IN43A-1161 INVITED
Using Interactive Visualization to Analyze Solid Earth Data and Geodynamics Models
The geological sciences are challenged to manage and interpret increasing volumes of data as observations
and simulations increase in size and complexity. Major projects such as EarthScope and GeoEarthScope are
producing the data needed to characterize the structure and kinematics of Earth's surface and interior at
unprecedented resolution. At the same time, high-performance computing enables high-precision and fine-
detail simulation of geodynamics processes, complementing the observational data. To facilitate
interpretation and analysis of these datasets, to evaluate models, and to drive future calculations, we have
developed methods of interactive visualization with a special focus on using immersive virtual reality (VR)
environments to interact with models of Earth's surface and interior. VR has traditionally been used primarily
as a presentation tool allowing active navigation through data. Reaping the full intellectual benefits of
immersive VR as a tool for accelerated scientific analysis requires building on the method's strengths, that is,
using both 3D perception and interaction with observed or simulated data. Our approach to VR takes
advantage of the specialized skills of geoscientists who are trained to interpret geological and geophysical
data generated from field observations. Interactive tools allow the scientist to explore and interpret
geodynamic models, tomographic models, and topographic observations, while feature extraction tools
support quantitative measurement of structures that emerge from numerical simulations or field observations.
The use of VR technology enables us to improve our interpretation of crust and mantle structure and of
geodynamical processes. Mapping tools based on computer visualization allow virtual "field studies" in
inaccessible regions, and an interactive tool allows us to construct digital fault models for use in numerical
models. Using the interactive tools on a high-end platform such as an immersive virtual reality room known as
a Cave Automatic Virtual Environment (CAVE), enables the scientist to stand in data three-dimensional
dataset while taking measurements. The CAVE involves three or more projection surfaces arranged as walls
in a room. Stereo projectors combined with a motion tracking system and immersion recreates the experience
of carrying out research in the field. This high-end system provides significant advantages for scientists
working with complex volumetric data. The interactive tools also work on low-cost platforms that provide
stereo views and the potential for interactivity such as a Geowall or a 3D enabled TV. The Geowall is also a
well-established tool for education, and in combination with the tools we have developed, enables the rapid
transfer of research data and new knowledge to the classroom. The interactive visualization tools can also be
used on a desktop or laptop with or without stereo capability. Further information about the Virtual Reality
User Interface (VRUI), the 3DVisualizer, the Virtual mapping tools, and the LIDAR viewer, can be found on the
KeckCAVES website, www.keckcaves.org.
http://www.keckcaves.org
IN43A-1162 INVITED
The use of Web-based GIS data technologies in the construction of geoscience instructional materials: examples from the MARGINS Data in the Classroom project
Web-accessible geospatial information system (GIS) technologies have advanced in concert with an expansion of data resources that can be accessed and used by researchers, educators and students. These resources facilitate the development of data-rich instructional resources and activities that can be used to transition seamlessly into undergraduate research projects. MARGINS Data in the Classroom (http://serc.carleton.edu/ margins/index.html) seeks to engage MARGINS researchers and educators in using the images, datasets, and visualizations produced by NSF-MARGINS Program-funded research and related efforts to create Web-deliverable instructional materials for use in undergraduate-level geoscience courses (MARGINS Mini-Lessons). MARGINS science data is managed by the Marine Geosciences Data System (MGDS), and these and all other MGDS-hosted data can be accessed, manipulated and visualized using GeoMapApp (www.geomapapp.org; Carbotte et al, 2004), a freely available geographic information system focused on the marine environment. Both "packaged" MGDS datasets (i.e., global earthquake foci, volcanoes, bathymetry) and "raw" data (seismic surveys, magnetics, gravity) are accessible via GeoMapApp, with WFS linkages to other resources (geodesy from UNAVCO; seismic profiles from IRIS; geochemical and drillsite data from EarthChem, IODP, and others), permitting the comprehensive characterization of many regions of the ocean basins. Geospatially controlled datasets can be imported into GeoMapApp visualizations, and these visualizations can be exported into Google Earth as .kmz image files. Many of the MARGINS Mini-Lessons thus far produced use (or have studentss use the varied capabilities of GeoMapApp (i.e., constructing topographic profiles, overlaying varied geophysical and bathymetric datasets, characterizing geochemical data). These materials are available for use and testing from the project webpage (http://serc.carleton.edu/margins/). Classroom testing and assessment of the Mini- Lessons begins this Fall.
IN43A-1163 INVITED
Using GeoMapApp as an Analytical Tool for the Journey From Data Visualization to Synthesis
The potential to explore and understand our world has forever changed since the appearance of the NASA
World Wind and Google Earth virtual globes. Now, in the duration of a single breath, we can zoom from the
planetary scale of an orbiting spacecraft down to a roadside outcrop and expose layers of information with
different and rich contents. But how do we digest all this information into new knowledge that explains the
processes that have shaped the land and oceans into their present configurations and behaviors? In our
opinion we need to transition beyond visualization to interactive inquiry of multiple datasets across a span of
expertise - from the classroom to the research laboratory. Although the virtual globe enables an
unprecedented means as revolutionary as the textural search engine to discover information, presently most
data on the WEB are not adequately described with metadata to make the subsequent steps of analysis
productive. We have begun to address this limitation by linking GeoMapApp to databases in the earth and
ocean sciences where content has been vetted for thoroughness, accuracy and global coverage. With
structure in the content, the virtual globe can then manipulate these databases in what if? exercises,
compare the various attributes of a dataset with each other via graphs and symbols, and correlate results
across different scientific domains. We will show examples of such data integration using the results of four
decades of ocean floor drilling, the focal mechanisms from thousands of earthquakes, and the chemistry of
the volcanic bedrock along the crest of the mid-ocean ridge. A synthesis of ocean drilling shows the
dependency of the sediment and faunal content on bedrock age, subsidence history and plate motions
relative to the past equator and deserts. A synthesis of earthquake rupture shows focal mechanism
dependency on segmentation of the plate boundaries. Patterns in the chemistry of erupted lava are
intricately related to the fine-scale segmentation of magma bodies on the small scale and the motions of
spreading centers relative to the underlying mantle on the large scale.
http://www.geomapapp.org
IN43A-1164 INVITED
Ridge 2000 Data Integration and Synthesis
The Ridge 2000 (R2K) program is transitioning toward an increased emphasis on integration and synthesis
of data acquired on multi-disciplinary expeditions focused on understanding the geo-biological processes
associated with hydrothermal systems on mid-ocean ridges (MORs). This phase of the program will focus not
only on the compilation of existing data, but also on integration of results across disciplines, and development
of models that examine the linkages between spreading, hydrothermal, and ecosystem processes. During
this phase of the program, data from throughout the global mid-ocean ridge system will be important to
achieve a more holistic understanding of MOR processes and how they relate to the Ridge 2000 data sets
from each Integrated Study Site (ISS).
A series of workshops were held in Fall 2008 to bring together researchers from each ISS to help coordinate
the integration and synthesis phase of the program. While most cruises conducted during the R2K program
have been cataloged and basic metadata made available through R2K Data Portal (http://www.marine-
geo.org/ridge2000), additional data, including derived and interpreted data sets from R2K-funded
expeditions and other ridge-related expeditions are important to make available during this phase in the
program. Once data are available, a variety of data access and visualization tools including GeoMapApp,
Google Earth, and IVS-Fledermaus can be used to help coordinate analysis and integration efforts. We focus
on highlighting potential scientific applications made possible with currently available software tools, and
report on the R2K community feedback and utilization of data bases and visualization tools brought to light
during the Fall 2008 workshops.
http://www.marine-
geo.org/ridge2000
IN43A-1165
From data to synthesis on the cheap: the Western Basin and Range Community Seismic Velocity Model
Despite decades of effort it has been difficult for scientists to make complete assessments of the earthquake
hazards faced by Nevada urban areas. The region is riddled with hundreds of active faults, few of them
yielding much detail to researchers; earthquake waves will propagate from major fault zones through a
geologically complex crust pocked with hundreds of basins large and small; and the urban basins bear few
recordings of strong motions and had hardly been studied for their site conditions until recently.
Nevada's potential hazard was brought into sharp focus during the spring of 2008 by the M6.0 Wells
earthquake and the West Reno earthquake swarm, including the M5.0 Mogul event. Neither sequence
occurred on a known fault; the downtown core of Wells was demolished; and surprisingly high ground
motions were recorded in West Reno neighborhoods. These alarming developments give special impetus to
the Western Basin and Range (WBR) Community Seismic Velocity Model (CVM) Working Group, which was
formed with US Geological Survey support in January 2008. This broad-based group of scientists and
engineers has the objective of creating a CVM to allow computation of expected shaking from scenario
earthquakes affecting Nevada urban areas.
One consensus the group identified was to follow the design of the Southern California Earthquake Center
and Wasatch Front CVMs. However, WBRCVM development and applications are more challenging: the
former CVMs address a few large basins that contain the main scenario faults, while in the WBR most of the
hazard may originate on faults well outside the urban basins. So the WBRCVM must contain geological and
geophysical information describing hundreds of basins at a range of scales, from regional crustal
tomography down to local site conditions. Some areas have accurately measured properties, while others are
little studied and must assume interpolated or projected values. The challenge for this group, with the very
limited budgets available, is to produce a WBRCVM able to match any of the wealth of new recordings in
scenario computations. Very simple 3D trials have been able to match the peak ground velocities recorded
around the Wells earthquake, but not from the Mogul earthquake.
http://www.seismo.unr.edu/gbcvm
IN43A-1166
The EarthScope Array Network Facility: application-driven low-latency web-based tools for accessing high-resolution multi-channel waveform data
Since April 2004 the EarthScope USArray seismic network has grown to over 400 broadband stations that
stream multi-channel data in near real-time to the Array Network Facility in San Diego. Providing secure, yet
open, access to real-time and archived data for a broad range of audiences is best served by a series of
platform agnostic low-latency web-based applications.
We present a framework of tools that interface between the world wide web and Boulder Real Time
Technologies Antelope Environmental Monitoring System data acquisition and archival software. These tools
provide audiences ranging from network operators and geoscience researchers, to funding agencies and the
general public, with comprehensive information about the experiment. This ranges from network-wide to
station-specific metadata, state-of-health metrics, event detection rates, archival data and dynamic report
generation over a stations two year life span.
Leveraging open source web-site development frameworks for both the server side (Perl, Python and PHP)
and client-side (Flickr, Google Maps/Earth and jQuery) facilitates the development of a robust extensible
architecture that can be tailored on a per-user basis, with rapid prototyping and development that adheres to
web-standards.
http://anf.ucsd.edu
IN43A-1167
Services for Data Transformation and Analysis
The potential flexibility, reusability and accessibility of a service oriented architecture (SOA) makes it an compelling option to scientific data centers when choosing a programmatic interface for client software that must facilitate the discovery and access of data holdings. While this provides convenient new pathways to access, there is a clear need for server-side data transformation, analysis, and visualization processes available on-demand, closer to the data repository, as part of a user's service oriented workflow. This enables transaction efficiencies and potentially reduces the need to copy data locally before processing. The IRIS DMC is answering this need by presenting web service interfaces to a number of decoupled processing services, which will act directly on data holdings at the center and present users only with the desired transformed products. The end products can take the form of recomputed signal values, simplified output files for convenient use, and imagery based on sampled data and sensor metadata. We will present a poster highlighting IRIS's current computational services and provide a glimpse of future strategies that will offer the user a flexible suite of computing tools that are easy to access, easy to automate, and provide added value to the raw data we collect and distribute.
IN43A-1168
Rolling Deck to Repository I: Designing a Database Infrastructure
The NSF-supported academic research fleet collectively produces a large and diverse volume of scientific
data, which are increasingly being shared across disciplines and contributed to regional and global
syntheses. As both Internet connectivity and storage technology improve, it becomes practical for ships to
routinely deliver data and documentation for a standard suite of underway instruments to a central shoreside
repository. Routine delivery will facilitate data discovery and integration, quality assessment, cruise planning,
compliance with funding agency and clearance requirements, and long-term data preservation.
We are working collaboratively with ship operators and data managers to develop a prototype "data
discovery system" for NSF-supported research vessels. Our goal is to establish infrastructure for a central
shoreside repository, and to develop and test procedures for the routine delivery of standard data products
and documentation to the repository. Related efforts are underway to identify tools and criteria for quality
control of standard data products, and to develop standard interfaces and procedures for maintaining an
underway event log.
Development of a shoreside repository infrastructure will include:
1. Deployment and testing of a central catalog that holds cruise summaries and vessel profiles. A cruise
summary will capture the essential details of a research expedition (operating institution, ports/dates,
personnel, data inventory, etc.), as well as related documentation such as event logs and technical reports.
A vessel profile will capture the essential details of a ship's installed instruments (manufacturer, model, serial
number, reference location, etc.), with version control as the profile changes through time. The catalog's
relational database schema will be based on the UNOLS Data Best Practices Committee's recommendations,
and published as a formal XML specification.
2. Deployment and testing of a central repository that holds navigation and routine underway data.
Based on discussion with ship operators and data managers at a workgroup meeting in September 2008, we
anticipate that a subset of underway data could be delivered from ships to the central repository in near-
realtime - enabling the integrated display of ship tracks at a public Web portal, for example - and a full data
package could be delivered post-cruise by network transfer or disk shipment. Once ashore, data sets could
be distributed to assembly centers such as the Shipboard Automated Meteorological and Oceanographic
System (SAMOS) for routine processing, quality assessment, and synthesis efforts - as well as transmitted to
national data centers such as NODC and NGDC for permanent archival.
3. Deployment and testing of a basic suite of Web services to make cruise summaries, vessel profiles,
event logs, and navigation data easily available. A standard set of catalog records, maps, and navigation
features will be published via the Open Archives Initiative (OAI) and Open Geospatial Consortium (OGC)
protocols, which can then be harvested by partner data centers and/or embedded in client applications.
http://researchvesseldata.us
IN43A-1169
Rolling Deck to Repository II: Getting Control of Provenance and Quality
Data are now being re-used by a far wider circle of researchers and students than ever before, across a broader range of disciplines. Before, during, and after a field program, the exchange and merging of information is a dynamic, iterative and sometimes anonymous process that may occur years, or decades, after the data were initially acquired. Collaborative steps are currently being taken toward a prototype data discovery system that can be extended to the entire NSF-supported academic research fleet, including both large and small operations. With emerging shipboard Internet capabilities, near real-time transport of selected data from the rolling deck to a repository can become a reality. Much of the effort is being devoted to data preservation and access, to the identification of standard products, and to the generation of appropriate metadata, with a streamlined methodology that can be adopted by a wide range of vessel operating institutions. Beyond the challenge of just finding data, currently researchers and students struggle repeatedly to make sense of downloaded data objects. Across institutions there are wide variations in quality control procedures, most commonly none, and virtually no external communication of the history of transformations applied to a data set. For example, the status of corrections for sound velocity or roll-bias, or beampoint editing for multibeam swath seafloor mapping sonar systems are often challenging to experts and baffling to users from other disciplines. With greater re-use of data comes the greater likelihood of interpretation of artifacts as features. Tracking data provenance and quality is a problem today, and the situation will become more critical as data are more widely and rapidly disseminated. We will identify quality control criteria for standard shipboard data products, and existing tools for quality assessment with automation potential (or lack thereof). We invite discussion of an XML institutional quality certificate approach that may be used throughout the life cycle of a data object.
IN43A-1170
Rolling Deck to Repository III: Shipboard Event Logging
Data gathered during NSF-supported scientific research cruises represent an important component of the overall oceanographic data collection. The Rolling Deck to Repository (R2R) pilot project aims to improve access to basic shipboard data and ultimately reduce the work required to provide that access. Improved access will be achieved through identification of best practices for shipboard data management, identification of standard metadata and data products from research cruises, development of metadata schemas to describe a research cruise, and development of a prototype data discovery system that could be used by the entire NSF-supported academic research fleet. Shoreside data managers will work collaboratively with ship technicians and vessel operators to develop approaches that scale from smaller coastal vessels to larger open ocean research vessels. One of the coordinated subprojects within the R2R project will focus on development of a shipboard event logging system that would incorporate best practice guidelines, a metadata schema and new and existing applications to generate a scientific sampling event log in the form of a digital text file. A cruise event logging system enables researchers to record digitally all scientific sampling events and assign a unique event identifier to each entry. Decades of work conducted within large coordinated ocean research programs (JGOFS, GLOBEC, WOCE and RIDGE) have shown that creation of a shipboard sampling event log can facilitate greatly the subsequent integration of data sets from individual investigators. In addition to providing a quick way to determine what types of data might have been collected during a cruise, the sampling event log can be used to visualize the relationship, both temporally and geospatially, between the diverse types of sampling events conducted during a research cruise. Research questions in marine ecology or modeling projects are inherently multi-disciplinary and require access to a variety of data types. Improvements in cruise metadata reporting, including a sampling event log that could be contributed routinely to a centralized data repository, should improve access to research cruise data and facilitate accurate reuse of those data by colleagues.
IN43A-1171
Exploring Multidisciplinary Data Sets through Database Driven Search Capabilities and Map-Based Web Services
Relational databases containing geospatially referenced data enable the construction of robust data access
pathways that can be customized to suit the needs of a diverse user community. Web-based search
capabilities driven by radio buttons and pull-down menus can be generated on-the-fly leveraging the power
of the relational database and providing specialists a means of discovering specific data and data sets.
While these data access pathways are sufficient for many scientists, map-based data exploration can also be
an effective means of data discovery and integration by allowing users to rapidly assess the spatial co-
registration of several data types.
We present a summary of data access tools currently provided by the Marine Geoscience Data System
(www.marine-geo.org) that are intended to serve a diverse community of users and promote data integration.
Basic search capabilities allow users to discover data based on data type, device type, geographic region,
research program, expedition parameters, personnel and references. In addition, web services are used to
create database driven map interfaces that provide live access to metadata and data files.
http://www.marine-geo.org
IN43A-1172
Enabling Science Integration through the Marine Geoscience Data System Media Bank
The Marine Geoscience Data System Media Bank (http://media.marine-geo.org) was constructed to enable
the sharing of high quality images, illustrations and animations among members of the science community
and to provide a new forum for education and public outreach (EPO). The initial focus of Media Bank was to
serve Ridge 2000 research and EPO efforts, but it was constructed as a flexible system that could
accommodate media from other multidisciplinary marine geoscience research initiatives. Media Bank
currently contains digital photographs, maps, 3-D visualizations, and video clips from the Ridge 2000 and
MARGINS focus sites as well as the Antarctic and Southern Ocean. We actively seek contributions of other
high quality marine geoscience media for inclusion in Media Bank.
Media Bank is driven by a relational database backend, enabling image browsing, sorting by category,
keyword search functionality, and the creation of media galleries. All media are accompanied by a descriptive
figure caption that provides easy access to expert knowledge to help foster data integration across
disciplines as well as EPO efforts. In addition to access to high quality media, Media Bank also provides basic
metadata including geographic position, investigator name and affiliation, as well as copyright information,
and links to references and relevant data sets. Since media are tied to geospatial coordinates, a map-based
interface is also provided for access to media.
http://media.marine-geo.org
IN43A-1173
Development of a Uniform Classification Tool for Peridotites Through the use of Geochemical Data From the EarthChem Deep Lithosphere Dataset
Geochemical data management projects such as PetDB, GEOROC, SedDB, and EarthChem provide access to disciplinary solid earth data widely used for domain research. The complexity of geochemical data often poses challenges to cross-disciplinary researchers who need simple tools to bridge gaps in specialized domain knowledge. Geophysical measurements collected through EarthScope require correlation with mantle xenolith geochemical and physical sample characteristics derived from modal mineralogy and rock classification. For many samples from the literature, information relevant for properly classifying a rock is often incomplete, incorrect, or altogether missing. However, major oxide data are commonly reported for deep crustal and mantle xenoliths and provide a viable means for classification, similar to the TAS (Total Alkali Silica) classification used for igneous rocks. In partnership with EarthScope researchers, an analogous approach for the classification of peridotites is being developed based on SiO2-CaO-Al2O3-MgO (SCAM) whole rock data and mineral composition obtained from analyses served and archived in the EarthChem Deep Lithosphere Dataset. We present a comparison of results of peridotite classifications based on the bulk chemical classification and observed and calculated modal data as well as a visualization scheme that can be implemented as a database query tool to facilitate the acquisition of data based on a uniform classification scheme.
IN43A-1174
The CoreWall Project: An Update for 2008
CoreWall suite provides a set of tools for coring scientists to observe and analyze geological cores through
high resolution imagery and physical sensor data. It helps scientists efficiently construct the overall mind map
and generate more research ideas while the coring expedition is happening with high-resolution visualization
and flexible interactivity. It has been used in various expeditions including individual lake core scientists in
National Lacustrine Core Repository Laboratory in University of Minnesota, Earthscope's San Andreas Fault
Observatory at Depth (SAFOD), and Antarctica Geological Drilling (ANDRILL).
The function of visual stratigraphy correlation has been further enhanced to allow two pieces of software in
the suite, Correlator (which replace SPLICER/SAGAN applications for IODP) and Corelyzer (which is the
visualization tool in CoreWall suite), to work side-by-side with both data logs and high-resolution imagery. An
imagery-only splice composition function will also be provided for scientists to generate the common splice
table output.
In 2008 Corelyzer starts working with International Continental Scientific Drilling Project's (ICDP) Drilling
Information System (DIS) and Paleotological Stratigraphic Interval Construction and Analysis Tool (PSICAT)
to have a interoperable data exchange format. The format is based on XML and allows DIS to expose the
collected recovery data to client applications like Corelyzer and PSICAT. Corelyzer will be able to visualize the
imagery downloaded from DIS server and export the user generated interpretation annotations back to DIS.
We also propose the idea of "Rich Media Core Archive" for flexible and easy to distribute core exchange and
management. The JPEG-2000 format provides both feature-preserving high compression ratio encoding and
advanced features like multi-resolution, multi-tile imagery delivery and the ability to embed XML data files. We
can embed not only geospatial metadata but also geometry features, data logs and interpretations into a
self-contained JPEG-2000 section core archive file. It could potentially be useful to both scientists and
educators in the classroom environment.
http://www.corewall.org/
IN43A-1175
Design and Development of a Spectral Library for Different Vegetation and Landcover Types for Arctic, Antarctic and Chihuahua Desert Ecosystem
All objects on the Earth's surface absorb and reflect portions of the electromagnetic spectrum. Depending on the composition of the material, every material has its characteristic spectral profile. The characteristic spectral profile for vegetation is often used to study how vegetation patterns at large spatial scales affect ecosystem structure and function. Analysis of spectroscopic data from the laboratory, and from various other platforms like aircraft or spacecraft, requires a knowledge base that consists of different characteristic spectral profiles for known different materials. This study reports on establishment of an online and searchable spectral library for a range of plant species and landcover types in the Arctic, Anatarctic and Chihuahuan desert ecosystems. Field data were collected from Arctic Alaska, the Antarctic Peninsula and the Chihuahuan desert in the visible to near infrared (IR) range using a handheld portable spectrometer. The data have been archived in a database created using postgre sql with have been made publicly available on a plone web-interface. This poster describes the data collected in more detail and offers instruction to users who wish to make use of this free online resource.