The GeoWall www'geowall'org

1
GeoWall Consortium
2
The big idea

• Theme
• the Earths surface (critical zone) is the
  environment. But we cannot quantitatively answer
  even relatively simple questions about its
  response to climatic and other changes, or
  provide tools to manage it effectively. Why?
• Problem
• Earth-surface science has been hindered by
  disciplinary fragmentation and a tradition of
  descriptive research and training
• Solution
• a center focused on developing an
  integrated, predictive, quantitative
  understanding of Earth-surface dynamics

3

• Desktop Watersheds Discover and advance the
  fundamental relations needed to predict landscape
  evolution and to model the coupling of ecosystem,
  landscape, and land-use dynamics

4
LIDAR

• National Center for Airborne Laser Mapping,
  National Center for Earth Surface Dynamics

5
Landscape Controls on River-to-Forest Exchange
Topography ? algal mats ? insect emergence ? bat
foraging
6
LIDAR Issues

• Large Size (10-100s Terabytes per state)
• Low cost (0.50 acre)
• Users rarely understand how the data was produced
  or processed
• Standard GIS software cant deal with the file
  sizes
• Useful to biologists, geomorphologists, first
  responders and political organizations
• Used as a map base to integrate other data

7
Jurassic Tank

• Created by non
• computer nerd
• geologists
• Extremely valuable
• to many users

8
The Need to Fuse Data
Synthetic 3D cube
Synthetic seismic (L. Pratson)
Original Image
Depositional History
9
Jurassic Tank Data Management and Visualization
Issues

• Size of a full run
• 11.43m X 5.80m X 1.52m
• 580 1cm slices
• 1.3 Gigabytes per slice (24 bit)
• 200 micron scan
• 756 Gigabytes for a full run
• 252 billion voxels
• Derived products are always made
• Standard displays are 1-3M pixels
• We need more real estate!

10
Fusion within Large Volumes

• Generic software doesnt work
• Memory inefficient
• Must run on modest computers
• Oil Patch software currently used
• Kingdom Suites, GoCAD, EarthVision
• To targeted at seismic and reservoir modeling
• Software suited for Earth Scientists
• Volume rendering with embedded surfaces,
  litholigic description, etc. Generic
  visualization software wont work.

11
Core Repositories

• One of the family jewels in the earth sciences
• Repositories at TAMU, Lamont, Minnesota, Florida
  State, USGS, state agencies, oil companies, etc.
• Chronos seems to be the key to database
  integration

12
Data FusionPersonal GeoWall2/CoreWall

• 3-8 Tile PC
• Prototype 8.6-15 Mpixels
• 6,000-20,000
• Software specifically written for the geologic
  community

LacCore/LRC/U of Minn
13
Dataflow and Database
LacCore/EVL/JOI/LDEO/INSTAAR
14
Core NavigatorShip Shoal Area, LouisianaChris
Jenkins, INSTAAR
15
Depth and Real Estate in Earth Science Display
Devices
16
GeoWall

• Stereo Projection in the classroom, office and
  lab
• Computer hardware from Best Buy
• Passive stereo
• 20 people at a time
• Under 10,000
• Small and Portable

University of Minnesota Department of Geology and
Geophysics
17
Number of GeoWalls
Currently 500
18
GeoWall 2/LambdaVision

• Tiled LCD Wall
• Prototype is 8000 x 3600 pixels (25M)
• Second generation (LambdaVision) is 55 panel,
  100,000,000 pixels
• Scaleable
• 60-500k

19
Magic Carpet
National Center for Earth-surface
Dynamics/Science Museum of Minnesota/ Electronic
Visualization Lab, UIC
20
Electronic Visualization Lab, UIC/Optiputer, USGS
21
Final Points

• Many Cyberinfrastructure users dont want to know
  that they are using it
• Must be transparent to the user
• Need targeted visualization applications
• Most of the potential users are not in the
  computational world
• Visualization is a key to bring together
  interdisciplinary groups