Abstracts and Schedule: DTM Science at 2018 AGU Fall Meeting

The 2018 AGU Fall Meeting at the Walter E. Washington Convention Center, Dec. 10, 2018. Photo: Roberto Molar Candanosa, DTM.
Monday, December 10, 2018 

Helen Janiszewski: V13D-0120: Receiver function imaging of magmatic- and subduction-related structures beneath arc volcanoes: A case study at Cleveland Volcano, Alaska | Monday, 10 December 2018 | 13:40 - 18:00

Cleveland Volcano is one of the most active volcanoes in the central-Aleutian island arc and represents one of the shallow-end members of depth to subducting slab at only ~ 65 km below the volcano. It was the site of a temporary deployment of twelve broadband seismometers from August 2015 – July 2016. We calculate P-to-s receiver functions using these as well as two permanent Alaska Volcano Observatory stations, which have operated since 2014, to determine new seismic constraints on the crustal structure beneath Cleveland Volcano and the depth to the subducting crust. Receiver functions are sensitive to boundaries in seismic velocity properties, and are typically used to image discontinuities such as the Moho or interfaces associated with subducting crust. They are also sensitive to crustal magmatic reservoirs and are particularly useful for imaging these features in the mid- to lower-crust, although the application and development of this technique for such systems has been limited. At Cleveland Volcano we image a clear P-to-s conversion from the Moho discontinuity. However, its arrival time relative to the initial P-wave typically varies up to two seconds at a given station with later arrivals systematically corresponding to ray paths that have passed through crust directly beneath the volcano. If we assume this variation is solely due to changes in the Moho depth, we predict that Moho topography beneath the volcano varies on the order of 20 km over horizontal distances of < 5 km. A more likely explanation is that slow shear-wave velocities associated with the magmatic system beneath Cleveland Volcano contribute to the travel time variability, with perhaps a secondary contribution from smaller variations in Moho topography. The density of the seismic array yields crossing ray paths for the P-to-s conversions, allowing us to investigate the potential for a quasi-local tomographic shear velocity inversion beneath the volcano. Finally, while investigating slab depth at these stations requires crustal velocity constraints that account for the volcanic structures, we do observe arrivals that are consistent with a shallow subducting slab beneath Cleveland Volcano using a single station that does not have ray paths that pass through the crust beneath the volcano.


Peter Driscoll: GP21C-0673: Paleomagnetic Biases Inferred from Numerical Dynamos and the Search for Geodynamo Evolution | Tuesday, 11 December 2018 |08:00 – 12:20

The paleomagnetic record is central to our understanding of the history of the Earth. The orientation and intensity of magnetic minerals preserved in ancient rocks indicate the geodynamo has been alive since at least the Archean and possibly the Hadean. A paleomagnetic signature of the solidification of the inner core, arguably the singular most important event in core history, however, has remained elusive. Why? In pursuit of this signature we investigate a suite of numerical dynamo simulations from a paleomagnetic perspective to explore how long the field should be time-averaged to obtain a Geocentric Axial Dipole (GAD) field, stable virtual paleomagnetic directions, and accurate virtual global field intensities. Generally we find obtaining accurate intensities requires longer time averaging than obtaining accurate directions. When polarity reversal frequency is low running averages of local field intensity and inclination produce reliable estimates of the underlying dipole moment independent of site latitude. However, when heat flux and reversal frequency are increased (>2 per Myr) we find that local observations tend to underestimate virtual dipole moment (VDM) by up to 50% and overestimate virtual axial dipole moment (VADM) by up to 150%. These biases in virtual dipole moment estimates can persist over more than 500 kyr. A latitudinal dependence is found where VDM underestimates the true dipole moment more at low latitudes, while VADM overestimates the true axial dipole moment more at high latitudes, and the most accurate estimates occur at mid-latitudes. The cause for these biases appears to be a contamination of the time averaged field by non-GAD terms, which grows with reversal frequency. We derive a scaling law connecting reversal frequency and site paleolatitude to paleointensity bias (ratio of observed to the true value). In application to the PINT paleointensity record we find these biases produce little change to the overall trend of a relatively flat but scattered intensity over the Precambrian. A more careful intensity adjustment applied during periods when the reversal frequency is known could reveal previously obscured features in the paleointensity record.


Doug Hemingway: GP21C-0666: Lunar swirl morphology constrains the geometry, magnetization, and origins of lunar magnetic anomalies | Tuesday, 11 December 2018 | 08:00 - 12:20

Lunar swirls are collections of finely structured bright and dark surface markings, alternating over length scales of typically 1–5 km. If swirls are the result of plasma interactions with crustal magnetic anomalies, or electrostatic or magnetic sorting of fine materials, the magnetic field orientation must vary over similar length scales. This requires that the associated source bodies be both shallow and narrow in horizontal extent. The correspondingly restricted volume of the source bodies in turn implies strong rock magnetization. Here we show that if ~300 nT surface fields are necessary to produce observable swirl markings, the required rock magnetization must be >0.5 A/m, even for very shallow sources, and likely closer to ~2 A/m or more. This strong source rock magnetization, together with the geometric constraints that favor magmatic structures such as dikes or lava tubes, requires a mechanism to enhance the magnetic carrying capacity of the rocks. We propose that heating associated with magmatic activity could thermochemically alter host rocks and impart them with magnetizations an order of magnitude stronger than is typical of lunar mare basalts. Our results both place constraints on the geometry and magnetization of the source bodies, and provide clues about the possible origins of the Moon's crustal magnetic anomalies.


Anais Bardyn: P23G-3520: Global Composition of Dust at Comet 67P/Churyumov-Gerasimenko as Measured by the COSIMA/Rosetta Mass Spectrometer | Tuesday, 11 December 2018 | 13:40 - 18:00

COSIMA was the time-of-flight secondary ion mass spectrometer on board the Rosetta orbiter. From August 2014 to September 2016 the instrument performed in situ analyses of the dust particles ejected from comet 67P/Churyumov-Gerasimenko, before and after perihelion. COSIMA collected more than 35,000 particles at a low impact velocity (< 10 m/s) and analyzed about 250 of them with size ranging from ~50 to ~1000 µm. We will report the global composition of the cometary dust as deduced from COSIMA measurements. The average elemental composition measured for 67P's dust will be compared to previous results obtained from the Giotto and Vega missions for comet 1P/Halley and the Stardust mission for comet 81P/Wild 2, to the composition of Chondritic Porous Interplanetary Dust Particles (CP-IDPs) and to the CI chondrite composition.


Kathleen McKee: V23K-0187: Explosion volume flux comparison using seismically derived tilt, infrasound, and gas data at Stromboli Volcano, Italy | Tuesday, 11 December 2018 | 13:40 - 18:00

Basaltic volcanoes are characterized by low-level explosions and lava fountaining, but are capable of violent subplinian to plinian eruptions. Stromboli Volcano has been observed with seismometers and tiltmeters to deform prior to explosion [Genco and Ripepe, 2010]. Quantitatively linking this precursory deformation, often interpreted as inflation due to an influx of magma and gas, with the volume of subsequent erupted material (gas and tephra) will be helpful for understanding eruption dynamics and in hazard mitigation, as we can better constrain expected volatile output from monitored volume input. To these aims, we temporarily deployed 7 seismometers, 7 infrasonic microphones, an FTIR, FLIR, gravimeter, and MultiGAS at Stromboli Volcano, Italy in May 2018. We use these data in combination with data from permanently installed seismometers, infrasound sensors, tiltmeters, and UV cameras to examine the volatile budget of Stromboli by comparing shallow volume input derived from broadband, tilt-affected seismic and tilt data with volume output from infrasound and gas data. First, we characterize and locate the seismically-derived tilt source using a nonlinear moment tensor inversion method [Tape and Tape, 2012; Waite and Lanza, 2016] and use the result to quantify the volatile volume input. We will attempt to characterize the input volume density using gravity data. Independently, we characterize and locate the infrasonic explosion source and quantify the output volume using an infrasonic waveform inversion technique that accounts for the influence of topography by computing numerical Green's functions by way of a 3-D finite difference time domain method [Kim et al., 2015]. The seismically estimated volume input and infrasonically estimated volume output will then be validated with volume output estimated from gas observations. This quantitative examination of tilt, infrasound and gas data aims to advance our ability to determine the size of eruption prior to its occurrence.


Hélène Le Mével: V23K-0194: Analysis of Seismo-Gravity Signals Associated with Volcanic Explosions Recorded at Stromboli, Italy in May 2018 | Tuesday, 11 December 2018 | 13:40 - 18:00

In May 2018, we temporarily deployed 7 seismometers, 7 infrasonic microphones, an FTIR, FLIR, and MultiGAS station at Stromboli Volcano, Italy. Simultaneously, we conducted a continuous gravity experiment with the goal of characterizing the source of inflation-deflation-inflation cycles observed in tilt time series for stations located within 1 km of Stromboli's summit vents. A Scintrex CG-6 gravimeter recorded gravity continuously at a sampling rate of 10 Hz at three different locations around the summit crater. For each gravity site we use the nearest broadband seismometer to estimate the inertial acceleration component of the gravity signal due to ground motion. In addition, ambient temperature, barometric pressure, and wind speed were recorded locally to identify any correlation with the gravity changes. After correcting for the long term drift and Earth tides, the residual gravity variations are analyzed in the frequency domain using the Fourier and Continuous Wavelet Transforms. We interpret the residual gravity variations in terms of volumetric changes associated with mass transport in the upper conduit. Data from the seismo-acoustic network provide the timing and duration of each explosive event. By isolating the gravity signature of the Strombolian explosions, we are able to study mass transport in the upper volcanic conduit on the timescale of ~500 seconds. Finally, the gravity-estimated volumes of gas input to the system can be compared to the volatile budget estimated from gas measurements and from the seismically-derived tilt source. This gravity experiment represents an important step towards the identification of precursory gravity changes prior to volcanic explosive events and a test of the utility of continuous gravity in real-time volcano monitoring.


Elodie Brothelande: V23G-0143: Damage modeling at Piton de la Fournaise: linking seismicity and deformation patterns leading to eruptions | Tuesday, 11 December 2018 | 13:40 - 18:00

Seismicity rates and surface deformation records have just started to be used together in large-scale damage models of volcanic edifices. In these models, daily seismicity (VT events) can be interpreted as an indicator of the progress of damage, which plays a critical role in determining the mechanical behavior of the volcano. Seismicity and deformation can therefore be linked through the time-dependent evolution of the edifice elastic parameters. At various basaltic volcanoes, the analysis of seismic and GPS time-series of inter and pre-eruptive phases reveals a change in the volcano mechanical behavior a few weeks to months prior to eruptions, leading to an acceleration in seismic and deformation rates before the eruption. Piton de la Fournaise volcano (Reunion Island) is one of the most active hotspot volcanoes, with very frequent eruptions over the last three decades, and it is also equipped with one of the best monitoring network. Taking advantage of this exceptional record, we performed a systematic study of inter-eruptive time-series at Piton de la Fournaise since 2005 using damage models in order to identify pre-eruptive patterns, understand the changes in the volcano mechanical behavior, determine whether the repetitive characteristics of the signal could be interpreted as precursors, and characterize the long-term evolution of the reservoir dynamics.


Brad Peters: V31D-0146: A framework for efficient chemical separation of tungsten and other trace elements using organic acids for high-precision mass spectrometry | Wednesday, 12 December 2018 | 08:00 - 12:20

High-precision measurements of isotopic compositions by mass spectrometry require sample solutions that are highly pure with respect to the element of interest. In order to achieve such high elemental purity, often for elements that are present in trace to ultratrace amounts in basaltic samples, chemical separation protocols are employed to separate major (e.g., Fe, Mg, Al, Ca, Ti) and trace elements while retaining the selected analyte. These protocols traditionally employ acidic solutions that typically include a suite of strong acids and hydrofluoric acid. Some other common-practice protocols utilize organic acids. Organic acids are particularly useful for separating elements with similar chemical behaviors on standard cation and anion resins. For example, separation of Ti from W in large (>1 gram) basaltic samples has traditionally employed acetic acid, which preferentially forms complexes with W that are retained on anion resins. We improve the efficiency of this separation protocol by instead using citric acid, which has highly disparate complexation behavior for Ti and W. The protocol is capable of producing a high W yield (approaching 100%) and can reduce the Ti/W ratio of basaltic samples from ~105 to <1 following a standard cation separation protocol. We then examine the chemical behaviors that enable such efficient separation of trace elements using organic acids, including pH, redox state, ligand availability, and sample size. Given the large range of organic acids available, careful selection of acid-analyte pairs and eluant formulae may enable a variety of high-efficiency separation techniques with minimal adaptation of our framework method. We illustrate this principle by adapting our citric acid protocol for high-efficiency separation of Mo and Sn (both typically with 90-100% yields), and further explore using salicylic acid for high-efficiency separation of other ultratrace metallic elements. These applications of our framework method reveal that organic acids are a currently underutilized tool for chemical separation protocols, particularly those protocols designed for use with large sample masses. Further, since many organic acids are inert and pose relatively little health risk compared to strong acids, our framework may be particularly for mitigating laboratory safety concerns.


Tim Jones: DI33A-08: Tungsten Isotopes in Mantle Plumes: Heads it's Positive, Tails it's Negative (Invited) | Wednesday, 12 December 2018 |15:25 - 15:40

The lowermost mantle is driven to Earth's surface by mantle plumes, providing a volcanic record of its structure and composition. Plumes consist of a head and tail, which melt to form large igneous provinces (LIPs) and ocean island basalts (OIBs), respectively. Recent analyses have shown that such sites of volcanism exhibit tungsten (W) isotope heterogeneity that was created in the first ~60 million years of our solar system's evolution. Moreover, the isotopic signature found in LIPs differs to that found in OIBs, revealing that the melt products of plume heads must be dominated by a different ancient mantle reservoir to that of plume tails. However, existing geodynamical studies show that plume heads and tails sample the same deep-mantle source region, and, therefore, cannot account for any systematic differences in composition. Here we present a suite of numerical simulations of thermo-chemical plumes and an isotopic model for W sources in the mantle. Our results show that the W isotope systematics of LIPs and OIBs can, under certain conditions, arise as a dynamical consequence of plumes forming in a heterogeneous, thermo-chemical boundary layer. We also show that ultra low-velocity zones (ULVZs), which sit on the core-mantle boundary (CMB), likely contribute to the chemical diversity observed in OIBs but not LIPs. This study places geochemical observations from Earth's surface in a geodynamically consistent framework and illuminates their relationship with seismically imaged features of the deep mantle.


Stephen Golden: IN41C-38: Digitization of Carnegie Analog Broadband Seismograph Tapes | Thursday, 13 December 2018 | 08:54 - 08:57

Between 1965 and 2003, the Carnegie Institution of Washington's Department of Terrestrial Magnetism has operated a continuous network of 9 broadband seismographs spread over the world. Data from these instruments was recorded on analog magnetic tapes, to be read and analyzed using a custom playback system. The tapes were stored in a shielded room, which kept them in reasonably good condition until today. The tapes are currently undergoing a digitization effort with the goal of making this dataset publicly available through archival in SEED format at the Incorporated Research Institutions for Seismology Data Management Center (IRIS DMC). The seismographs, designed by Selwyn Sacks, were among the earliest broadband instruments, sensing between >100 seconds and ~30 Hz. To preserve the instruments' high dynamic range with minimum distortion, each component was simultaneously recorded onto 5 tape tracks, using different amplifier and filter settings. Station amplitudes were calibrated at regular intervals. A time code was continuously recorded on a separate tape track. Three stations were installed in the Andes, three in Japan, one in Papua New Guinea, one in Iceland, and one in Washington DC. All but two stations were three component installations. Early data from this network, especially from the 1960s and 70s, may still have significant scientific value due to the rareness of continuous broadband recordings from that period. Although absolute time accuracy does not hold up to today's standards, the dataset appears suitable for single station analysis methods such as shear wave splitting or receiver functions. Digitization required custom hardware and software to be restored or newly developed. After initial digitization, each tape's time track is decoded and the waveform data integrated and resampled to correct for tape speed variations. The output is reviewed for data quality and coverage, and the results compiled alongside scanned field notes into a meta-data catalog. As of late 2018, this data preservation effort is still ongoing. This presentation provides an overview of the current project status, in the hope of receiving feedback on possible improvements before making the archive public, while also sharing information about the availability of these data with the broader scientific community.


Diana Roman: V43J-0285: Modulation of seismic activity in Kilauea's East Rift Zone by summit inflation and deflation | Thursday, 13 December 2018 | 13:40 - 18:00

Kilauea Volcano's East Rift Zone (ERZ) has been the site of magma transport and eruption since 1983. Thus, an understanding of the controls on ERZ seismicity is key to interpreting changes in seismicity in an eruption forecasting context. We analyzed earthquake and deformation (multi-temporal InSAR and GPS) data during May 25-November 5, 2007. This period includes a major dike intrusion and a small fissure eruption in the ERZ (the Father's Day eruption), a small dike intrusion between Pu u O o and Kupaianaha vents, a slow-slip southern flank event, multiple inflation and deflation of Kilauea's summit, and variable rates of ERZ seismicity. We calculated double-couple fault-plane solutions (FPS) for all earthquakes located in the ERZ during this period. FPS indicate that ERZ seismicity involves strike-slip faulting, with P-axes oriented East-West, throughout the study period and area. Modeled GPS geodetic data indicate deflation of a shallow reservoir at Halemaumau from May 25 through June 16 and June 20 through 26, periods characterized by relatively low ERZ seismicity rates (~3 located earthquakes/day). The Halemaumau reservoir then inflated between June 27-July 21, accompanied by relatively high rates of ERZ seismicity (~ 8 located earthquakes per day). From July 21 through the end of the study period, deformation observed by GPS and ENVISAT MT-InSAR geodetic data was characterized by deflation of a reservoir located beneath the south caldera, and ERZ seismicity was minimal (~0.2 located earthquakes/day) although the Puu Oo vent continued to erupt. We model Coulomb stress changes induced by modeled reservoir inflation/deflation on NW-striking left-lateral faults and find that summit reservoir inflation promotes ERZ faulting with the observed sense of slip, and deflation of the modeled summit reservoir impedes ERZ faulting. Thus, rates of seismicity in the ERZ appear to be strongly modulated by stresses arising from deformation of Kilauea's summit reservoir. This result indicates that, at volcanoes with complex and laterally extensive plumbing systems, it is important to consider the influence of distal deformation sources on changes in seismicity rates throughout the volcanic system.


Shi Joyce Sim: DI43C-0045: Melt focusing at mid ocean ridges | Thursday, 13 December 2018 | 13:40 - 18:00

Magmatism at mid-ocean ridges generates new oceanic crust and accounts for 90% of global volcanism. The oceanic crust is emplaced in a narrow neo-volcanic region, whereas basaltic melt is generated in a wide region beneath mid-ocean ridges as suggested by geophysical surveys. Melt focusing mechanisms, such as ridge suction and decompaction layers, have been proposed. We present results from a suite of two-phase models applied to the mid-ocean ridges with varying half spreading rate using a new open source model, Melt in the Mantle beneath Mid-ocean ridges (M3LT). M3LT is built using TerraFERMA, the Transparent Finite Element Rapid Model Assembler. M3LT solves for the magma migration (McKenzie, 1984) equations with thermal evolution, non-linear rheology and melting that depends on pressure and temperature. The amount of oceanic crust generated by the M3LT models are consistent with observations from active seismic surveys. There are three melt focusing mechanisms present in our models: ridge suction, decompaction layers and melting pressure focusing. The dominant melt focusing mechanism determines how the majority of melt moves before reaching the ridge axis and has implications for geophysical and geochemical observations. We assume a bulk viscosity of form Rζη/φ where Rζ is the bulk viscosity constant, η is the mantle shear viscosity and φ is the porosity. For a reasonable shear viscosity ( η = 1018-1019 Pa s) and small porosities due to melt extraction, bulk viscosity constant Rζ (1-20) can change melt focusing. For a lower bulk viscosity, decompaction layers are dominant while for a larger bulk viscosity, melting pressure focusing is dominant. The latter is favored since melt rich decompaction layers have not been observed by geophysical methods whereas the melting pressure mechanism focuses melt in regions where melt is observed below the ridge axis. Further analyses suggest that the focusing mechanisms vary with spreading rate and that bulk viscosity should be on the higher end of the established range or a new formulation might be needed.


Jesse Reimink: V43A-04: Pushing beyond current limits on Nd-isotope ratio measurement precision | Thursday, 13 December 2018 | 14:25 - 14:40

The long-lived 147Sm-143Nd isotope system has been a widely used geochronometer and geochemical tool for several decades. Recently, analytical developments have allowed for exploration of the utility of the geochemical twin, the short-lived 146Sm-142Nd system, where the parent isotope, 146Sm has a half-life of ~100 Ma. Analysis of this short-lived system has proven a valuable tool to address various problems in geo/cosmochemistry, including the timing of crust formation on the Earth and Moon, meteorite differentiation mechanisms, entrainment of early-formed reservoirs in the deep mantle, and the nucleosynthetic building blocks of the terrestrial planets. Natural variation in 142Nd/144Nd, however, is in the fifth decimal place, or on the order of ~40 parts per million, making precise and accurate measurements a complex, but essential, task. Current analytical constraints, both in TIMS and ICPMS analyses, limit the internal precision of any one 142Nd/144Nd analysis to a few parts per million. This severely limits the utility of the short-lived Sm-Nd system for detailed interrogation of the earliest events in our Solar System. Here we review the factors that are currently limiting precision in our double-filament TIMS measurements (e.g. Garcon et al., Chem. Geol. 2018), as well as possible paths forward. We will highlight recent work on a so-called cavity-ion source that has the potential to increase ion production in a thermal ionization mass spectrometer. Using both a modified, DTM-built, single-collector TIMS instrument, as well as an Oak Ridge National Lab multi-collector Triton mass spectrometer modified to operate in cavity mode, we are testing cavity designs and materials to document the utility of cavity ion sources for high precision Nd-isotope measurements. We will discuss the possibilities of such an instrument to produce a new level of precision in Nd-isotope measurements, with the goal of pushing precision in to the sub-ppm range.


My E.I. Riebe: P52B-02: D/H in Photochemically Produced Organic Dust Analogs: Friday, 14 December 2018 | 10:35 - 10:50

Organic matter in meteorites is enriched in D compared to the Sun. In the most primitive meteorites, organics have bulk δD values of up a few 1,000‰ and contain sub-µm- to µm-sized anomalous regions with δD values up to a few 10,000‰. Relatively complex organic molecules can be made in the laboratory by UV irradiation of astrophysically relevant ices under vacuum at temperatures of ~10‒15 K [1]. In space, this process could have occurred in ice mantles on mineral grains in interstellar clouds and/or in the protosolar nebula. Organic materials that formed from ices may have been irradiated further in the protosolar nebula, resulting in alteration of their structures, chemical compositions, and/or isotopic compositions [2,3]. We simulate such additional irradiation by exposing photoproducts made in the laboratory to X-rays. Samples were produced at the Astrochemistry Laboratory of NASA Ames. Starting ice mixtures of H2O, CH3OH, CO, and NH3, with or without addition of naphthalene (C10H8), were irradiated with UV, heated up to room temperature in vacuum, and in some cases the resulting organic residues were further irradiated with different doses of X-rays at the National Synchrotron Radiation Research Center (NSRRC) in Taiwan. Hydrogen isotopic compositions were measured with the Carnegie Cameca NanoSIMS 50L. In most samples, the photoproducts have H that is enriched in D, with δD up to ~1400‰. This shows that photochemistry does not only have the potential to "pass on" D enrichment from isotopically heavy molecules in interstellar clouds to their photoproducts [4], but that the chemistry itself also results in increased D/H ratios. The effects of X-ray irradiation are complex, with samples becoming both enriched and depleted in D compared to the composition prior to X-ray irradiation. [1] Bernstein, M. P. et al. (1995). ApJ, 454: 327-344. [2] De Gregorio B.T. et al. (2010) GCA, 74: 4454-4470. [3] Ciesla F.J., and Sandford S.A. (2012) Science, 336: 452-454. [4] Sandford, S. A. et al. (2000) ApJ, 538: 691-697.


Jonathan Tucker: V52A-04: A high carbon content of the Hawaiian mantle from olivine-hosted melt inclusions | Friday, 14 December 2018 | 11:05 - 11:20

The deep mantle carbon content and flux represent significant uncertainties in global volatile cycles and distributions. Here, we present CO2 concentrations measured in 437 primitive olivine-hosted melt inclusions from Hualalai, Kilauea, Koolau, Loihi, and Mauna Loa to constrain the Hawaiian mantle CO2 content and flux. Quantification of melt inclusion CO2 is complicated by the ubiquitous presence of vapor or "shrinkage" bubbles. We measure exsolved shrinkage bubble CO2 by a volumetric method, and add that to dissolved CO2 in order to reconstruct total melt inclusion CO2 concentrations. We observe that cryptic solid carbon-bearing phases precipitate from the CO2 vapor onto bubble walls, sequestering C from bubbles. Consequently, direct measurement of bubble CO2 by Raman spectroscopy can severely underestimate the total inclusion CO2. We also demonstrate that addition of vapor bubble CO2 to dissolved CO2 results in ~50% uncertainty in total melt inclusion CO2 concentrations, primarily due to uncertainty in 3-D melt inclusion geometry. This same uncertainty is also applicable to Raman studies. Based on our data set, we estimate that parental melts from five Hawaiian volcanoes have between 0.33 and 0.99 wt% CO2, implying a flux of 3.2 Tg CO2/year. Mantle sources for the volcanoes have between 200 and 430 ppm CO2, with an average of 320 ± 50 ppm CO2, suggesting that the Hawaiian plume is significantly more C-rich than the MORB mantle, by at least a factor of ~3. The high CO2 concentration of the Hawaiian plume could originate from deeply subducted surficial C, retention of juvenile C, or both.


Larry Nittler: P52A-05: Disk Processes Diversify Planetary Compositions: Lessons from Mercury for Exoplanetary Science | Friday, 14 December 2018 | 11:20 - 11:35

Two types of data commonly used to constrain the compositions of exoplanets are the bulk composition of host stars and mass-radius relations (when both properties can be determined). However, the former rests on assumptions that planets are not substantially elementally fractionated compared to their host stars and the latter depends on equations-of-state calculated with specific assumptions of redox conditions. The diversity of planetary compositions in our solar system belie these assumptions and demonstrate the importance of protoplanetary disk (PPD) processes in setting compositions and internal structures of planets. An extreme example is given by Mercury, which has long been known to have an anomalously high density indicating that its metallic core makes up an anomalously large fraction of its mass. Published data from the MESSENGER spacecraft revealed additional geochemical surprises. For example, Mercury has high abundances of the moderately volatile elements Na, S, Cl, and K. The high S together with very low surface Fe abundance indicates that Mercury formed under highly reducing conditions (fO2 ~IW-7 to IW-3, cf. ~IW+2 for Earth). This has important implications for bulk composition, because at low O fugacity, Si becomes much more siderophile and large amounts of Si may be present in Mercury's core. Therefore, Mercury's bulk Fe/Si ratio is estimated to lie between ~2-17; the large range reflects uncertainty in the origin and evolution of Mercury as well as in partitioning of elements at low oxygen fugacity. High Si in the core would lead to highly non-solar ratios of other major elements to Si, e.g., Mg/Si, as well. Moreover, meteorite compositions indicate the presence of highly oxidized regions of the solar PPD, e.g. R chondrites. Planets formed from such materials may have much smaller than typical cores. The origins of both Mercury's large core and the range of redox conditions observed in the inner solar system is unknown, but are certainly due to PPD processes. Redox variations are likely related to the abundance and transport of water ice, which in turn depend critically on many details of the starting conditions and evolutionary path of a PPD. That a single disk can produce planets with such a wide compositional range indicates the need for caution when interpreting more limited exoplanet datasets.


Kei Shimizu: V52A-07: Partial degassing and regassing of CO2 in CO2 undersaturated mid-ocean ridge basalts | Friday, 14 December 2018 | 11:50 - 12:05

Carbon flux from the Earth's upper mantle to the atmosphere is an important constraint for understanding Earth's climate stability and habitability. Increased carbonate content in the mantle lowers its solidus temperature thus affecting physical properties through increasing partial melting. CO2 has been shown to behave similarly to highly incompatible elements (e.g. Ba) during mantle melting from natural samples [1] and experiments [2]. Hence, CO2 and Ba contents in CO2 undersaturated depleted mid-ocean ridge basalts (D-MORBs) have been used to estimate the CO2/Ba ratio and carbon content in the Earth's upper mantle. However, a potential problem is that CO2 undersaturated D-MORBs may have lost CO2 through a process of partial degassing and mixing [3]. In this process, near fractional mantle melts are stored in the crust or uppermost mantle, where CO2 supersaturated low degree melts partially degas before mixing with CO2 undersaturated high degree melts. If CO2 undersaturated D-MORBs form through such process, they would have partially degassed CO2 and their CO2/Ba ratios may not necessarily be representative of their mantle sources. In this study, we report on the compositions of olivine-hosted melt inclusions in D-MORBs from the Siqueiros and Garrett transform faults. The melt inclusions are CO2 undersaturated and highly depleted in incompatible elements. Despite being CO2 undersaturated, their CO2/Ba ratios show significant ranges of 144±53 (Garrett) and 90±34 (Siqueiros). Our calculations indicate that, while partial degassing and mixing can explain the observed range in CO2/Ba ratio in the melt inclusions, the correlation coefficients between CO2/trace element and 1/trace element ratios of such process do not match those observed in the Siqueiros and Garrett melt inclusions. Instead, according to our preliminary model, the correlation coefficients for Siqueiros melt inclusions suggest that the partially degassed CO2 can re-dissolve into CO2 undersaturated higher degree melts prior to mixing. In this model, there is no net degassing of CO2 from the magmas such that the average CO2/Ba ratio of Siqueiros melt inclusions can provide a reliable estimate of the CO2/Ba ratio of their mantle source.


Jessica Arnold: P53G-3050: Refractive Index Measurements of a Solar System Organic Analog | Friday, 14 December 2018 | 13:40 - 18:00

Carbon compounds are prevalent throughout the solar system and also occur within dust in extrasolar systems. Debris disks contain dust produced by the collision of planetesimals and thus provide insight into the planet formation process. However, understanding the properties of this dust requires being able to model spectrophotometric scattered light images of disks. The composition, size, shape, and porosity of the dust grains all affect key wavelength-dependent components of such models, for example, the scattering efficiency and the phase function. Dust grain composition is parameterized in light scattering models as the optical constants, i.e. the wavelength-dependent real and imaginary indices of refraction. Hence, these values are necessary maximize the use of future and existing data sets in order to further our understanding of solar system objects and exoplanetary dust. We use transmission micro-FTIR spectroscopy to estimate the optical constants of a formaldehyde-derived polymer in the visible-near infrared region using the Kramers-Kronig approach. This polymer is similar in molecular structure to organic solids from Comet 81P/Wild2 and carbonaceous chondrite samples. We compare these measurements to those of other solar system and extrasolar organic analogs, such as amorphous carbon produced under various conditions.


Lara Wagner: T53A-04: The Peruvian Flat Slab Sag: Constraints from the Pucallpa Seismic Nest | Friday, 14 December 2018 | 14:25 - 14:40

The broad along-strike extent of the Peruvian flat slab has raised many questions about the necessary conditions for flat slab formation. Previous work has suggested the presence of two bathymetric features (one in the north, one in the south) which, in addition to other necessary conditions, play critical roles in providing the slab buoyancy needed for the formation and persistence of the world's largest modern flat slab. Between these two subducted features, there is a debate as to the precise geometry of the downgoing plate, and the effect of along-strike variability of this slab geometry on surface tectonics. In this study, we focus on a previously undescribed seismic nest located in central Peru beneath the Amazon basin city of Pucallpa. We present a description of the nest compared to other established nests, and then discuss the implications of its location and state of stress on our understanding of the geometry and evolution of the Peruvian flat slab and overriding lithosphere. Our results indicate that the Pucallpa nest demarcates the northern margin of a sag in the horizontally subducting Nazca plate in central Peru. This change in slab geometry (from flat to sagging) may be accommodated by the presence of the downgoing Mendaña Fracture Zone which provides a region of weakness along which abrupt slab geometry changes can occur. The geometry of the sag, as constrained by this cluster, closely corresponds to deformation and heat flow at the surface, suggesting a possible component of dynamic uplift in this region.