SF32A-01 INVITED 10:20h
A Distributed Computing Infrastructure for Computational Thermodynamic Calculations of Solid-Liquid Phase Equilibria
Software tools like MELTS (Ghiorso and Sack, 1995, CMP 119:197) and its derivatives (Ghiorso et al., 2002, G3 3:10.1029/2001GC000217) are sophisticated calculators used by geoscientists to quantify the chemistry of melt production, transport and storage. These tools utilize computational thermodynamics to evaluate the equilibrium state of the system under specified external conditions by minimizing a suitably constructed thermodynamic potential. Like any thermodynamically based tool, the principal advantage in employing these techniques to model igneous processes is the intrinsic ability to couple the chemistry and energetics of the evolution of the system in a self consistent and rigorous formalism. Access to MELTS is normally accomplished via a standalone X11-based executable or as a Java-based web applet. The latter is a dedicated client-server application rooted at the University of Chicago. Our on-going objective is the development of a distributed computing infrastructure to provide "MELTS-like" computations on demand to remote network users by utilizing a language independent client-server protocol based on CORBA. The advantages of this model are numerous. First, the burden of implementing and executing MELTS computations is centralized with a software implementation optimized to a compute cluster dedicated for that purpose. Improvements and updates to MELTS software are handled locally on the server side without intervention of the user and the server-model lessens the burden of supporting the computational code on a variety of hardware and OS platforms. Second, the client hardware platform does not incur the computational cost of performing a MELTS simulation and the remote user can focus on the task of incorporating results into their model. Third, the client user can write software in a computer language of their choosing and procedural calls to the MELTS library can be executed transparently over the network as if a local language-compatible library of routines is being accessed. Fourth, the flexibility of calling library functions means that the client has more control over the configuration and output of the MELTS calculation. Fifth, if the client computer is a multi-processor compute cluster capable of issuing parallel requests to the MELTS "remote" library, then these requests may be in turn parallelized to the server compute cluster to enhance throughput and performance. Application of this computational model to fluid dynamical simulations of melting and transport in the Earth's mantle is envisioned. Further information and example clients for utilizing the current prototype library for distributed computing applications can be found at http://melts.uchicago.edu.
SF32A-02 INVITED 10:40h
Global Oceanic Basalt Geochemistry From EarthChem Databases
For the past 21 years, global trace element systematics in oceanic basalts have been systematically developed, using mass spectrometry, by the MPI geochemistry department in Mainz, starting with Ba-Rb-Cs, and continuing with K-U-Th, Nb-U-Th, Pb-Ce, Pr-Mo, Nb-Ta, Sn-REE, Sb-REE, and Y-Ho relations. These were complemented by other groups, contributing e.g. Sr-REE relationships and more refined Nb-Ta systematics. Goal of these investigations was to establish relative trace element compatibilities during mantle melting, the corresponding enrichment and depletion patterns in MORB, OIB, subduction-related volcanics and the continental crust, and the relationships to Bulk Silicate Earth abundances through comparisons with element abundances in meteorites and in the continental crust (see e.g. [1]). Most of these studies were initially based on extremely limited data sets, often fewer than 100 analyses, because routine analytical techniques such as XRF and INAA were inadequate, either in sensitivity or accuracy or both, for many of the elements of interest. The advent of ICPMS technology has increased the volume of available, useable data enormously. The most recent development of laser source ICPMS is accelerating the acquisition of comprehensive trace element data even more dramatically. Although the general quality of recent trace element analyses has improved significantly, there remain large differences in reliability between published data because of varying analytical uncertainties and sample alteration. Thus, quality assurance remains an enormous task. In any case, now and in the foreseeable future, it will be impossible to assess global geochemical data without the use of comprehensive databases. Such databases are now available under http://www.earthchem.org/, comprising http://beta.petdb.ciesin.columbia.edu/; http://georoc.mpch-mainz.gwdg.de/, and http://navdat.geo.ku.edu/ . Unfortunately, the use of such databases is not without pitfalls. Often, appropriate metadata are inadequate or absent, and a significant fraction of the primary data and metadata contain errors, either in the original publication or in the transfer to the database. Therefore, analysis of these databases must be performed by experienced geochemists, who are best qualified to recognize erroneous or low-quality data. Even so, elimination of such erroneous or poor data frequently requires subjective judgment. Such evaluations should then produce "expert data sets", which can subsequently be used by non-experts, including non-specialist geochemists, geophysicists, and geologists. A partial expert evaluation of global MORB geochemistry has been prepared by Su [2], who gives segment-by-segment MORB averages. We are preparing an expert data set for global trace element and isotope geochemistry of ocean island basalts. On the basis of these data, we reevaluate global compatibility relationships. At this stage, these relationships are remarkably similar to those summarized by Hofmann [3], and by Sun and McDonough [4], using extremely limited data. This validates the original methodology used by these authors. At the same time, basalt suites containing geochemical anomalies can now be more clearly delineated. Thus, EarthChem databases allow us to develop a comprehensive view of global chemical differentiation of the Earth. [1] Hofmann, A.W. (2003) in Treatise on Geochemistry, ed. R.W. Carlson, Vol.2, 61-101. [2] Su,, Y.J. (2002) PhD Thesis, Columbia University, 472p. [3] Hofmann, A.W. (1988) Earth Planet. Sci. Lett. 90, 297-314. [4] Sun, S.-S. & McDonough, W.F. (1989) Geol. Soc. Spec. Publ. 42 (eds. A.D. Saunders & M.J. Norry), 313-345.
SF32A-03 INVITED 11:00h
Use of the PetDB Database for Calculating the Composition of the MORB-Source
We used the PetDB database to obtain a large number of geochemical analyses of mid-ocean ridge basalts (MORB). These analyses were used to derive average MORB compositions, each fulfilling a number of different criteria, with the ultimate goal to constrain the composition of the mantle from which these rocks are derived by partial melting. Here, we will discuss search strategies to efficiently retrieve subsets of data from the PetDB. One of the largest advantages of the database is the possibility to select samples based on metadata. For example, for one of the average MORB compositions calculated, the compiled data were filtered to include only samples dredged from depths in excess of one thousand meters. Similarly, the ability to choose samples based on tectonic setting, freshness or any other non-chemical parameter is a convenient way to pre-select data by using non-chemical parameters. Extracting such non-chemical data from the literature would otherwise have been a tremendous task. Another advantage of databases is their ability to organize large amounts of data based on the individual needs of the user. However, during our "mining" of PetDB we had severe problems compiling large amounts of data. The present version of PetDB does not provide compilations (over the internet) that contain more than approximately 1000 samples. Given that by far more than 1000 analyses exist, even for individual ridge sections, this is a severe limitation of the PetDB. This is compounded by the fact that combination of smaller datasets requires significant manipulation because the number of geochemical parameters returned for each search is different depending on the availability of the selected parameters. It has been our experience that the present database is effective for retrieving small amounts of data. Retrieving the large amounts of geochemical analyses on global MORB (finally $>$5000 samples with more than 50 geochemical parameters were used), however, required development of strategies to retrieve the desired data. The pre-compiled data-sets offered recently by the PetDB, however, help using large datasets with a large number of geochemcial parameters. On the other hand, the interface of the PetDB does, at the moment, not provide more advanced methods of data selection, e.g. obtaining data within a certain range of a chemical parameter (for example La/Sm or Mg#). This is made possible in other interfaces (GEOROC) and allows selection of datasets for highly specialized purposes. In summary, the inability of the PetDB to handle large amounts of samples ($>$1000) required some effort to extract and combine small dataset into the ultimately required large compilation of geochemical analyses on a large number of data. Without the PetDB database, however, the compilation and calculation of averages would have been impossible to accomplish within a reasonable time frame.
SF32A-04 INVITED 11:20h
The Curious Decoupling of Magmatism and Plate Tectonics During the Cenozoic in Western North America: Insight From the NAVDAT Database
Since the widespread acceptance of plate tectonics, magmatism in the western U.S. has been explained by subduction along the west coast of North America and destruction of the subduction system by development of the San Andreas transform fault system. However, re-analysis of space-time patterns of magmatism in western North America calls many of these inferred patterns of magmatism into question. Animation of space-time patterns found in the developing NAVDAT dataset (which currently hosts about 10,000 Cenozoic age and/or geochemical analyses; navdat.geongrid.org), demonstrates that: (1) subduction-type (e.g., intermediate) volcanism is poorly linked to the subduction system; (2) there is little evidence that slab windows controlled magmatism; (3) magmatism was clearly migratory, but not in ways that can be explained by plate-tectonic processes; and (4) magmatism was migratory at length scales ranging from 1000s of km (continental) to 10s of km (county). Several space-time patterns are evident in the NAVDAT animations, including: (1) a sweep from Montana into Nevada from 50 to about 20 Ma; (2) a clockwise sweep around the Colorado Plateau from New Mexico to southern Nevada, from about 30 to 15 Ma; (3) a burst of magmatism at about 16 Ma in northern Nevada, followed by outward sweeps to Yellowstone, central Oregon, and the Sierra Nevada; (4) a burst of magmatism in the Sierra Nevada at 3.5 Ma; and (5) several local migrations, including from Phoenix north onto the Colorado Plateau and from the San Francisco Bay area north to the Geysers geothermal field. Some of these patterns have been tied to specific events (e.g., impingement of the Yellowstone plume and Pliocene delamination), but the others are difficult to relate to plate-tectonic events. They may be caused by local tectonic events (propagating rifts?), minor convective rolls in the asthenosphere, lithospheric delamination, or delamination of a flat Laramide slab. Whatever their origin, database animation provides a powerful tool for examining these space-time patterns.
http://navdat.geongrid.org
SF32A-05 11:40h
EarthRef.org: Exploring aspects of a Cyber Infrastructure in Earth Science and Education
EarthRef.org is the common host and (co-) developer of a range of earth science databases and IT resources providing a test bed for a Cyberinfrastructure in Earth Science and Education (CIESE). EarthRef.org data base efforts include in particular the Geochemical Earth Reference Model (GERM), the Magnetics Information Consortium (MagIC), the Educational Resources for Earth Science Education (ERESE) project, the Seamount Catalog, the Mid-Ocean Ridge Catalog, the Radio-Isotope Geochronology (RiG) initiative for CHRONOS, and the Microbial Observatory for Fe oxidizing microbes on Loihi Seamount (FeMO; the most recent development). These diverse databases are developed under a single database umbrella and webserver at the San Diego Supercomputing Center. All the data bases have similar structures, with consistent metadata concepts, a common database layout, and automated upload wizards. Shared resources include supporting databases like an address book, a reference/publication catalog, and a common digital archive making database development and maintenance cost-effective, while guaranteeing interoperability. The EarthRef.org CIESE provides a common umbrella for synthesis information as well as sample-based data, and it bridges the gap between science and science education in middle and high schools, validating the potential for a system wide data infrastructure in a CIESE. EarthRef.org experiences have shown that effective communication with the respective communities is a key part of a successful CIESE facilitating both utility and community buy-in. GERM has been particularly successful at developing a metadata scheme for geochemistry and in the development of a new electronic journal (G-cubed) that has made much progress in data publication and linkages between journals and community data bases. GERM also has worked, through editors and publishers, towards interfacing databases with the publication process, to accomplish a more scholarly and database friendly data publication environment, and to interface with the respective science communities. MagIC has held several workshops that have resulted in an integrated data archival environment using metadata that are interchangeable with the geochemical metadata. MagIC archives a wide array of paleo and rock magnetic directional, intensity and magnetic property data as well as integrating computational tools. ERESE brought together librarians, teachers, and scientists to create an educational environment that supports inquiry driven education and the use of science data. Experiences in EarthRef.org demonstrates the feasibility of an effective, community wide CIESE for data publication, archival and modeling, as well as the outreach to the educational community.
http://earthref.org/
SF32A-06 INVITED 12:00h
SESAR: Addressing the need for unique sample identification in the Solid Earth Sciences
The study of solid earth samples is key to our knowledge of Earth's dynamical systems and evolution. The data generated provide the basis for models and hypotheses in all disciplines of the Geosciences from tectonics to magmatic processes to mantle dynamics to paleoclimate research. Sample-based data are diverse ranging from major and trace element abundances, radiogenic and stable isotope ratios of rocks, minerals, fluid or melt inclusions, to age determinations and descriptions of lithology, texture, mineral or fossil content, stratigraphic context, physical properties. The usefulness of these data is critically dependent on their integration as a coherent data set for each sample. If different data sets for the same sample cannot be combined because the sample cannot be unambiguously recognized, valuable information is lost. The ambiguous naming of samples has been a major problem in the geosciences. Different samples are often given identical names, and there is a tendency for different people analyzing the same sample to rename it in their publications according to local conventions. This situation has generated significant confusion, with samples often losing their "history", making it difficult or impossible to link available data. This has become most evident through the compilation of geochemical data in relational databases such as PetDB, NAVDAT, and GEOROC. While the relational data structure allows linking of disparate data for samples published in different references, linkages cannot be established due to ambiguous sample names. SESAR is a response to this problem of ambiguous naming of samples. SESAR will create a common clearinghouse that provides a centralized registry of sample identifiers, to avoid ambiguity, to systematize sample designation, and ensure that all information associated with a sample would in fact be unique. The project will build a web-based digital registry for solid earth samples that will provide for the first time a way to uniquely name and identify samples on a global scale, along with the generation of barcodes for sample labeling. We will describe a prototype of the registry, demonstrating its structure, user-friendly web interface, and functionality, and outline future plans for further enhancement of the system, pertaining to interoperability within the Geoscience Cyber-infrastructure. With the ability to track a sample through its history, the system will facilitate the ability of investigators to build on previously collected data on samples as new measurements are made or techniques are developed. The unique identifiers will also dramatically advance interoperability among existing and emerging data and information management systems for sample-based data such as CHRONOS, EarthChem, SedDB, PaleoStrat, opening an extensive range of new opportunities for discovery and for interdisciplinary approaches in research.
http://www.geosamples.org