Derrill M. Kerrick

Metamorphic devolatilization of subducted oceanic metabasalts: implications for seismicity, arc magmatism and volatile recycling

Subducted oceanic metabasalts are believed to be a primary source of volatiles for arc magmatism and fluid-induced seismicity. From phase equilibria computed for an average oceanic metabasalt we present a model for subduction zone devolatilization for pressures up to 6 GPa (∼180 km). Along high temperature geotherms complete dehydration occurs under forearcs, whereas dehydration does not occur along low temperature geotherms. For intermediate geotherms, major dehydration occurs under subarcs and provides a subjacent H2O source for arc volcanism. Decarbonation is negligible along cold and intermediate geotherms and limited along high temperature geotherms. Because decarbonation is limited for all subducted carbonate-bearing lithologies, transfer of CO2 from subducted slabs to arc magmas may be triggered by aqueous fluid infiltration. Metabasalt devolatilization could induce seismicity in forearcs (high temperature geotherms) and subarcs (intermediate geotherms); however, because of the lack of devolatilization, metabasalts would not be a fluid source for seismicity with low temperature geotherms. Along low temperature geotherms, limited devolatilization of subducted oceanic metabasalts and marine sediments in forearcs and subarcs provides a mechanism for return of volatiles to the deeper mantle.

Statistical thermodynamic models for ideal oxide and silicate solid solutions, with application to plagioclase

Abstract Activity-composition relations are derived for ideal substitutional solid solutions through the Helmholtz free energy expressed in terms of the partition function. For solutions of the type (A, B) u Z w involving mixing on one type of atom site, ideal activities of end-member components are expressed by: a A u Z w = ( X A u Z w ) u , and a B u Z w = ( X B u Z w ) u . With multi-site mixing excluding charge balance restrictions, as in (A, B) α u (C, D) β v Z w , the ideal activity of an end-member component such as A u C v Z w is calculated as: a A u C v Z w = ( X α A ) u ( X β c ) v . These expressions support the ‘ionic solid solution model for the activities of components in ideal solid solutions. Ideal solution models for coupled substitutions involving charge balance are considered using plagioclase as an example. Ideal activity expressions for solid solution of albite and anorthite are derived with and without adherence to the Al avoidance principle. Mixing models involving local electrostatic balance are contrasted with those involving independent, random mixing of Na-Ca and Al-Si. Of several possible ideal solution models for plagioclase, only that specifying complete Al-Si ordering and local electrostatic neutrality yields activities conforming to Raoults Law.

Quantification of deep CO2 fluxes from Central Italy. Examples of carbon balance for regional aquifers and of soil diffuse degassing

Abstract In Central Italy non-volcanic CO 2 is discharged by focused degassing (strong diffuse emission and vents) and by high-CO 2 groundwater. 3 He / 4 He data and the carbon isotopic composition of CO 2 are compatible with derivation from mantle degassing and/or metamorphic decarbonation. The gases produced at depth accumulate in permeable reservoirs composed of Mesozoic carbonates. When total pressure (roughly corresponding to p CO 2 ) of the reservoir fluid exceeds hydrostatic pressure, a free gas phase forms gas reservoirs within the permeable host rocks from which gases may escape toward the surface. This process generates both the focused vents and the CO 2 -rich springs which characterise the study area. The storage and expulsion of CO 2 is controlled by fractures and faults and/or structural highs of permeable carbonate formations. Influx of deep CO 2 into the overlying groundwater yields a widespread elevated p CO 2 anomaly in the Tyrrhenian Central Italy aquifers. These aquifers release CO 2 to the atmosphere when groundwater is discharged at the surface from springs. The groundwater degassing flux is estimated from the carbon balance of regional aquifers computed by coupling aquifer geochemistry with isotopic and hydrogeological data. The resulting production rate of deep CO 2 ranges from 4×10 5 to 9×10 6 mol y −1 km −2 . In concert with the regional geologic setting, the deep CO 2 production rate increases westward. In the aquifers with anomalously high p CO 2 , the average CO 2 influx rate of the anomalous areas is several times higher than the value derived by Kerrick et al. [Kerrick, D.M., McKibben, M.A., Seward, T.M., Caldeira, K., 1995. Convective hydrothermal CO 2 emission from high heat flow regions. Chem. Geol., 121 (1995) 285–293.] as baseline for CO 2 emission from areas of high heat flow. The flux of CO 2 lost to the atmosphere from water emitted from springs is of the same order of magnitude as the influx of deep CO 2 into the aquifer.

Present and past nonanthropogenic CO2 degassing from the solid earth

Global carbon cycle models suggest that CO2 degassing from the solid Earth has been a primary control of paleoatmospheric CO2 contents and through the greenhouse effect, of global paleotemperatures. Because such models utilize simplified and indirect assumptions about CO2 degassing, improved quantification is warranted. Present-day CO2 degassing provides a baseline for modeling the global carbon cycle and provides insight into the geologic regimes of paleodegassing. Mid-ocean ridges (MORs) discharge 1–3 × 1012 mol/yr of CO2 and consume ∼3.5 × 1012 mol/yr of CO2 by carbonate formation in MOR hydrothermal systems. Excluding MORs as a net source of CO2 to the atmosphere, the total CO2 discharge from subaerial volcanism is estimated at ∼2.0–2.5 × 1012 mol/yr. Because this flux is lower than estimates of the global consumption of atmospheric CO2 by subaerial silicate weathering, other CO2 sources are required to balance the global carbon cycle. Nonvolcanic CO2 degassing (i.e., emission not from the craters or flanks of volcanos), which is prevalent in high heat flow regimes that are primarily located at plate boundaries, could contribute the additional CO2 that is apparently necessary to balance the global carbon cycle. Oxidation of methane emitted from serpentinization of ultramafics and from thermocatalysis of organic matter provides an additional, albeit unquantified, source of CO2 to the atmosphere. Magmatic CO2 degassing was probably a major contributor to global warming during the Cretaceous. Metamorphic CO2 degassing from regimes of shallow, pluton-related low-pressure regional metamorphism may have significantly contributed to global warming during the Cretaceous and Paleocene/Eocene. CO2 degassing associated with continental rifting of Pangaea may have contributed to the global warming that was initiated in the Jurassic. During the Cretaceous, global warming initiated by CO2 degassing of flood basalts, and consequent rapid release of large quantities of CH4 by decomposition of gas hydrates (clathrates), could have caused widespread extinctions of organisms.

Methane: An equation of state with application to the ternary system H2O-CO2-CH4

Abstract The following hardsphere modified Redlich-Kwong (HSMRK) equation of state was obtained by least squares fitting to available P-V-T data for methane (P in bars; T in Kelvins; v in cm3 mol−1; b = 60.00 cm3 mol−1; R = 83.14 cm3barmol−1K−1): P RT(1 + y + y 2 −y 3 v(1−y) 3 ) - c(T) + d(T) v + e(T) v 2 /v(v + b)T 1 2 y = b 4v c(T) = 13.403 × 106 + (9.28 × 104)T + 2.7 T2d(T) = 5.216 × 109 − (6.8 × 106)T + (3.28 × 103)T2e(T) = (−2.3322 × 1011) + (6.738 × 108)T + (3.179 × 105)T2 For the P-T range of experimental data used in the fit (50 to 8600 bars and from 320 to 670 K), calculated volumes and fugacity coefficients for CH4 relative to experimentally determined volumes and fugacity coefficients have average percent deviations of 0.279 and 1.373, respectively. The HSMRK equation, which predicts linear isochores over a wide P-T range, should yield reasonable estimates of fugacity coefficients for CH4 to pressures and temperatures well outside the P-T range of available P-V-T data. Calculations for the system H2O-CO2-CH4, using the HSMRK equations for H2O and CO2 of Kerrick and Jacobs (1981) and the HSMRK equation for CH4 of this study, indicate that compared to the binary H2O-CO2 system, small amounts of CH4 in the ternary system H2O-CO2-CH4 slightly increases the activity of H2O, and significantly decreases the activity of CO2.

Modeling open system metamorphic decarbonation of subducting slabs

Fluids derived from the devolatilization of subducting slabs play a critical role in the melting of the mantle wedge and global geochemical cycles. However, in spite of evidence for the existence and mobility of an aqueous fluid phase during subduction metamorphism, the effect of this fluid on decarbonation reactions in subducting lithologies remains largely unquantified. In this study we present results from thermodynamic modeling of metamorphic devolatilization of subducted lithologies for pressures up to 6 GPa using an approach which considers fluid fractionation from source lithologies and infiltration from subjacent lithologies. This open system approach in which fluid flow is an intrinsic component of the chemical model offers an alternative to closed system models of subduction zone decarbonation. In general, our models simulating pervasive fluid flow in subducting lithologies predict CO2 fluxes measured from volcanic arcs more closely than models which assume purely channelized flow. Despite the enhanced effect of H2O-rich fluid infiltration on subduction decarbonation, our results support the hypothesis that CO2 is returned to the deep mantle at convergent margins, particularly in cool and intermediate subduction zones. Our results demonstrate that for most subduction zones, a significant proportion of the CO2 derived from the slab is lost beneath the fore arc, and therefore CO2 flux estimates based on measurements within the volcanic arc alone may significantly underestimate the slab-derived CO2 flux for individual margins. Nevertheless, our predicted global slab-derived CO2 flux from convergent margins of 0.35–3.12 × 1012 mols CO2/yr is in good agreement with previous estimates of global arc volcanic flux. Because our predicted global slab-derived CO2 flux is significantly less than atmospheric CO2 drawdown by chemical weathering, significant CO2 emission from other geologic regimes (e.g., hot spots) would be required to balance the global carbon cycle.

Dynamics of carbon dioxide emission at Mammoth Mountain, California

Mammoth Mountain, a dormant volcano in the eastern Sierra Nevada, California, has been passively degassing large quantities of cold magmatic CO2 since 1990 following a 6-month-long earthquake swarm associated with a shallow magmatic intrusion in 1989. A search for any link between gas discharge and volcanic hazard at this popular recreation area led us to initiate a detailed study of the degassing process in 1997. Our continuous monitoring results elucidate some of the physical controls that influence dynamics in flank CO2 degassing at this volcano. High coherence between variations in CO2 efflux and variations in atmospheric pressure and wind speed imply that meteorological parameters account for much, if not all of the variability in CO2 efflux rates. Our results help explain differences among previously published estimates of CO2 efflux at Mammoth Mountain and indicate that the long-term (annual) CO2 degassing rate has in fact remained constant since ∼1997. Discounting the possibility of large meteorologically driven temporal variations in gas efflux at other volcanoes may result in spurious interpretations of transients that do not reflect actual geologic processes.

Metamorphic controls on seismic velocity of subducted oceanic crust at 100–250 km depth

Abstract Most circum-Pacific subduction zones at 100–250 km depth contain layers in which seismic velocities are ca. 5% slower than in the adjacent mantle. We compute seismic velocities from thermodynamic data for equilibrium metabasalt mineralogies, determined by free energy minimization, at subduction zone conditions. Lawsonite stability has a profound effect on seismic velocities of subducted oceanic metabasalts. Velocity reductions of 3–7% are estimated for lawsonite–eclogites derived by metamorphism of hydrothermally altered oceanic basalt subducted along relatively cool geotherms, whereas a 2–4% velocity increase is characteristic of anhydrous eclogites within the coesite stability field. The restricted depth extent of low-velocity layers is explicable through the influence of the coesite–stishovite transition, which reduces lawsonite stability at high pressure. This transition also increases the positive velocity anomaly in anhydrous eclogites to 4–6%, an effect that may account for deep high-velocity layers. The quality of the match between the properties of lawsonite–eclogite and low-velocity layers supports the contention that significant quantities of volatiles are retained within the oceanic crust beyond sub-arc depths. Because the velocity anomalies are explicable in terms of equilibrium phase relations, we find no reason to invoke metastability of metamorphic reactions to explain the low-velocity layers.

Subduction of ophicarbonates and recycling of CO2 and H2O

Because subducted serpentinites may release significant quantities of volatiles, high-pressure phase equilibria were computed for two end-member hydrothermally altered mantle harzburgite protoliths (ophicarbonates): calcite + antigorite + brucite and calcite + antigorite + talc. For both bulk compositions, most of the H 2 O released by metamorphic dehydration occurs at subarc depths; thus dehydration of serpentinites could be a major source for H 2 O in arc magmas. In contrast, for both model compositions a significant fraction of the original carbonate is retained to depths exceeding 200 km. Consequently, deep subduction of ophicarbonate rocks of the oceanic lithosphere and/or downward drag of mantle wedge ophicarbonates provide a mechanism for carbonating the mantle and thus a potentially significant CO 2 source for deep mantle melts. The probable CO 2 sources for arc magmas are metamorphic decarbonation of marine sediments and/or carbonated mafic volcanics in the subducted slab.

Geochemistry of Quaternary travertines in the region north of Rome (Italy): structural, hydrologic and paleoclimatic implications

In the Tyrrhenian region of central Italy, late Quaternary fossil travertines are widespread along two major regional structures: the Tiber Valley and the Ancona–Anzio line. The origin and transport of spring waters from which travertines precipitate are elucidated by chemical and isotopic studies of the travertines and associated thermal springs and gas vents. There are consistent differences in the geochemical and isotopic signatures of thermal spring waters, gas vents and present and fossil travertines between east and west of the Tiber Valley. West of the Tiber Valley, δ13C of CO2 discharged from gas vents and δ13C of fossil travertines are higher than those to the east. To the west the travertines have higher strontium contents, and gases emitted from vents have higher 3He/4He ratios and lower N2 contents, than to the east. Fossil travertines to the west have characteristics typical of thermogene (thermal spring) origin, whereas those to the east have meteogene (low-temperature) characteristics (including abundant plant casts and organic impurities). The regional geochemical differences in travertines and fluid compositions across the Tiber Valley are interpreted with a model of regional fluid flow. The regional Mesozoic limestone aquifer is recharged in the main axis of the Apennine chain, and the groundwater flows westward and is discharged at springs. The travertine-precipitating waters east of the Tiber Valley have shallower flow paths than those to the west. Because of the comparatively short fluid flow paths and low (normal) heat flow, the groundwaters to the east of the Tiber Valley are cold and have CO2 isotopic signatures, indicating a significant biogenic contribution acquired from soils in the recharge area and limited deeply derived CO2. In contrast, spring waters west of the Tiber Valley have been conductively heated during transit in these high heat-flow areas and have incorporated a comparatively large quantity of CO2 derived from decarbonation of limestone. The elevated strontium content of the thermal spring water west of the Tiber Valley is attributed to deep circulation and dissolution of a Triassic evaporite unit that is stratigraphically beneath the Mesozoic limestone. U-series age dates of fossil travertines indicate three main periods of travertine formation (ka): 220–240, 120–140 and 60–70. Based on the regional flow model correlating travertine deposition at thermal springs and precipitation in the recharge area, we suggest that pluvial activity was enhanced during these periods. Our study suggests that travertines preserve a valuable record of paleofluid composition and paleoprecipitation and are thus useful for reconstructing paleohydrology and paleoclimate.