21Large Igneous Provinces (LIPs) have been emplaced throughout Earth's history, erupting that can release significant volumes of carbon into the ocean-atmosphere system (Bryan and , 2008; Coffin and Eldholm, 1994; Courtillot and Renne, 2003; Ernst et al., 2005).47 Ernst
48While the LIPs with the largest volumes (e.g. the Siberian Traps or the Central Atlantic
49Magmatic Province) are widely accepted to have triggered episodes of carbon cycle 50 perturbation, global warmth and ecological crisis (Grard et al., 2005; Sobolev et al., 2011; 51 Wignall, 2005 51 Wignall, , 2001, the case for LIPs with smaller (≤2x10 6 km 3 ) volumes (e.g., the Deccan
52Traps or the Columbia River Basalt), emitting sufficient CO 2 to cause a significant global 53 impact is a subject of debate (Caldeira and Rampino, 1990; Diester-Haass et al., 2009; 54 Kender et al., 2009; Self et al., 2006; Taylor and Lasaga, 1999 2011a; Erwin, 2006; Ganino and Arndt, 2010, 2009; Iacono-Marziano et al., 2012; Retallack 60 and Jahren, 2008; Svensen et al., 2009 Svensen et al., , 2004, and the potential impact of the recycling of 61 mafic crust into the LIP magma source (Sobolev et al., 2011 2013; Barry et al., 2013Barry et al., , 2010 Hooper, 1997 Hooper, , 1988 Reidel et al., 2013b; Wolff and Ramos, 78 2013) (Fig. 2), with the initiation of the CRB eruptions coinciding with the core of both the 79 MCO and the MCIE (Fig. 1). In detail, the peak of the carbon cycle perturbation occurs ca.
8016.0 Ma (Fig. 3) parameters, and the lack of dynamic ocean, atmosphere and terrestrial biosphere components.
119Both models are tuned with estimates for mid-Miocene parameters (Tables S1-3) burial, is also analysed by calculating a sensitivity index using the formula:where S is the sensitivity index, p the parameter value, Y the model results for p, and '
143denoting the parameter value and model results after the ± 10 % change in p (Haefner, 1996).
144A sensitivity index value of S ≤ ±0. the Siberian Traps) ( Table 2). An EAF of 2.0-6.4 also applies to this crustal-contamination 217 emission scenario.
218An additional source of cryptic degassing beyond magma degassing is the In Figure 4 we compare palaeorecords with model simulations for δ 13 C, CCD and 247 atmospheric CO 2. We find that sub-aerial basalt degassing alone has a negligible effect in our 248 simulations, but that adding cryptic degassing results in excursions similar to those seen in 249 the palaeorecords. In some respects these results are counter-intuitive because they simulate: carbon an isotopic value of -12 ‰ results in a δ 13 C excursion similar to but slightly larger 266 than in our best-fit simulation in MTW08 (Fig. S1), indicating that our main conclusions are 267 not particularly sensitive to uncertainty in this parameter.
268We find that in MTW08 an emission scenario of 1000 Pg C from the Steens and 3000-4000 Pg C of this emitted during just the GRB eruptions, best fits the palaeorecords.
279This emission range falls within the upper end of the range 460-6...