleocene Eocene Oligocene Miocene PD
Degassing rate
Calculated steady-state pCO2 UNI (ppm)
REG
VOL
VOL (ratio)
340 320 350 330 280
680 360 320 500 280
1315 930 360 475 280
1.35 1.46 1.05 0.98 1
where T is the seawater temperature, δ18O is the oxygen isotope value of the sedimentary carbonate data and δ18Osw is the δ18O of the seawater (fixed to 0‰ from 30 Ma to modern, and −1‰ for 65 and 52 Ma). The temperature in our VOL simulations does not have exact same trend in temperature as the 10 Myr-averaged temperature depicted by the oxygen isotopes (Fig. 6). While there is good agreement for the 52 Ma, 30 Ma and 15 Ma time periods, our VOL simulation tends to overestimate the sea surface temperature for the 65 Ma time period. Conversely, the temperature from the UNI simulation fits the oxygen isotope estimates better at the 65 Ma periods, modeled with values of atmospheric CO2 level of 340 ppm. We note here that this estimate corresponds to the atmospheric CO2 level estimated by Beerling and Royer (around 400 ppm on their curve). However, this is below the threshold value for glaciation of Antarctica and thus too low for the 65 Ma time period. Nevertheless, the values of the glaciation thresholds are based on modern and Oligocene continental configurations. At the beginning of the Paleocene however, Australia was still attached to Antarctica. This configuration may be in favor of a lower atmospheric CO2 threshold. With a larger continental mass centered around the pole, warmer summer temperature should induce lower snow accumulation, reducing the atmospheric CO2 threshold for southern continental glaciation. This hypothesis should be tested in a coupled ice-climate model in the future. The Eocene is a warm period in the model as the result of the intense degassing rate that can largely overcompensate for the elevated CO2 consumption linked to the northward drift of the Deccan basalts. The decrease in degassing rate from the Eocene to the Oligocene causes atmospheric CO2 to cross the threshold of Antarctic glaciation, which might have happened between 52 and
V. Lefebvre et al. / Earth and Planetary Science Letters 371-372 (2013) 203–211
Ma.
Paleo.
Eocene
Meso.
Oligo.
Miocene
Pli.
Ple.
CO2 consumption rate (1012 mol/yr)
208
Cenozoic Time (Myr)
Fig. 3. Evolution of the (a) granitic, of the (b) basaltic and of the (c) global weathering rates (in mol CO2/yr) at 280 ppm throughout the Cenozoic. Circles (squares) stand for conditions defined in the UNI (REG) simulation, see text for more details.
30 Ma. To narrow down when the glaciation of Antarctica started within this time window requires additional continental configurations. The model suggests that the mid-latitude averaged temperature at the Oligocene is close to the pre-industrial value (14 1C) and consistent with the presence of an ice sheet on Antarctica. During the Miocene, the modeled atmospheric CO2 and temperature increases in response to a decrease of continental weathering. The northward drift of India and Africa tends to reduce the areas of granitic and basaltic lithologies below the ITCZ during that time period. The resulting steady state CO2 level is 475 ppm, which is consistent with the 400–500 ppm estimation of Kürschner et al. (2008) at the beginning of the Middle Miocene (15.5 Ma). Our result is also consistent with the modeling work by Hamon et al. (2012) but disagrees with estimates from marine phytoplankton carbon isotopes which estimate a pCO2 as low as 280 ppmv (Pagani et al., 2005, 2011). However, a warmer Miocene is in agreement with the record of the Mid-Miocene Climate Optimum (MMCO) (Kürschner et al., 2008; Mosbrugger et al., 2005; Zachos et al., 2001) which could be related at least in our model to a decrease in the CO2 consumption rate by weathering.
5. Conclusions We used a spatially resolved climate-carbon model (GEOCLIM) and investigated the influence of continental drift on the long-
term evolution of the Cenozoic atmospheric CO2. Our model shows that the continental configuration should set the required conditions for the onset of a large-scale glaciation at the beginning of the Cenozoic. This is true even when we imposed in the model a decrease of the weathering owing to the absence of major uplift during the early Cenozoic. The dispersed configuration and the location of a large continental area (North Africa, northern South America) within the ITCZ promoted the CO2 consumption by weathering, forcing CO2 to remain low. Using a 2D representation of the basaltic provinces in the model, we found that Paleocene had higher atmospheric CO2 because of the location of the Deccan traps in the subtropics, which triggered lower CO2 consumption by weathering. The weathering effect of this northward drift of the Deccan traps at the Eocene is so high that atmospheric CO2 is too low in triggering the buildup of the Antarctic ice sheet before the observed pattern. This result is in disagreement with the scenario suggested by Kent and Muttoni (2008). Here we suggest that Antarctic ice sheet did not develop before the Oligocene because of a high degassing rate at the Eocene. In other words, our model results suggest that the transition from a greenhouse Eocene to an icehouse Oligocene is mainly the result of a reduction in CO2 degassing rate. Finally during the Miocene period, our model results show that the northward drift of Africa and India causes CO2 removal to decrease with the weathering, followed by an increase in the atmospheric CO2 level to up to 500 ppm. The modeling of the
V. Lefebvre et al. / Earth and Planetary Science Letters 371-372 (2013) 203–211
209
PD
SST (°C)
Silicate weathering rate (1010 mol/yr)
15Ma
52Ma
30Ma
65Ma
Fig. 4. Evolution of the annual silicate weathering rate in mol CO2/yr and of the annual averaged air temperature above the ocean for the paleogeographical reconstructions used in this study at 280 ppm CO2. Conditions used to calculate the weathering rates are those defined for the REG simulation.
Ma.
Paleo.
Eocene
Meso .
Oligo.
Miocene
Pli.
Ple.
CO2 consumption rate (1012 mol/yr)
Ethiopia Deccan Emeishan Indonesia
Cenozoic Time (Myr)
Fig. 5. CO2 consumption by silicate weathering history of the four main basaltic provinces during the Cenozoic as calculated in the REG simulation.
effect of the continental drift on atmospheric CO2 favors the presence of a Middle Miocene Climate Optimum. Finally, we found that the decrease of the atmospheric CO2 level from 500 ppm
during the Miocene to its pre-industrial values is induced by an increase in silicate weathering, due to a more intense South-East Asian monsoon.
210
V. Lefebvre et al. / Earth and Planetary Science Letters 371-372 (2013) 203–211
25 India efficient weathering Antarctic ice sheet N. Hemisphere ice sheet
Temperature (°C)
20
15
M a. 5
Paleo.
Eocene
Oligo.
Miocene
P li.
Ple.
10
Cenozoic
Meso.
60
50
40
30
20
10
0
Time (Myr) Fig. 6. Evolution of the mid-latitude (32–581N and S) sea surface temperature calculated from a carbonate δ18O compilation (Prokoph et al., 2008). Gray squares are the 10 Myr averaged temperature of this dataset; the vertical bars stand for the standard deviation. The blue triangles are the 30–601S and N sea surface averaged temperature calculated by FOAM accounting for a variable solid Earth degassing (VOL simulation). The vertical bars represent the maximum and minimum calculated temperatures occurring in these areas. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Acknowledgments This work was supported by the ANR project COLORS (No. ANR09-JCJC-0105) and by the ANR project CALIMERO (No. ANR-09BLAN-0221). We thank two anonymous reviewers for their constructive comments. We thank F. Monteiro, N. MacBean and S. Taylor for improving the readability of this contribution.
Appendix A. Supplementary matrials Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.epsl.2013.03.049.
References Anderson, T.F., Arthur, M.A., 1983. Stable isotopes of oxygen and carbon and their application to sedimentological and paleoenvironnemental problems. In: Arthur, M.A., Anderson, T.F., Kaplan, I.R., Veizer, J., Land, L. (Eds.), Stable Isotopes in Sedimentary Geology. Society of Economic Paleontologists and Mineralogists Tulsa, Oklahoma, USA, pp. 1–151. Beerling, D.J., Royer, D.L., 2011. Convergente Cenozoic CO2 history. Nature Geoscience 4, 418–420. Berner, R.A., 2004. A model for calcium, magnesium and sulfate in seawater over phanerozoic time. Am. J. Sci. 304, 438–453. Besse, J., Courtillot, V., 2002. Apparent and true polar wander and the geometry of the geomagnetic field over the last 200 Myr. J. Geophys. Res. 107, 2300. Bouchez, J., Gaillardet, J., Lupker, M., Louvat, P., France-Lanord, C., Maurice, L., Armijos, E., Moquet, J.-S., 2012. Floodplains of large rivers: weathering reactors or simple silos? Chem. Geol. 332–333, 166–184. Bluth, G.J.S., Kump, L.R., 1991. Phanerozoic paleogeology. Am. J. Sci. 291, 284–308. Chen, C.H., Lee, C.Y., Shinjo, R.I., 2008. Was there Jurassic paleo-Pacific subduction in South China?: constraints from 40Ar/39Ar dating, elemental and Sr–Nd–Pb isotopic geochemistry of the Mesozoic basalts. Lithos 106, 83–92. Corgne, A., Maury, R.C., Lagabrielle, Y., Bourgois, J., Suarez, M., Cotten, J., Bellon, H., 2001. La diversité des basaltes de Patagonie à la latitude du point triple du Chili (461–471lat. S) : données complémentaires et implications sur les conditions de la subduction. C. R. Acad. Sci.—Ser. IIA—Earth Planet. Sci. 333, 363–371. Courtillot, V.E., Renne, P.R., 2003. On the ages of flood basalt events. C. R. Geosci. 335, 113–140. DeConto, R.M., Pollard, D., 2003. Rapid Cenozoic glaciation of Antarctica induced by declining atmospheric CO2. Nature 421, 245–249. DeConto, R.M., Pollard, D., Wilson, P.A., Palike, H., Lear, C.H., Pagani, M., 2008. Thresholds for Cenozoic bipolar glaciation. Nature 455, 652.
Dercourt, J., Ricou, L.E., Vrielynck, B., 1993. Atlas Tethys Palaeoenvironments Maps. Gauthier-Villars, Paris. Derry, L.A., FranceLanord, C., 1996. Neogene growth of the sedimentary organic carbon reservoir. Paleoceanography 11, 267–275. Dessert, C., Dupré, B., François, L.M., Schott, J., Gaillardet, J., Chakrapani, G., Bajpai, S., 2001. Erosion of Deccan Traps determined by river geochemistry: impact on the global climate and the 87Sr/86Sr ratio of seawater. Earth Planet. Sci. Lett. 188, 459–474. Dessert, C., Dupré, B., Gaillardet, J., Francois, L.M., Allègre, C.J., 2003. Basalt weathering laws and the impact of basalt weathering on the global carbon cycle. Chem. Geol. 202, 257–273. Donnadieu, Y., Godderis, Y., Bouttes, N., 2009. Exploring the climatic impact of the continental vegetation on the Mesozoic atmospheric CO2 and climate history. Clim. Past 5, 85–96. Donnadieu, Y., Godderis, Y., Pierrehumbert, R., Dromart, G., Fluteau, F., Jacob, R., 2006. A GEOCLIM simulation of climatic and biogeochemical consequences of Pangea breakup. Geochem. Geophys. Geosyst., 7, http://dx.doi.org/10.1029/ 2006GC001278. Engebretson, D.C., Kelley, K.P., Cashman, H.J., Richards, M.A., 1992. 180 million years of subduction. GSA Today 2, 94–100. France-Lanord, C., Derry, L.A., 1997. Organic carbon burial forcing of the carbon cycle from Himalayan erosion. Nature 390, 65–67. Gaffin, S., 1987. Ridge volume dependence on sea-floor generation rate and inversion using long-term sealevel change. Am. J. Sci. 287, 596–611. Gaillardet, J., Dupre, B., Louvat, P., Allegre, C.J., 1999. Global silicate weathering and CO2 consumption rates deduced from the chemistry of large rivers. Chem. Geol. 159, 3–30. Galy, A., France-Lanord, C., 1999. Weathering processes in the Ganges–Brahmaputra basin and the riverine alkalinity budget. Chem. Geol. 159, 31–60. Galy, V., France-Lanord, C., Beyssac, O., Faure, P., Kudrass, H., Palhol, F., 2007. Efficient organic carbon burial in the Bengal fan sustained by the Himalayan erosional system. Nature 450, U406–U407. Gibbs, M.T., Bluth, G.J.S., Fawcett, P.J., Kump, L.R., 1999. Global chemical erosion over the last 250 my: variations due to changes in paleogeography, paleoclimate, and paleogeology. Am. J. Sci. 299, 611–651. Goddéris, Y., Donnadieu, Y., de Vargas, C., Pierrehumbert, R.T., Dromart, G., van de Schootbrugge, B., 2008. Causal or casual link between the rise of nanoplankton calcification and a tectonically-driven massive decrease in Late Triassic atmospheric CO2? Earth Planet. Sci. Lett. 267, 247–255. Goddéris, Y., Donnadieu, Y., Lefebvre, V., Le Hir, G., Nardin, E., 2012. Tectonic control of continental weathering, atmospheric CO2, and climate over Phanerozoic times. C. R. Geosci. 344, 652–662. Goddéris, Y., François, L.M., 1996. Balancing the Cenozoic carbon and alkalinity cycles: constraints from isotopic records. Geophys. Res. Lett. 23, 3743–3746. Goddéris, Y., Joachimski, M.M., 2004. Global change in the Late Devonian: modelling the Frasnian–Famennian short-term carbon isotope excursions. Palaeogeogr., Palaeoclimatol., Palaeoecol. 202, 309–329. Gough, D.O., 1981. Solar interior structure and luminosity variations. Sol. ar Phys. 74, 21–34.
V. Lefebvre et al. / Earth and Planetary Science Letters 371-372 (2013) 203–211
Hamon, N., Sepulchre, P., Donnadieu, Y., Henrot, A.-J., François, L., Jaeger, J.-J., Ramstein, G., 2012. Growth of subtropical forest in Miocene Europe: the roles of carbon dioxide and Antarctic ice volume. Geology 40, 567–570. Herold, N., Seton, M., Muller, R.D., You, Y., Huber, M., 2008. Middle Miocene tectonic boundary conditions for use in climate models. Geochem. Geophys. Geosyst, 9, http://dx.doi.org/10.1029/2008GC002046. Honthass, C., Bellon, H., Kepezhinskas, P.K., Maury, R.C., 1995. Nouvelles datations 40 K–40Ar du magmatisme Crétacé à Quaternaire du Kamchatka du Nord (Russie). Elsevier, Paris, France. Kent, D.V., Muttoni, G., 2008. Equatorial convergence of India and early Cenozoic climate trends. Proc. Natl. Acad. Sci. USA 105, 16065–16070. Kerrick, D.M., Caldeira, K., 1999. Was the Himalayan orogen a climatically significant coupled source and sink for atmospheric CO2 during the Cenozoic? Earth Planet. Sci. Lett. 173, 195–203. Kump, L.R., Arthur, M.A., 1997. Global chemical erosion during the Cenozoic: weatherability balances the budget. In: Ruddiman, W. (Ed.), Tectonic Uplift and Climate Change. Plenum Publishing Co., New York, pp. 399–426. Kürschner, W.M., Kvaček, Z., Dilcher, D.L., 2008. The impact of Miocene atmospheric carbon dioxide fluctuations on climate and the evolution of terrestrial ecosystems. Proc. Natl. Acad. Sci. USA 105, 449–453. Kuznetsov, N.B., Soboleva, A.A., Udoratina, O.V., Hertseva, M.V., Andrelchev, V., 2007. Pre-Ordovician tectonic evolution and volcano-plutonic associations of the Timanides and northern Pre-Uralides, northeast part of the East European Craton. Gondwana Res. 12, 305–323. Larson, R.L., 1991. Latest pulse of Earth—evidence for a Midcretaceous superplume. Geology 19, 547–550. Lefebvre, V., Servais, T., François, L., Averbuch, O., 2010. Did a Katian large igneous province trigger the Late Ordovician glaciation? A hypothesis tested with a carbon cycle model. Palaeogeogr. Palaeoclimatol.,Palaeoecol 296, 310–319. Lunt, D.J., Foster, G.L., Haywood, A.M., Stone, E.J., 2008. Late Pliocene Greenland glaciation controlled by a decline in atmospheric CO2 levels. Nature 454, 1102–1105. Marshall, H.G., Walker, J.C.G., Kuhn, W.R., 1988. Long-term climate change and the geochemical cycle of carbon. J. Geophys. Res.—Atmos. 93, 791–801. Meffre, S., Direen, N.G., Crawford, A.J., Kamenetsky, V., 2004. Mafic volcanic rocks on King Island, Tasmania: evidence for 579 Ma break-up in east Gondwana. Precambrian Res. 135, 177–191. Moquet, J.-S., Crave, A., Viers, J., Seyler, P., Armijos, E., Bourrel, L., Chavarri, E., Lagane, C., Laraque, A., Casimiro, W.S.L., Pombosa, R., Noriega, L., Vera, A., Guyot,
211
J.-L., 2011. Chemical weathering and atmospheric/soil CO2 uptake in the Andean and foreland Amazon basins. Chem. Geol. 287, 1–25. Mosbrugger, V., Utescher, T., Dilcher, D.L., 2005. Cenozoic continental climatic evolution of Central Europe. Proc. Natl. Acad. Sci. USA 102, 14964–14969. Nardin, E., Godderis, Y., Donnadieu, Y., Le Hir, G., Blakey, R.C., Puceat, E., Aretz, M., 2011. Modeling the early Paleozoic long-term climatic trend. Geol. Soc. Am. Bull. 123, 1181–1192. Oliva, P., Viers, J., Dupre, B., 2003. Chemical weathering in granitic environments. Chem. Geol. 202, 225–256. Pagani, M., Huber, M., Liu, Z.H., Bohaty, S.M., Henderiks, J., Sijp, W., Krishnan, S., DeConto, R.M., 2011. The role of carbon dioxide during the onset of Antarctic glaciation. Science 334, 1261–1264. Pagani, M., Zachos, J.C., Freeman, K.H., Tipple, B., Bohaty, S., 2005. Marked decline in atmospheric carbon dioxide concentrations during the Paleogene. Science 285, 876–879. Prokoph, A., Shields, G.A., Veizer, J., 2008. Compilation and time-series analysis of a marine carbonate δ18O, δ13C, 87Sr/86Sr and δ34S database through Earth history. Earth-Sci. Rev. 87, 113–133. Raymo, M.E., Ruddiman, W.F., 1992. Tectonic forcing of Late Cenozoic climate. Nature 359, 117–122. Rowley, D.B., 2002. Rate of plate creation and destruction: 180 Ma to present. Geol. Soc. Am. Bull. 114, 927–933. Sitch, S.N., Smith, B., Prentice, I.C., Arneth, A., Bondeau, A., Cramer, W., Kaplan, J.O., Levis, S., Lucht, W., Sykes, M.T., Thonicke, K., Venevsky, S., 2003. Evaluation of ecosystem dynamics, plant geography and terrestrial carbon cycling in the LPJ dynamic global vegetation model. Global Change Biol. 9, 161–185. Walker, J.C.G., Hays, P.B., Kasting, J.F., 1981. A negative feedback mechanism for the long-term stabilization of Earth’s surface-temperature. J. Geophys. Res.—Atmos. 86, 9776–9782. West, A.J., Galy, A., Bickle, M., 2005. Tectonic and climatic controls on silicate weathering. Earth Planet. Sci. Lett. 235, 211–228. Wignall, P.B., 2001. Large igneous provinces and mass extinction. Earth-Sci. Rev. 53, 1–33. Wold, C.N., Hay, W.W., 1990. Estimating ancient sediment. Am. J. Sci. 290, 1069–1089. Zachos, J.C., Pagani, M., Sloan, L., Thomas, E., Billups, K., 2001. Trends, rhythms, and aberrations in global climate 65 Ma to Present. Science 292, 686–693. Zhisheng, A., Kutzbach, J.E., Prell, W.L., Porter, S.C., 2001. Evolution of Asian monsoons and phased uplift of the Himalayan Tibetan plateau since Late Miocene times. Nature 411, 62–66.
Contents lists available at SciVerse ScienceDirect
Earth and Planetary Science Letters journal homepage: www.elsevier.com/locate/epsl
Was the Antarctic glaciation delayed by a high degassing rate during the early Cenozoic? Vincent Lefebvre a,b,n, Yannick Donnadieu a, Yves Goddéris b, Frédéric Fluteau c, Lucie Hubert-Théou d a
Laboratoire des Sciences du Climat et de l’Environnement, CNRS-CEA, CEA Saclay, Orme des Merisiers, F-91190 Gif sur Yvette, France Géosciences-Environnement Toulouse, Observatoire Midi-Pyrénées CNRS, 14 Avenue Edouard Belin, F-31400 Toulouse, France c Institut de Physique du Globe de Paris, UMR 7154, Sorbonne Paris Cité, 1 rue Jussieu, F-75238 Paris, France d Department of Earth and Planetary Sciences, McGill University, Montreal, Canada b
art ic l e i nf o
a b s t r a c t
Article history: Received 3 August 2012 Received in revised form 25 March 2013 Accepted 31 March 2013 Editor: J. Lynch-Stieglitz Available online 24 April 2013
The Cenozoic is a period of major climatic changes marked by the formation of the Antarctic ice sheet at the Eocene/Oligocene (E/O) boundary. The opening of the southern ocean seaways and the decrease in atmospheric CO2 are two processes generally evoked to explain this E/O cooling. The debate is still ongoing but modeling studies tend to demonstrate that the decrease in atmospheric CO2 is the main driver of the cooling. However, uncertainties persist on what drove the decrease in atmospheric CO2 during the Cenozoic. In this study, we investigate the impact of continental drift, lithology distribution and volcanic degassing rates on the atmospheric carbon dioxide concentration over the Cenozoic within a coupled climate-carbon model (GEOCLIM). In the model, the continental drift results in driving low atmospheric CO2 levels during the Eocene and the Oligocene. The dispersed configuration and the location of a large continental area (North Africa, northern South America) within the Inter Tropical Convergence Zone (ITCZ) promote CO2 consumption by weathering, forcing CO2 to remain low. Icehouse conditions are also promoted by the drifting of India and the weathering of the Deccan basalts in the ITCZ during the Eocene, and by the weathering of the Ethiopian traps during the Oligocene. To prevent the building up of the Antarctic ice sheet at the Eocene, the model needs enhanced solid Earth degassing flux by 50% so that atmospheric CO2 levels stay above the glacial threshold (750 ppm). We find that the decrease in atmospheric CO2 from the Eocene to the Oligocene is probably due to a reduction in the source of volcanic CO2 rather than an increase in silicate weathering. The model results furthermore suggest that during the Miocene period, the northward drifting of both the African plate and India (including the Deccan traps) might have decreased the continental surface exposed to chemical weathering, therefore generating higher CO2 values. Finally, the uplift of the Tibetan plateau from the Miocene to the present-day induces in the model an increase in silicate weathering through the intensification of the South-Eastern Asian monsoon, causing atmospheric CO2 to come back to a pre-industrial value. & 2013 Elsevier B.V. All rights reserved.
Keywords: Cenozoic cooling Antarctic glaciations carbon cycle modeling Cenozoic CO2 silicate weathering
1. Introduction It is well documented that the Cenozoic is a period of transition from the warm Cretaceous climate toward the cooler present-day (Zachos et al., 2001). The two most salient features of the Cenozoic are the growth of the ice sheet over Antarctica at the Eocene– Oligocene boundary, and the onset of the Northern hemisphere glaciation about 5 million yr ago. For both of these events, a threshold of atmospheric CO2 to initiate glaciation has been estimated to be 750 ppm for Antarctica (DeConto and Pollard, 2003), and about n Corresponding author at: Laboratoire des Sciences du Climat et de l’Environnement, CNRS-CEA, CEA Saclay, Orme des Merisiers, F-91190 Gif sur Yvette, France. Tel.: +33 169 088 798. E-mail address: [email protected] (V. Lefebvre).
0012-821X/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.epsl.2013.03.049
280 ppm, a much lower level, for the Northern hemisphere (DeConto et al., 2008). Any reconstructions of the long-term CO2 evolution during the Cenozoic must account for these two threshold values and respective timing. Until now, the most commonly invoked mechanism for the Cenozoic CO2 drawdown has been the reduction of CO2 by weathering of silicate rocks promoted by the Himalayan uplift (Raymo and Ruddiman, 1992). However the weathering of silicate rocks exposed in the Himalayan orogen is alone not large enough to account for the observed drawdown in atmospheric CO2 (France-Lanord and Derry, 1997; Galy and France-Lanord, 1999). The associated burial of organic carbon in the Bengal fan might have helped to contribute in drawing CO2 down (Galy et al., 2007). This organic carbon sink is quite efficient today and throughout the Neogene, (Derry and FranceLanord, 1996). Other orogens could
204
V. Lefebvre et al. / Earth and Planetary Science Letters 371-372 (2013) 203–211
have been additional important contributors to the Cenozoic CO2 drawdown. Moquet et al. (2011) showed that silicate weathering in the Andes substantially contributes to the CO2 consumption today. Furthermore, weathering may take place in floodplains at the foot of active orogens (Bouchez et al., 2012). Despite years of debate, we have to acknowledge that, up to now, no quantitative model describing the integrated effects of mountain uplift on the carbon cycle evolution has been proposed. The reason for this is rooted in the complexity of the processes at play. In particular, a process-based description of the time evolution of the mechanical erosion is still lacking. Another mechanism for the Cenozoic cooling is the general decrease in the solid Earth degassing rates, as shown by Gaffin (1987) and Engebretson et al. (1992), who suggested that ocean crust production has halved since the Cretaceous. On the contrary, Larson (1991) and Rowley (2002) suggested that the overall degassing rate showed little variation over the last 100 Ma. According to these reconstructions, the past solid Earth degassing rate is still largely unconstrained, despite its central role in the evolution of global climate during the Cenozoic A third mechanism for the atmospheric CO2 evolution is better constrained over the Cenozoic and refers to the continental drift. When moving either in latitude or in longitude, the continents go across various climatic belts. These displacements strongly influence the chemical weathering of rocks. Marshall et al. (1988) were the first to propose the existence of a causal link between the continental configuration, the silicate rock weathering and the atmospheric CO2 level, using a simple zonally averaged model of the land distribution. Bluth and Kump (1991) and Gibbs et al. (1999) showed with a GCM, and a realistic continental drift history, that the end Mesozoic–early Cenozoic periods have been characterized by an efficient silicate weathering promoted by intense rainfall. More recently, the continental drift effect on atmospheric CO2 and global climate has been explored with the GEOCLIM model for the whole Phanerozoic (Donnadieu et al., 2006; Goddéris et al., 2008, 2012; Nardin et al., 2011). Superimposed on the continental drift, several large igneous provinces (LIPs) erupted during the Cenozoic or slightly before (Deccan trapps, Ethiopian trapps; Courtillot and Renne, 2003). Together with older basaltic provinces, and because they weather much faster than continental shield rocks (Dessert et al., 2003), LIPs constitute potential hot spots for CO2 consumption when carried by continental drift into warm and humid areas (Kent and Muttoni, 2008; Lefebvre et al., 2010). In this contribution, we use the coupled carbon-climate model GEOCLIM to quantify the impact of the Cenozoic drift of the continents and the LIPs on the time evolution of the atmospheric CO2.
2. Model description 2.1. General framework of the GEOCLIM model We use the most recent version of the GEOCLIM model described by Donnadieu et al. (2009). The model couples the 3-D climate model FOAM, the dynamic global vegetation model LPJ (Sitch et al., 2003) and the box model for the geological carbon and alkalinity global cycle COMBINE (Goddéris and Joachimski, 2004). The atmospheric component of FOAM is a parallelized version of NCAR's Community Climate Model 2 (CCM2) with the upgraded radiative and hydrologic physics incorporated in CCM3 v. 3.2. The atmosphere runs at R15 spectral resolution (4.51 7.51) with 18 levels. For this study, we use FOAM in mixed-layer mode where the atmosphere is coupled to a 50-m mixed-layer ocean, which parameterizes heat transport through diffusion, in order to save
computation time (one GEOCLIM simulation requires up to 12 GCM simulations). The vegetation model LPJ is a Dynamic Global Vegetation Model of vegetation biogeography and vegetation/soil biogeochemistry. Starting from the climate, soil and atmospheric information as input, it dynamically computes the transient vegetation composition in terms of plant functional groups and their associated carbon and water budgets. The calculated vegetation distribution directly impacts on (1) the albedo of each grid cell, (2) its thermal properties, (3) its roughness, and (4) its potential evaporation. In this contribution, we aim to calculate the steady-state pCO2 associated with five continental configurations spanning the whole Cenozoic. On a geological timescale, the global carbon and alkalinity cycles are balanced: F vol ¼ F silw þ ðF od −F ow Þ
ð1Þ
where Fvol is the total solid Earth CO2 degassing flux and Fsilw is the total CO2 consumed by continental silicate rocks (Walker et al., 1981). Fod is the burial of organic carbon and Fow the oxidation of sedimentary reduced carbon exposed at the continental surfaces. The difference, Fod−Fow, fluctuated over the Cenozoic (Goddéris and François, 1996). However, because in the present study we focus on estimating the role of continental drift and its impact on silicate weathering on the Cenozoic climate evolution, we deliberately set the Fod−Fow term to 0. This assumes steady state of the organic carbon sub-cycle. Under this assumption, Eq. (1) simplifies as follows: F vol ¼ F silw
ð2Þ
Fsilw depends on climate (runoff and air temperature), amongst others, and climate depends on atmospheric CO2 pressure. Accounting for these dependencies, Eq. (1) is readily solved for the steadystate atmospheric CO2. The advantage of GEOCLIM when compared to previously published models such as GEOCARB (Berner, 2004) is to resolve the continental climate spatially, so that climatic parameters such as runoff, air temperature and consumption by continental silicate weathering, are calculated at each grid cell. We can rewrite Eq. (1) including the impact of the continental configuration: ncont
F vol ¼ ∑ F jsilw ðT j ; runof f j Þ j¼1
ð3Þ
where ncont is the number of continental grid elements, Tj and runoffj the air temperature and runoff calculated by the GCM for the grid element j respectively. FOAM is run for each paleogeographical configuration at various CO2 concentrations (from 160 to 2400 ppm) accounting for the secular trend in the solar constant. Each simulation is run for 30 yr (until steady state is reached). Assuming that the modeled climate system behaves quasi-linearly, we can estimate the climatic variables (temperature and runoff) for any pCO2 and continental configuration to calculate the F jsilw term on each continental grid cell. Then, if Fvol is known, we can estimate the atmospheric CO2 level necessary for equilibrating Fvol and Fsilw (Eq. (2)). 2.2. Model representation of the lithology The geochemical model differentiates weathering of basaltic and granitic lithologies. For the granitic lithologies, total CO2 consumption through weathering is based on the work by Oliva et al. (2003), who compiled data for about 100 small granitic catchments (Eq. (4)). They showed that the cation flux released by granitic weathering is proportional to the runoff and is a function of temperature, assuming an apparent activation energy of 48.7 kJ/mol. The CO2 consumption is proportional to the cation
V. Lefebvre et al. / Earth and Planetary Science Letters 371-372 (2013) 203–211
flux, because in freshwater the cationic charge is mostly balanced by HCO3− anions. CO2 consumption is then proportional to the runoff and is a function of temperature with the same activation energy as the total cation flux. graw npixel E 1 1 F graw ¼ ∑ kgraw ⋅runof f j ⋅areaj ⋅exp a − ð4Þ Tj T0 R j¼1 where Fgraw is the total atmospheric CO2 consumption via granitic weathering, R is the perfect gas constant, Eagraw is the apparent activation energy of 48.7 kJ/mol, areaj is the surface of the continental grid cells. Tj and runoffj are the mean annual ground air temperature and the mean annual runoff at the grid cell j respectively. Weathering of basaltic lithologies is calculated according to Dessert et al. (2001): " # npixel Ebasw 1 1 a F basw ¼ ∑ kbasw ⋅runof f j ⋅areaj ⋅exp ð5Þ − Tj T0 R j¼1 where Fbasw is the total atmospheric CO2 consumption flux through basaltic weathering. Ebasw is fixed at 42.3 kJ/mol. kbasw and kgraw are a calibrated on present day climatic conditions provided by a FOAM simulation in a mixed-layer configuration. By assuming a total CO2 consumption by silicate weathering (Fgraw+Fbasw) equal to 13.6 1012 mol/yr (Dessert et al., 2003; Gaillardet et al., 1999) with a 30% contribution of basaltic lithologies and by using the present day climatic fields from FOAM, kbasw (6.25 109 with runoff in cm/ yr and area in 106 km2) and kgraw (1.23 109 with runoff in cm/yr and area in 106 km2) are determined.
we calculate the steady state CO2 for each time period assuming a constant degassing rate (6.8 1012 mol/yr). A uniform distribution of the basaltic rocks at each terrestrial grid cell is prescribed and calibrated in order for the basalt weathering to reach a 30% contribution of the total silicate weathering flux taken at present day (Dessert et al., 2003). These simulations aim to show the effect of the continental drift. In the second set of simulations (REG simulation), the consumption of CO2 by continental silicate weathering is calculated based on the spatial distribution of the basaltic provinces. Fig. 1 shows prescribed distribution of basaltic provinces plotted in the atmospheric resolution of the model FOAM. The size of each basaltic province is listed in Table 1. In the REG simulation, solid Earth degassing rate is fixed to present day value. In the third set of simulations (VOL), igneous provinces are explicitly accounted for as in the REG simulation, but we now consider a time-varying degassing rate throughout the Cenozoic. This is deduced from the reconstruction of the subduction rate provided by Engebretson et al. (1992) (see discussion below). These simulations are summarized in Table 2.
3. Results 3.1. Role of the continental drift The most striking result comes from the UNI simulations where atmospheric CO2 stays relatively constant oscillating around 340 ppmv over the whole Cenozoic (Fig. 2; Table 3). This is not
2.3. Model configurations We use five geographic reconstructions of the Cenozoic including Early Paleocene (65 Ma), Early Eocene (52 Ma), the middle of the Oligocene (30 Ma), Middle Miocene (15 Ma) and Present-Day (PD). The Miocene geography comes from Herold et al. (2008). The land– ocean configurations for 65 Ma, 52 Ma and 30 Ma are built from a synthesis of paleomagnetic data, hot spot tracks and geologic constraints (Besse and Courtillot, 2002; Dercourt et al., 1993). Earth's orbital parameters are set to present-day values for the five configurations. Solar radiation is assumed to evolve through time according to the stellar evolution models (from 0.65% reduction in the Paleocene to 0.15% in the Miocene, Gough, 1981). We use the estimated CO2 thresholds below which ice is allowed to accumulate on Antarctica and Greenland at 750 ppm (DeConto and Pollard, 2003) and 280 ppm (DeConto et al., 2008; Lunt et al., 2008) respectively. Based on these estimations, and because the FOAM GCM does not include an ice sheet model allowing the onset of an ice sheet, (1) we assume the absence of ice sheets in all geographies when CO2 is above 750 ppm, (2) we prescribe the existence of both ice sheets (North and South) for CO2 levels lower or equal to 280 ppm, and (3) we prescribe the Antarctic ice sheet for CO2 levels between 280 and 750 ppm. The ice sheet relief and albedo values for Antarctica and Greenland are taken from their modern states. 2.4. Setup of the atmospheric CO2 simulations Over the Cenozoic, mechanical erosion appears to increase worldwide in response to the erosion of the major Cenozoic orogens (Kump and Arthur, 1997; Wold and Hay, 1990). In order to account for a fluctuating role of physical erosion in our simulations, potentially promoting silicate weathering (West et al., 2005), we simply multiply our granitic silicate weathering rate by a fR factor which varies from 0.75 at 65 Ma to 1 at 15 Ma (method and values are from Berner (2004)). We then run three sets of simulations. In the first one (UNI),
205
Col.
Is.
Sib. Kam.
Tim. Dec.
Cent. Am. present at :
Par.
Emei.
Eth.
Indo.
Kar.
65, 52, 30, 15, PD Pat. 52, 30, 15, PD 30, 15, PD 15, PD PD
Tas.
Fig. 1. Present-day localization of the main basaltic provinces and their age of formation. Comparing the area of an igneous province with the area of the model grid cell, we define the fraction of the cell covered by the igneous province (the rest being covered by shield rocks).
Table 1 Ages and areas of the main basaltic provinces considered in this study (Chen et al., 2008; Corgne et al., 2001; Courtillot and Renne, 2003; Dessert et al., 2003; Honthass et al., 1995; Kuznetsov et al., 2007; Meffre et al., 2004; Wignall, 2001). Basaltic provinces
Age (Ma)
Area (106 km2)
Kamchakta (Kam.) Patagonia (Pat.) Columbia River (Col.) Ethiopia (Eth.) Island/Greenland (Isl.) Deccan (Dec.) Central America (Cent. Am.) Indonesia/SE Asia (Indo.) Parana (Par.) Karoo (Kar.) Siberia (Sib.) Emeishan (Emei.) Timan/NW Russia (Tim.) Tasmania (Tas.)
5–12 10 16 30 60 65 90 90–250 133 184 250 259 550 580
0.17 0.21 0.2 0.8 0.1 0.5 0.31 0.54 0.57 0.22 1.5 0.5 0.18 0.33
Total
6.13
206
V. Lefebvre et al. / Earth and Planetary Science Letters 371-372 (2013) 203–211
Table 2 Description of the parameters used in the simulations. Simulation
Basalt repartition
Mechanical erosion (65–15 Ma)
Degassing rate
Aims
UNI REG
Uniform Explicit with the modern geographical distribution (Fig. 1) Explicit with the modern geographical distribution (Fig. 1)
0.75–1 0.75–1
Constant (6.8 1012 mol/yr) Constant (6.8 1012 mol/yr)
Continental drift Continental drift+basalt
0.75–1
Deduced from Engebretson et al. (1992) (Fig. S1)
Degassing rate
VOL
because the carbon cycle is insensitive to the change in continental configuration, but rather the result of efficient weathering during the Cenozoic damping CO2 down to low levels. Indeed, using the same model, Donnadieu et al. (2009) found higher CO2 levels for pre-Cenozoic continental configurations. In addition, subtle changes in the paleogeography of our model such as the northward drift of Pangea between the Carnian and the Rhaetian time periods (i.e. 220 and 200 Ma) can trigger an atmospheric CO2 fluctuation of 2000 ppm (Goddéris et al., 2008). To illustrate the effect of paleogeography in our model, we compare global weathering rates from climatic simulations with identical atmospheric CO2 level (280 ppmv, i.e. the pre-industrial value) (Fig. 3). In these series of simulations, the basaltic weathering rates are substantially larger in the 65, 52 and 30 Ma configurations and lower in the 15 Ma configuration than the modern value (Fig. 3b). For the 65, 52 and 30 Ma paleogeographies, the trend in granitic weathering rates is slightly different because of the erosion corrective factor that implicitly draws down the weathering rates of the granitic surfaces by 25% (Fig. 3a). As a result, the global weathering rate, the sum contribution of basaltic and granitic lithologies, stays below (but close to) the present-day level (Fig. 3c). With lower global weathering rates, atmospheric CO2 levels stay above the pre-industrial value in the model (Fig. 2). In this set of experiments, we did not aim to quantify the effect of the physical erosion of active orogens on the global carbon cycle. However our results suggest that the weathering feedback remains quite efficient during the early Cenozoic even in the absence of intense physical erosion. Modeled CO2 levels for the 65 Ma and 15 Ma time periods are well within the range of estimates, while modeled CO2 levels of the Oligocene and Eocene are at the lower end estimates (Fig. 2). However, all CO2 levels from the UNI simulations are lower than the threshold for initiating an extensive glaciation in Antarctica. In these simulations, the transition to an icehouse world would be triggered from the beginning of the Cenozoic. Because the main mechanism in the UNI simulation is the continental drift forcing the global climate to be cold well before the effective start of the glaciation, we conclude that another mechanism is required to explain the evolution of the long-term CO2 trend (and climatic trend) throughout the Cenozoic.
and 65 Ma-old Deccan traps are still located in subtropical regions where runoff is low, inducing smaller global silicate weathering rates (4.87 1012 mol of CO2/yr for the Paleocene vs. 6.8 1012 mol of CO2/yr for the modern, see Figs. 3 and 4). The lower basaltic weathering rate at the Paleocene is compensated in the model by higher atmospheric CO2 concentration in order to keep global silicate weathering and total solid Earth degassing in balance (Eq. (1)). It is only during the Eocene that the Deccan traps finally leave the subtropical latitudes to enter the ITCZ increasing basaltic and silicate weathering (Fig. 4). An additional contribution at the Oligocene to the global silicate weathering comes from the weathering of the Ethiopian traps below the ITCZ (Figs. 3 and 4). During the Miocene, the northward drift of Africa and India reduced the basaltic weathering over the Ethiopian traps and the Deccan traps respectively, resulting in higher CO2 concentration while reaching steady state in the model (Figs. 3 and 4). Finally, from the MidMiocene to the present day, the silicate weathering flux increases in response to the strengthening of precipitation and runoff in southeastern China and Indonesia due to the Tibetan plateau uplift and subsequent intensification of the South East Asian monsoon (Fig. 4). Observations show a more intense monsoon in Asia during the last 8–9 Ma (Zhisheng et al., 2001). The displacement of the Emeishan traps and the Indonesian basaltic archipelago in the vicinity of the ITCZ emphasizes this pattern of increased weathering over the last 15 Ma. The modeled evolution in the weathering of the main basaltic provinces, calculated for each time slice of the REG simulation, confirms this scenario (Fig. 5). The model shows that weathering specifically of the Deccan traps reaches a maximum at 52 Ma, followed by a decrease during the Oligocene to reach relatively low rates from the Miocene to today. The Ethiopian traps are rapidly weathered during the Oligocene, directly after their onset, and consume less CO2 during the Miocene and the presentday. Finally, our model simulations show that CO2 concentration levels kept decreasing in the last 15 Ma because of a strengthening of the East Asian monsoon that resulted in stronger weathering of Emeishan and Indonesia basaltic provinces. The explicit calculation of the effect of the basaltic provinces leads to a general increase in CO2 from the UNI to the REG simulations. Nevertheless, atmospheric CO2 remains too low, i.e. below the Antarctic glaciation threshold, during the Paleocene (680 ppm) and the Eocene (360 ppm).
3.2. Role of the weathering of the LIPs 3.3. Role of the degassing rate The REG simulation (where basaltic provinces are spatially distributed) shows larger fluctuations in the atmospheric CO2 levels than the UNI simulation. In REG, atmospheric CO2 concentration stays relatively high during the Paleocene (680 ppm) and the Miocene (500 ppm), and drops to about 340 ppm at the Eocene and Oligocene (Fig. 2; Table 3). To separate the different causes in the REG CO2 trend, we calculate the silicate weathering flux for each reconstruction set with 280 ppm of CO2 (Figs. 3 and 4). The model results show that the surface of basaltic provinces experiencing a warmer and wetter climate is smaller during the Paleocene than during the rest of the Cenozoic. This is because the Ethiopian traps have not been emplaced at that time, and the 258 Ma-old Emeishan
The evolution of the global degassing rate throughout the Cenozoic, and more generally throughout the Earth history, remains poorly constrained. There are (1) no direct measurements of the Cenozoic fluctuations of the solid Earth degassing and (2) no clear consensus between the various reconstructions of the degassing flux through the geological past. For instance, recent re-evaluations suggest that because the ridge production is estimated to have stayed roughly constant for the last 180 Ma, the degassing rate should have had little fluctuations since the middle Jurassic (Rowley, 2002). However, as already noticed by Kerrick and Caldeira (1999), processes other than ridge production can influence
V. Lefebvre et al. / Earth and Planetary Science Letters 371-372 (2013) 203–211 5000 4000
UNI REG VOL
3000
Antarctic ice sheet N. Hemisphere ice sheet
2000
pCO2 (ppmv)
207
1000
Antarctic ice sheet threshold
500 400 300 N.H. threshold
Ma.
Paleo.
Eocene
Oligo.
Pli.
Cenozoic
Meso.
70
Miocene
Ple.
200
60
50
40
30
20
10
0
Time (Myr) Fig. 2. Atmospheric CO2 levels simulated by GEOCLIM. Black squares are the pCO2 data compiled by Beerling and Royer (2011). Occurrence of Antarctic and Northern Hemisphere ice sheets are taken from Zachos et al. (2001). Because our knowledge of the paleogeography is tied to the time resolution of the paleomagnetic data, we assume that the steady-state CO2 levels are valid within a 5 Ma time window and give the multi-million years baseline of the CO2 evolution, missing potential rapid shifts.
the Earth degassing rate. During the Paleocene and the Eocene, Kerrick and Caldeira (1999) suggested that the Earth degassing rate may have increased in response to regional metamorphism of the Tethysian orogeny, following the subduction of carbonates accumulated on the seafloor of the European and Asian margins. Using this information along with Engebretson et al.'s (1992) work on subduction history, we estimate that the global degassing rate was higher between 130 and 50 Myr than today (Fig. A1). We investigate here the potential impact of this degassing rate history on modeled atmospheric CO2. Increasing the degassing rate by 50% results in our model an increase of atmospheric CO2 to about 1310 ppmv at the Paleocene and 930 ppmv at the Eocene (Fig. 2; Table 3). These new CO2 levels are then well above the Antarctic glaciation threshold (750 ppm) and consequently more consistent with the Cenozoic climate history. Including the timing estimates of degassing rate by Engebretson et al. (1992) did not change our steady state CO2 levels at the Oligocene and Miocene calculated in the REG simulation. Details of the continental temperature, runoff and precipitation minus evaporation can be found in the Appendix (Figs. A2–A8).
4. Discussions The long-term CO2 calculated in the VOL simulation is in good agreement with the evolution of the climate during the Cenozoic. The intensity of weathering during the Paleocene and Eocene requires an enhanced CO2 source (Engebretson et al., 1992) to allow a greenhouse climate that prevents the Antarctic glaciation at that time. However, in our modeling study, we assumed that the organic carbon sub-cycle (i.e. the balance between the organic carbon burial and the sedimentary organic carbon oxidation over continents) is at steady state. Instead of a high solid Earth degassing rate, a low organic carbon burial (or strong kerogen oxidation) could alternatively lead to higher CO2 levels during the Early Cenozoic. This is a possibility based on the carbon isotope budget, which suggests that organic carbon burial has increased over the course of the Cenozoic (Derry and FranceLanord, 1996). We use the compilation of δ18O provided by Prokoph et al. (2008) to calculate the surface seawater temperature in midlatitudes (30–601N and S) using the paleothermometer equation of Anderson and Arthur (1983): T ¼ 16−4:14ðδ18 O−δ18 Osw Þ þ 0:13ðδ18 O−δ18 Osw Þ2
ð6Þ
Table 3 Steady-state atmospheric CO2 level calculated for each simulation. Period
Paleocene Eocene Oligocene Miocene PD
Degassing rate
Calculated steady-state pCO2 UNI (ppm)
REG
VOL
VOL (ratio)
340 320 350 330 280
680 360 320 500 280
1315 930 360 475 280
1.35 1.46 1.05 0.98 1
where T is the seawater temperature, δ18O is the oxygen isotope value of the sedimentary carbonate data and δ18Osw is the δ18O of the seawater (fixed to 0‰ from 30 Ma to modern, and −1‰ for 65 and 52 Ma). The temperature in our VOL simulations does not have exact same trend in temperature as the 10 Myr-averaged temperature depicted by the oxygen isotopes (Fig. 6). While there is good agreement for the 52 Ma, 30 Ma and 15 Ma time periods, our VOL simulation tends to overestimate the sea surface temperature for the 65 Ma time period. Conversely, the temperature from the UNI simulation fits the oxygen isotope estimates better at the 65 Ma periods, modeled with values of atmospheric CO2 level of 340 ppm. We note here that this estimate corresponds to the atmospheric CO2 level estimated by Beerling and Royer (around 400 ppm on their curve). However, this is below the threshold value for glaciation of Antarctica and thus too low for the 65 Ma time period. Nevertheless, the values of the glaciation thresholds are based on modern and Oligocene continental configurations. At the beginning of the Paleocene however, Australia was still attached to Antarctica. This configuration may be in favor of a lower atmospheric CO2 threshold. With a larger continental mass centered around the pole, warmer summer temperature should induce lower snow accumulation, reducing the atmospheric CO2 threshold for southern continental glaciation. This hypothesis should be tested in a coupled ice-climate model in the future. The Eocene is a warm period in the model as the result of the intense degassing rate that can largely overcompensate for the elevated CO2 consumption linked to the northward drift of the Deccan basalts. The decrease in degassing rate from the Eocene to the Oligocene causes atmospheric CO2 to cross the threshold of Antarctic glaciation, which might have happened between 52 and
V. Lefebvre et al. / Earth and Planetary Science Letters 371-372 (2013) 203–211
Ma.
Paleo.
Eocene
Meso.
Oligo.
Miocene
Pli.
Ple.
CO2 consumption rate (1012 mol/yr)
208
Cenozoic Time (Myr)
Fig. 3. Evolution of the (a) granitic, of the (b) basaltic and of the (c) global weathering rates (in mol CO2/yr) at 280 ppm throughout the Cenozoic. Circles (squares) stand for conditions defined in the UNI (REG) simulation, see text for more details.
30 Ma. To narrow down when the glaciation of Antarctica started within this time window requires additional continental configurations. The model suggests that the mid-latitude averaged temperature at the Oligocene is close to the pre-industrial value (14 1C) and consistent with the presence of an ice sheet on Antarctica. During the Miocene, the modeled atmospheric CO2 and temperature increases in response to a decrease of continental weathering. The northward drift of India and Africa tends to reduce the areas of granitic and basaltic lithologies below the ITCZ during that time period. The resulting steady state CO2 level is 475 ppm, which is consistent with the 400–500 ppm estimation of Kürschner et al. (2008) at the beginning of the Middle Miocene (15.5 Ma). Our result is also consistent with the modeling work by Hamon et al. (2012) but disagrees with estimates from marine phytoplankton carbon isotopes which estimate a pCO2 as low as 280 ppmv (Pagani et al., 2005, 2011). However, a warmer Miocene is in agreement with the record of the Mid-Miocene Climate Optimum (MMCO) (Kürschner et al., 2008; Mosbrugger et al., 2005; Zachos et al., 2001) which could be related at least in our model to a decrease in the CO2 consumption rate by weathering.
5. Conclusions We used a spatially resolved climate-carbon model (GEOCLIM) and investigated the influence of continental drift on the long-
term evolution of the Cenozoic atmospheric CO2. Our model shows that the continental configuration should set the required conditions for the onset of a large-scale glaciation at the beginning of the Cenozoic. This is true even when we imposed in the model a decrease of the weathering owing to the absence of major uplift during the early Cenozoic. The dispersed configuration and the location of a large continental area (North Africa, northern South America) within the ITCZ promoted the CO2 consumption by weathering, forcing CO2 to remain low. Using a 2D representation of the basaltic provinces in the model, we found that Paleocene had higher atmospheric CO2 because of the location of the Deccan traps in the subtropics, which triggered lower CO2 consumption by weathering. The weathering effect of this northward drift of the Deccan traps at the Eocene is so high that atmospheric CO2 is too low in triggering the buildup of the Antarctic ice sheet before the observed pattern. This result is in disagreement with the scenario suggested by Kent and Muttoni (2008). Here we suggest that Antarctic ice sheet did not develop before the Oligocene because of a high degassing rate at the Eocene. In other words, our model results suggest that the transition from a greenhouse Eocene to an icehouse Oligocene is mainly the result of a reduction in CO2 degassing rate. Finally during the Miocene period, our model results show that the northward drift of Africa and India causes CO2 removal to decrease with the weathering, followed by an increase in the atmospheric CO2 level to up to 500 ppm. The modeling of the
V. Lefebvre et al. / Earth and Planetary Science Letters 371-372 (2013) 203–211
209
PD
SST (°C)
Silicate weathering rate (1010 mol/yr)
15Ma
52Ma
30Ma
65Ma
Fig. 4. Evolution of the annual silicate weathering rate in mol CO2/yr and of the annual averaged air temperature above the ocean for the paleogeographical reconstructions used in this study at 280 ppm CO2. Conditions used to calculate the weathering rates are those defined for the REG simulation.
Ma.
Paleo.
Eocene
Meso .
Oligo.
Miocene
Pli.
Ple.
CO2 consumption rate (1012 mol/yr)
Ethiopia Deccan Emeishan Indonesia
Cenozoic Time (Myr)
Fig. 5. CO2 consumption by silicate weathering history of the four main basaltic provinces during the Cenozoic as calculated in the REG simulation.
effect of the continental drift on atmospheric CO2 favors the presence of a Middle Miocene Climate Optimum. Finally, we found that the decrease of the atmospheric CO2 level from 500 ppm
during the Miocene to its pre-industrial values is induced by an increase in silicate weathering, due to a more intense South-East Asian monsoon.
210
V. Lefebvre et al. / Earth and Planetary Science Letters 371-372 (2013) 203–211
25 India efficient weathering Antarctic ice sheet N. Hemisphere ice sheet
Temperature (°C)
20
15
M a. 5
Paleo.
Eocene
Oligo.
Miocene
P li.
Ple.
10
Cenozoic
Meso.
60
50
40
30
20
10
0
Time (Myr) Fig. 6. Evolution of the mid-latitude (32–581N and S) sea surface temperature calculated from a carbonate δ18O compilation (Prokoph et al., 2008). Gray squares are the 10 Myr averaged temperature of this dataset; the vertical bars stand for the standard deviation. The blue triangles are the 30–601S and N sea surface averaged temperature calculated by FOAM accounting for a variable solid Earth degassing (VOL simulation). The vertical bars represent the maximum and minimum calculated temperatures occurring in these areas. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Acknowledgments This work was supported by the ANR project COLORS (No. ANR09-JCJC-0105) and by the ANR project CALIMERO (No. ANR-09BLAN-0221). We thank two anonymous reviewers for their constructive comments. We thank F. Monteiro, N. MacBean and S. Taylor for improving the readability of this contribution.
Appendix A. Supplementary matrials Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.epsl.2013.03.049.
References Anderson, T.F., Arthur, M.A., 1983. Stable isotopes of oxygen and carbon and their application to sedimentological and paleoenvironnemental problems. In: Arthur, M.A., Anderson, T.F., Kaplan, I.R., Veizer, J., Land, L. (Eds.), Stable Isotopes in Sedimentary Geology. Society of Economic Paleontologists and Mineralogists Tulsa, Oklahoma, USA, pp. 1–151. Beerling, D.J., Royer, D.L., 2011. Convergente Cenozoic CO2 history. Nature Geoscience 4, 418–420. Berner, R.A., 2004. A model for calcium, magnesium and sulfate in seawater over phanerozoic time. Am. J. Sci. 304, 438–453. Besse, J., Courtillot, V., 2002. Apparent and true polar wander and the geometry of the geomagnetic field over the last 200 Myr. J. Geophys. Res. 107, 2300. Bouchez, J., Gaillardet, J., Lupker, M., Louvat, P., France-Lanord, C., Maurice, L., Armijos, E., Moquet, J.-S., 2012. Floodplains of large rivers: weathering reactors or simple silos? Chem. Geol. 332–333, 166–184. Bluth, G.J.S., Kump, L.R., 1991. Phanerozoic paleogeology. Am. J. Sci. 291, 284–308. Chen, C.H., Lee, C.Y., Shinjo, R.I., 2008. Was there Jurassic paleo-Pacific subduction in South China?: constraints from 40Ar/39Ar dating, elemental and Sr–Nd–Pb isotopic geochemistry of the Mesozoic basalts. Lithos 106, 83–92. Corgne, A., Maury, R.C., Lagabrielle, Y., Bourgois, J., Suarez, M., Cotten, J., Bellon, H., 2001. La diversité des basaltes de Patagonie à la latitude du point triple du Chili (461–471lat. S) : données complémentaires et implications sur les conditions de la subduction. C. R. Acad. Sci.—Ser. IIA—Earth Planet. Sci. 333, 363–371. Courtillot, V.E., Renne, P.R., 2003. On the ages of flood basalt events. C. R. Geosci. 335, 113–140. DeConto, R.M., Pollard, D., 2003. Rapid Cenozoic glaciation of Antarctica induced by declining atmospheric CO2. Nature 421, 245–249. DeConto, R.M., Pollard, D., Wilson, P.A., Palike, H., Lear, C.H., Pagani, M., 2008. Thresholds for Cenozoic bipolar glaciation. Nature 455, 652.
Dercourt, J., Ricou, L.E., Vrielynck, B., 1993. Atlas Tethys Palaeoenvironments Maps. Gauthier-Villars, Paris. Derry, L.A., FranceLanord, C., 1996. Neogene growth of the sedimentary organic carbon reservoir. Paleoceanography 11, 267–275. Dessert, C., Dupré, B., François, L.M., Schott, J., Gaillardet, J., Chakrapani, G., Bajpai, S., 2001. Erosion of Deccan Traps determined by river geochemistry: impact on the global climate and the 87Sr/86Sr ratio of seawater. Earth Planet. Sci. Lett. 188, 459–474. Dessert, C., Dupré, B., Gaillardet, J., Francois, L.M., Allègre, C.J., 2003. Basalt weathering laws and the impact of basalt weathering on the global carbon cycle. Chem. Geol. 202, 257–273. Donnadieu, Y., Godderis, Y., Bouttes, N., 2009. Exploring the climatic impact of the continental vegetation on the Mesozoic atmospheric CO2 and climate history. Clim. Past 5, 85–96. Donnadieu, Y., Godderis, Y., Pierrehumbert, R., Dromart, G., Fluteau, F., Jacob, R., 2006. A GEOCLIM simulation of climatic and biogeochemical consequences of Pangea breakup. Geochem. Geophys. Geosyst., 7, http://dx.doi.org/10.1029/ 2006GC001278. Engebretson, D.C., Kelley, K.P., Cashman, H.J., Richards, M.A., 1992. 180 million years of subduction. GSA Today 2, 94–100. France-Lanord, C., Derry, L.A., 1997. Organic carbon burial forcing of the carbon cycle from Himalayan erosion. Nature 390, 65–67. Gaffin, S., 1987. Ridge volume dependence on sea-floor generation rate and inversion using long-term sealevel change. Am. J. Sci. 287, 596–611. Gaillardet, J., Dupre, B., Louvat, P., Allegre, C.J., 1999. Global silicate weathering and CO2 consumption rates deduced from the chemistry of large rivers. Chem. Geol. 159, 3–30. Galy, A., France-Lanord, C., 1999. Weathering processes in the Ganges–Brahmaputra basin and the riverine alkalinity budget. Chem. Geol. 159, 31–60. Galy, V., France-Lanord, C., Beyssac, O., Faure, P., Kudrass, H., Palhol, F., 2007. Efficient organic carbon burial in the Bengal fan sustained by the Himalayan erosional system. Nature 450, U406–U407. Gibbs, M.T., Bluth, G.J.S., Fawcett, P.J., Kump, L.R., 1999. Global chemical erosion over the last 250 my: variations due to changes in paleogeography, paleoclimate, and paleogeology. Am. J. Sci. 299, 611–651. Goddéris, Y., Donnadieu, Y., de Vargas, C., Pierrehumbert, R.T., Dromart, G., van de Schootbrugge, B., 2008. Causal or casual link between the rise of nanoplankton calcification and a tectonically-driven massive decrease in Late Triassic atmospheric CO2? Earth Planet. Sci. Lett. 267, 247–255. Goddéris, Y., Donnadieu, Y., Lefebvre, V., Le Hir, G., Nardin, E., 2012. Tectonic control of continental weathering, atmospheric CO2, and climate over Phanerozoic times. C. R. Geosci. 344, 652–662. Goddéris, Y., François, L.M., 1996. Balancing the Cenozoic carbon and alkalinity cycles: constraints from isotopic records. Geophys. Res. Lett. 23, 3743–3746. Goddéris, Y., Joachimski, M.M., 2004. Global change in the Late Devonian: modelling the Frasnian–Famennian short-term carbon isotope excursions. Palaeogeogr., Palaeoclimatol., Palaeoecol. 202, 309–329. Gough, D.O., 1981. Solar interior structure and luminosity variations. Sol. ar Phys. 74, 21–34.
V. Lefebvre et al. / Earth and Planetary Science Letters 371-372 (2013) 203–211
Hamon, N., Sepulchre, P., Donnadieu, Y., Henrot, A.-J., François, L., Jaeger, J.-J., Ramstein, G., 2012. Growth of subtropical forest in Miocene Europe: the roles of carbon dioxide and Antarctic ice volume. Geology 40, 567–570. Herold, N., Seton, M., Muller, R.D., You, Y., Huber, M., 2008. Middle Miocene tectonic boundary conditions for use in climate models. Geochem. Geophys. Geosyst, 9, http://dx.doi.org/10.1029/2008GC002046. Honthass, C., Bellon, H., Kepezhinskas, P.K., Maury, R.C., 1995. Nouvelles datations 40 K–40Ar du magmatisme Crétacé à Quaternaire du Kamchatka du Nord (Russie). Elsevier, Paris, France. Kent, D.V., Muttoni, G., 2008. Equatorial convergence of India and early Cenozoic climate trends. Proc. Natl. Acad. Sci. USA 105, 16065–16070. Kerrick, D.M., Caldeira, K., 1999. Was the Himalayan orogen a climatically significant coupled source and sink for atmospheric CO2 during the Cenozoic? Earth Planet. Sci. Lett. 173, 195–203. Kump, L.R., Arthur, M.A., 1997. Global chemical erosion during the Cenozoic: weatherability balances the budget. In: Ruddiman, W. (Ed.), Tectonic Uplift and Climate Change. Plenum Publishing Co., New York, pp. 399–426. Kürschner, W.M., Kvaček, Z., Dilcher, D.L., 2008. The impact of Miocene atmospheric carbon dioxide fluctuations on climate and the evolution of terrestrial ecosystems. Proc. Natl. Acad. Sci. USA 105, 449–453. Kuznetsov, N.B., Soboleva, A.A., Udoratina, O.V., Hertseva, M.V., Andrelchev, V., 2007. Pre-Ordovician tectonic evolution and volcano-plutonic associations of the Timanides and northern Pre-Uralides, northeast part of the East European Craton. Gondwana Res. 12, 305–323. Larson, R.L., 1991. Latest pulse of Earth—evidence for a Midcretaceous superplume. Geology 19, 547–550. Lefebvre, V., Servais, T., François, L., Averbuch, O., 2010. Did a Katian large igneous province trigger the Late Ordovician glaciation? A hypothesis tested with a carbon cycle model. Palaeogeogr. Palaeoclimatol.,Palaeoecol 296, 310–319. Lunt, D.J., Foster, G.L., Haywood, A.M., Stone, E.J., 2008. Late Pliocene Greenland glaciation controlled by a decline in atmospheric CO2 levels. Nature 454, 1102–1105. Marshall, H.G., Walker, J.C.G., Kuhn, W.R., 1988. Long-term climate change and the geochemical cycle of carbon. J. Geophys. Res.—Atmos. 93, 791–801. Meffre, S., Direen, N.G., Crawford, A.J., Kamenetsky, V., 2004. Mafic volcanic rocks on King Island, Tasmania: evidence for 579 Ma break-up in east Gondwana. Precambrian Res. 135, 177–191. Moquet, J.-S., Crave, A., Viers, J., Seyler, P., Armijos, E., Bourrel, L., Chavarri, E., Lagane, C., Laraque, A., Casimiro, W.S.L., Pombosa, R., Noriega, L., Vera, A., Guyot,
211
J.-L., 2011. Chemical weathering and atmospheric/soil CO2 uptake in the Andean and foreland Amazon basins. Chem. Geol. 287, 1–25. Mosbrugger, V., Utescher, T., Dilcher, D.L., 2005. Cenozoic continental climatic evolution of Central Europe. Proc. Natl. Acad. Sci. USA 102, 14964–14969. Nardin, E., Godderis, Y., Donnadieu, Y., Le Hir, G., Blakey, R.C., Puceat, E., Aretz, M., 2011. Modeling the early Paleozoic long-term climatic trend. Geol. Soc. Am. Bull. 123, 1181–1192. Oliva, P., Viers, J., Dupre, B., 2003. Chemical weathering in granitic environments. Chem. Geol. 202, 225–256. Pagani, M., Huber, M., Liu, Z.H., Bohaty, S.M., Henderiks, J., Sijp, W., Krishnan, S., DeConto, R.M., 2011. The role of carbon dioxide during the onset of Antarctic glaciation. Science 334, 1261–1264. Pagani, M., Zachos, J.C., Freeman, K.H., Tipple, B., Bohaty, S., 2005. Marked decline in atmospheric carbon dioxide concentrations during the Paleogene. Science 285, 876–879. Prokoph, A., Shields, G.A., Veizer, J., 2008. Compilation and time-series analysis of a marine carbonate δ18O, δ13C, 87Sr/86Sr and δ34S database through Earth history. Earth-Sci. Rev. 87, 113–133. Raymo, M.E., Ruddiman, W.F., 1992. Tectonic forcing of Late Cenozoic climate. Nature 359, 117–122. Rowley, D.B., 2002. Rate of plate creation and destruction: 180 Ma to present. Geol. Soc. Am. Bull. 114, 927–933. Sitch, S.N., Smith, B., Prentice, I.C., Arneth, A., Bondeau, A., Cramer, W., Kaplan, J.O., Levis, S., Lucht, W., Sykes, M.T., Thonicke, K., Venevsky, S., 2003. Evaluation of ecosystem dynamics, plant geography and terrestrial carbon cycling in the LPJ dynamic global vegetation model. Global Change Biol. 9, 161–185. Walker, J.C.G., Hays, P.B., Kasting, J.F., 1981. A negative feedback mechanism for the long-term stabilization of Earth’s surface-temperature. J. Geophys. Res.—Atmos. 86, 9776–9782. West, A.J., Galy, A., Bickle, M., 2005. Tectonic and climatic controls on silicate weathering. Earth Planet. Sci. Lett. 235, 211–228. Wignall, P.B., 2001. Large igneous provinces and mass extinction. Earth-Sci. Rev. 53, 1–33. Wold, C.N., Hay, W.W., 1990. Estimating ancient sediment. Am. J. Sci. 290, 1069–1089. Zachos, J.C., Pagani, M., Sloan, L., Thomas, E., Billups, K., 2001. Trends, rhythms, and aberrations in global climate 65 Ma to Present. Science 292, 686–693. Zhisheng, A., Kutzbach, J.E., Prell, W.L., Porter, S.C., 2001. Evolution of Asian monsoons and phased uplift of the Himalayan Tibetan plateau since Late Miocene times. Nature 411, 62–66.