40 Geoinformatics 2008—Data to Knowledge project at the San Diego Supercomputer Center (SDSC, and provide data visualization for various core-drilling com- http://portal.geongrid.org), and the development of Web munities. Corelyzer is the initial visual core description tool service interfaces at IRIS, UNAVCO, and Stanford. The portal developed for the CoreWall suite. Corelyzer allows scientists is implemented using the open-source portal infrastructure to collaborate over huge data visualizations on a desktop software, GridSphere, which supports the well-known Java workstation using one or more monitors with great interactiv- portlet interface, JSR 168, or the Portlet application program- ity and scalability. The main software architecture was devel- ming interface (API). It uses a set of “core” portlets that have oped using Java language with a native scene-graph library. been developed in GEON for data registration, searches, and The user-interface module and data-retrieval module were workspace services. written in pure Java. The scene-graph library was developed In this presentation, we will provide a report on the cur- in native C language with standard Open Graphics Library rent state of development of the EarthScope data portal. So (OpenGL) for efficient rendering. far, a preliminary deployment of the portal software has been conducted on systems at SDSC; initial designs have been accomplished for the StationDiscovery, DataDiscovery and Scalability DataPackaging services; and IRIS, UNAVCO, and Stanford Corelyzer was designed to be scalable. A graphics system have implemented the alpha version of the corresponding Web that employs level-of-detail (LOD) control and paging services, which runs on servers at their respective locations. (a method to conserve the amount of memory an image needs The beta version of these Web services will be demonstrated to load) was implemented inside Corelyzer. This graphics during the presentation. system allows scientists to load and interact smoothly with data representing thousands of meters of geological cores; one kilometer of core data produces roughly 30 gigabytes (GB) of Enhancing Core Drilling Workflows Through raw imagery. Advanced Visualization Technology By Yu-Chung Chen,1 Jason Leigh,1 Andrew Johnson,1 Visualization Capability Luc Renambot,1 Emi Ito,2 Paul Morin,3 Sean Higgins,4 Corelyzer supports hardware setups that range from a Frank Rack,5 Richard Levy,5 and Josh Reed6 single screen on a laptop computer to six liquid crystal display (LCD) panels connected to a single desktop workstation. The 1Electronic Visualization Laboratory and the Department of Computer Sci- system scales core images with different formats and resolu- ence, University of Illinois—Chicago, Chicago, Ill. tions to match the physical core sample size. The main user interface provides major data visualization capabilities for 2 Limnological Research Center, University of Minnesota, Minneapolis, core drilling, such as high-resolution core imagery, numerical Minn. core logging data, lithologic diagrams, smear slides, thin sec- 3Department of Geology and Geophysics, University of Minnesota, Min- tions, and user-generated free-form or structured annotations. neapolis, Minn.

4Consortium for Ocean Leadership, Washington, D.C. Software extensibility The Corelyzer source code was released under an open- 5Department of Geosciences, University of Nebraska—Lincoln, Lincoln, Nebr. source license and uses plain Extensible Mark-up Language (XML) file formats. Anyone can take this code and make 6Antarctic Geological Drilling Science Management Office, University of modifications to fit his or her needs. For example, with a Nebraska—Lincoln, Lincoln, Nebr. simple exporter module, the Drilling Information System (DIS) can export core data along with core imagery as a Core- Everywhere in the sciences, modern information technol- lyzer session file format, which enables all the core data to ogies change the way people work. New tools and equipment be loaded seamlessly into Corelyzer. Corelyzer also provides are constantly being developed to help scientists to process a a plug-in framework, which allows third-party developers huge amount of data and to observe detailed phenomenon that to extend its functionalities and capabilities; for example, they could not see before. We present the design and develop- support for lithologic diagrams was developed by Josh Reed, ment of an initial visual core description tool with collabora- who is the information technology manager of the Antarctic tion and annotation features that may be used for core drilling Geological Drilling project (ANDRILL, a third-party entity). expeditions. By observing the use of the tool during real core Moreover, for standardized core data (metadata) distribution, a drilling expeditions, we have learned how scientists make use “core feed” plug-in was designed to allow users to subscribe to of it and how it fits into modern core drilling workflows. core data description feeds defined in the standard syndication The CoreWall Suite is a set of tools designed to aid real- format. Users can look up the available feeds and subscribe to time stratigraphic correlation, create initial core descriptions, interesting core data, in the same manner as “.” The Geoinformatics 2008—Data to Knowledge 41 feed provides the metadata required to download and interpret order to easily access related data. The comments from the sci- actual imagery and numerical core log datasets. entists have been positive, and in the 2007 season, ANDRILL increased the number of CoreWall workstations to six for the entire science team. One CoreWall workstation was set up Deployment and Usage right at the drill site to help the drillers make on-the-spot drill- The CoreWall prototype has been used since 2006 by ing decisions based on collected data. the National Lacustrine Core Repository at the University of Minnesota and by the Lamont-Doherty Earth Observatory at Columbia University. At the end of 2006 and 2007, Core- An Analysis of Landscape Change Based on lyzer was used in core drilling expeditions by the ANDRILL Remote Sensing and Geographic Information project. In the 2006 season, ANDRILL deployed with two Systems in the Jinghe Basin, China CoreWall workstations with 30-inch LCD displays (fig. 1). The workstations mainly were used as follows: By Yonghua Zhao1 1. During the , a CoreWall workstation was placed alongside the physical cores on the tabletop to help with 1College of Earth Science and Land Resources Management, Chang’an core description. The visualization capability that allows University, Xi’an, China. zooming into the high-resolution images beyond a core’s physical scale while still maintaining details made the Using digital Landsat Thematic Mapper (TM) and setup act like an electronic microscope for the cores. This Enhanced Thematic Mapper (ETM+) imagery from 1986, capability made it easier to do more accurate and detailed 1995, and 2000 and a geographic information system (GIS), observations. For example, Dr. Franco Talarico (Univer- landscape changes were interpreted and analyzed in the Jinghe sity of Siena, Italy) used to manually draw all of the core basin (a region in China that is experiencing serious soil ero- clasts on paper in order to classify them. With CoreWall, sion problems) in order to provide basic data for local deci- he now conducts research more efficiently with modern sionmaking and for sustainable land use and management. The tools and techniques. results showed that between 1986 and 2000, most of the area covered in the basin was classified as grassland. The second 2. In the morning briefing, the other workstation was used largest area was cropland, the third was shrubland, and the for progress report explanations and tours of the core fourth was forestland. imagery alongside physical core samples. This worksta- Because they cover the most area, grassland and crops tion was set up in a public discussion area in order to help probably have the most important effect on the direction of the researchers conduct context-sensitive discussions that landscape change, ecological and environmental change, benefited from being able to see the images on the screen. the safety of the regional ecology, and so on, in this region. Although there were only two CoreWall workstation Forests have an important function in maintaining envi- setups for the entire science team, all involved personnel were ronmental quality, preventing soil erosion, or maintaining encouraged to install Corelyzer on their laptop computers in ecological balance in the region; however, the combined area of all forests (including scattered forested areas) and shrubland was less than 11 percent of the total basin area. Only the area classified as built-up land always showed an increase from 1986 to 2000. Areas of crops, forests and scattered forests, and unused land increased between 1986 and 1995 and decreased between 1995 and 2000. Bidirectional change of all landscape types was������������������������������������������������� more obvious between 1995 and 2000,��������������������� ���������and land�����- scape change was more obvious between 1986 and 1995. Above all, these changes showed that the landscape developed continuously and obviously was transformed before 1995; the landscape became regulated after 1995. Among the types of areas that showed an increase, crop areas increased the most, by about 4,165 hectares (ha) from 1986 to 1995. The second greatest increase was in areas classi- fied as shrubland, by about 2,207 ha between 1995 and 2000. Unused land increased by about 849 ha between 1986 and 1995, and grassland increased by about 816 ha between 1995 and 2000. The increase in the amount of area covered by water Figure 1. A Corelyzer set-up, which is running on an Apple was less than 94 ha from 1995 to 2000. computer and uses two Apple 30-inch cinema-display Among the types of areas that showed a decrease, grass- monitors. lands decreased the most, by about 3,125 ha from 1986 to