Carbon released by sill intrusion into young sediments measured through scientific drilling

Hydrocarbon formation driven by igneous sill intrusion into sedimentary basins has been extensively studied in both field and laboratory settings. Early scientific interest in using the steep thermal gradients surrounding sills as analogs for hydrocarbon maturation with burial (Simoneit et al., 1978) expanded into study of the metamorphic aureole of sills themselves (e.g., Saxby and Stephenson, 1987). The demonstration that sill intrusion in the North Atlantic Ocean basin could plausibly explain rapid early Eocene warming and the associated negative carbon isotope excursion (Svensen et al., 2004) was followed by studies linking these processes to end-Permian extinctions (Svensen et al., 2004; Kaiho et al., 2009), Early Jurassic warming (Svensen et al., 2007), and Cretaceous anoxia (Turgeon and Creaser, 2008). The discovery of sill-hosted hydrothermal systems in Guaymas Basin (Curray et al., 1982), Gulf of California, extended this interest to ecosystems supported by high-temperature alteration products (e.g., Zierenberg and Holland, 2004; Teske et al., 2014) and introduced the notion that processes driven by sill intrusion, including alteration and release of gases out of the seabed, persist through time in many marine settings and are not limited to singular igneous events.

International Ocean Discovery Program (IODP) Expedition 385 to Guaymas Basin (Fig. 1) explored the influence of systematic, widespread igneous sill intrusion on carbon cycling in young ocean basins. Guaymas Basin is a marginal basin that transitioned from rifting to seafloor spreading ca. 6 Ma (Lizarralde et al., 2007; Miller and Lizarralde, 2013) and thus has attributes in common with young continental rifts and back-arc basins, including marine and terrestrial sedimentation, high heat flow, and active magmatism. Seismic data show that active magmatism in Guaymas Basin is not confined to the divergent plate boundary, but it instead extends tens of kilometers off axis, where it is manifest as sill intrusions into sediments (Lizarralde et al., 2011). In contrast to on-axis hydrothermalism, off-axis sill magmatism has the potential to impact a much larger volume of sedimentary carbon (Lizarralde et al., 2011) and is analogous to sedimentary basin intrusions implicated in climate perturbations.

A key objective of Expedition 385 was to quantify sill-driven carbon release from off-axis sills. Sites U1545 and U1546 (Fig. 1) were chosen to address this objective by comparison of cores from the two sites, one affected by a massive sill intrusion and one unaffected. The close spacing (~1.1 km) of the sites and the relatively simple stratigraphy enable confident correlation between sites. The sill at Site U1546 is shallow enough to be accessible to scientific drilling, thick enough (~77 m) to have an unambiguous thermal impact, and thin enough to enable coring through the sill and into underlying sediments.

Sites U1545 and U1546 were cored to depths of 503.3 and 540.2 m below seafloor (mbsf), with high core recovery (Teske et al., 2021a, 2021b). The recovered sediments at both sites consist of a mix of biogenic and terrigenous components, primarily siliceous diatom tests, siliciclastic particles, and authigenic carbonate as micrite and nodules. The sediments are mostly laminated, implying limited burrowing and a low-oxygen seafloor environment. Total organic carbon (TOC) measured in core samples decreases from ~5 wt% at the seafloor tô2 wt% at 200 mbsf for both sites. Wireline logging measurements at each site, including seismic velocity, natural gamma radiation, and density, enabled accurate depth migration of the seismic profile crossing the sites (Fig. 2); detailed stratigraphic correlation (see the Supplemental Material¹); and the conversion from wt% TOC to volumetric weight of carbon (Fig. 3).

The 77-m-thick sill cored at Site U1546 is evident in the seismic profile, with a top at ~355 mbsf and notable deformation of the overlying sediments. The entire thickness of the sill was drilled, with 73% core recovery. The stratigraphic correlation between sites indicates that the youngest and oldest possible paleoseafloor horizons at the time of sill intrusion are marked by horizons Y and 3, respectively, in Figure 2. Biostratigraphic datums identified in cores (Teske et al., 2021a, 2021b) provide a sedimentation rate of ~102 cm/k.y. The sill thus intruded between 76 ka at 277 mbsf and 149 ka at 201 mbsf. In situ formation temperature measurements were made every ~30 m at each site. The observed linear geothermal gradient of ~223 °C/km indicates steady-state heat flow and implies thermal equilibration of the sill. Thus, the temperature of sediment surrounding the sill at the time of intrusion would have been similar to modern values of ~50–65 °C. Core observations of grain sizes and igneous textures suggest that the sill formed from a single magmatic pulse and cooled slowly enough to develop smooth transitions from basaltic to doleritic rock types (Teske et al., 2021b).

The metamorphic aureole of the sill at Site U1546 can be estimated from changes in TOC measured in core samples (Teske et al., 2021a, 2021b) and from the production index (PI) calculated from source-rock analyses (SRA; Teske et al., 2021c). PI, the ratio of free hydrocarbon to free plus cracking-generated hydrocarbon (both measured using SRA; Peters, 1986), enables oil-generation (0.1 < PI < 0.3) and gas-generation (PI > 0.3) zones to be inferred. PI values for U1546 samples are shown in Figure 3 along with green- and blue-shaded areas indicating the oil- and gas-generation zones inferred from that index. Above the upper aureole, average measured TOC is ~30 kg/m³ at correlative depths for both sites. Site U1546 values are consistently below that average, and they are also less than correlative U1545 values, beginning at ~16 m above the sill, which corresponds to the top of the oil-generation zone, and they decrease significantly beginning at ~8 m above the sill, similar to the gas-generation zone. Below the sill, average TOC for Site U1546 samples is near 40 kg/m³ below 455 mbsf, though the TOC of the deepest correlative Site U1545 samples, at ~31 kg/m³, is similar to values above the sill. A notable decrease in TOC upward toward the sill base occurs over ~15 m beginning at ~447 mbsf, consistent with the base of the gas-generation zone. The alteration of sediment is visibly evident as color and texture changes in cores from beneath the sill over a depth interval of ~5 m (Fig. 4).

The amount of “released carbon” is understood to be the difference between the amount of organic carbon in sediments before versus after the intrusion and alteration by a sill. Estimating carbon release involves defining the altered zone and subtracting estimates of organic carbon present after intrusion from carbon present before intrusion. The postintrusion case can be confidently estimated based on TOC measurements and SRA performed on sediment core samples from Site U1546 (Teske et al., 2021b, 2021c). The pre-intrusion case can similarly be estimated based on TOC measurements outside of the metamorphic aureole at Site U1546 and comparison with TOC measured in correlative strata from Site U1545 (Fig. 3).

The core-sample measurements of TOC within the aureole indicated a steep decrease toward the sill. We estimated TOC present within the aureole to decrease exponentially toward the sill from a reference value to zero, yielding a release of ~70% of the available organic carbon over a given interval. We considered two cases, one with carbon release from the oil-generation boundary to the sill (gas + oil case) and one from the gas-generation boundary (gas-only case). We used reference values of 30 kg C/m³ above the sill and 40 kg C/m³ below the sill. The reference values yielded pre-intrusion organic carbon spatial densities of 1364 kg C/m² and 843 kg C/m² for the gas + oil and gas-only cases. Subtracting calculated TOC within the aureole yielded spatial densities of 983 kg C/m² and 586 kg C/m² for the two cases, with an average of 784 kg C/m². Multiplying these densities by an estimated 4.2 km² area of the sill at Site U1546 (Fig. 1) yielded maximum and minimum amounts released of 4.12 and 2.46 Mt of C, or an average amount of 3.29 Mt of C released due to sill intrusion. This is equivalent to 3.86 Mt of CH₄ or 12.85 Mt of CO₂.

The estimated carbon released linked to the 77-m-thick sill at Site U1546 is substantial, equivalent to about one quarter of the annual global CO₂ flux from terrestrial volcanism (53.1 Mt of CO₂; Fischer et al., 2019). However, our estimate of 3.3 Mt of C is small for an intruded sill of this size relative to computations using published models. Modeling by Aarnes et al. (2010) predicts that a 100-m-thick, 4.2-km²-area sill intruding sediments with 2 wt% organic carbon would have a methane-generation potential of 73 Mt of CH₄. Iyer et al. (2013) predicted the same sill may release ~109 Mt of CH₄ in models that included fluid flow. The 10-fold difference between these model predictions and observations cannot be due solely to the 25% difference between observed and modeled sill thicknesses. Factors that impact sill-driven gas generation include intrusion depth, host-rock temperature, and sediment type. Models of sills intruded at depths ranging from 0.4 to 5.1 km indicate that intrusion depth is the dominant factor controlling methane mobilization (Galerne and Hasenclever, 2019).

Sill-driven organic matter pyrolysis is increasingly limited as intrusion depth shallows above 500 m. Marine sediment porosity commonly decreases with compaction from ~100% tô50% over the upper 500 mbsf, with a trend toward more linear decrease below 500 m (Tenzer and Gladkikh, 2014). The volumetric density of organic carbon increases with compaction, while the specific heat of the sediment decreases. Higher-porosity sediments therefore require more heat to raise their temperature, since the heat capacity of water is large, i.e., ~5 times that of sedimentary grains. The difference in aureole thickness above versus below the sill at Site U1546 and the corresponding factor-of-two difference in carbon release are likely due to the influence of compaction. The stratigraphic correlation indicates that ~36 m of sill thickness was accommodated by compaction of underlying sediments, with a corresponding ~12.5% increase in bulk density and an ~39% decrease in heat capacity of the compacted sediments, which is similar to the ~38% difference in aureole thickness below versus above the sill.

Intrusion depth also dictates host-rock temperature for a given geothermal gradient, with deeper intrusions requiring less heat to warm sediment to hydrocarbon-formation temperatures. The thermal gradient at Site U1546 is large, and the temperatures at thê230 m intrusion depth are similar to those at ~1 + km depth in many continental basins. Therefore, the relative inefficiency of carbon release for the sill at Site U1546 cannot be ascribed to low hostrock temperatures. Similarly, the organic matter within the sediments at Site U1546 consists primarily of type II kerogen (Teske et al., 2021a, 2021b). This would not limit hydrocarbon production given the availability of water to provide hydrogen (Seewald, 2003).

The sill that intruded at Site U1546 appears to have had an impact on carbon cycling within the upper 100 m of sediment, distinct from thermogenic alteration, and it persists to the present day. Pore-water sulfate and core-headspace methane concentrations measured from cores (Fig. 2; see the Supplemental Material) indicate a large difference in the depth of the sulfate/methane transition zone (SMTZ) between Sites U1545 (50 mbsf) and U1546 (110 mbsf). The SMTZ marks the intersection of sulfate and methane concentration profiles and represents the depth where the downward flux of sulfate from the water column is balanced by the upward flux of methane from deeper sediments (Iversen et al., 1985). Headspace gas measurements from cores indicate pore-water peaks reach ~10,000 mM H₂S and ~2.25 mM methane accumulation below the SMTZ (Teske et al., 2021a, 2021b), indicating active methanogenic and sulfate-reducing, methane-oxidizing microbial communities at both sites. The steeper sulfate gradient at Site U1545, however, points to a lower flux of methane at Site U1546 relative to Site U1545.

Compaction-driven advective fluid flux provides a plausible explanation for the difference in SMTZ depths between these sites. Advective flux delivers methane to the SMTZ from below and counters downward diffusion of seawater sulfate (Regnier et al., 2011). Vertically advecting fluids at Site U1546 are limited to fluids that are compacted only out of sediments between the seafloor and the igneous sill due to the low permeability of the sill. These sediments are on average older and thus more compacted than the sediment column at Site U1545. Sills with lateral dimensions that are large with respect to emplacement depth may thus limit the export of thermogenic gas out of the sediment column once the sill has cooled.

Scientific drilling results show that the impact of igneous sill intrusion into young, shallow basins differs from intrusions into older, deeper sedimentary basins. The large water content of young, shallow sediments requires substantial heat to raise sediment temperature to levels required for thermal cracking of kerogen, resulting in less hydrocarbon production for shallower intrusions. In addition, subsidence of a low-permeability igneous sill into compacting underlying sediments may inhibit upward migration of dissolved thermogenic gases. These effects themselves substantially impact the intruded sediments, which are left more compacted and hydrocarbon rich than those prior to intrusion, effectively accelerating the maturation through compaction and heating with burial that is typically observed within deep sedimentary basins unaffected by igneous intrusions.

This research used samples and data provided by the International Ocean Discovery Program (IODP). We sincerely thank the IODP staff and crew of the D/V JOIDES Resolution, and we also thank the two manuscript reviewers for thoughtful comments that improved the paper.