Budhendra Bhaduri

1
Overview of Geospatial Computing at ORNL

• Budhendra Bhaduri

Geographic Information Science
Technology Computational Sciences and Engineering
Division
2
ORNL A pioneer in Geospatial Science and
Technology
Rich History Spanning 30 years Defining
Geospatial Research Agendas Developing New
Algorithms, Software, and Data Conducting
Verification Validation Studies
3
Technology Highlights

• Geospatial Science
• Multi-resolution data analysis
• Advanced spatial modeling
• Data integration and visualization
• Geographic Information Systems (GIS)
• Decision Support Systems (DSS)
• Scientific Data Access
• Collection, Distribution, and Management
• Large scale data manipulation

4
Ongoing Research Highlights

• Analysis of surface drinking water systems and
  source water characterization
• I2IA Feature-based characterization of large
  image archives
• Transportation Routing Analysis GIS (TRAGIS)
• LandScan High resolution population distribution
  model

5
Source Water Characterization

• Georeferencing Community Water System (CWS)
  intakes to the National Hydrography Data (NHD)
• Delineating upstream contributory watersheds from
  the intake locations
• Characterizing upstream contributory watersheds
  with pesticide usage data

6
CWS Intakes

• 6945 intake location
• 6361 indexed to NHD
• 30 validated so far
• Characterization and visualization tools developed

7
Watershed Characterization

• Land cover/Land use
• Pesticide application
• 1992
• 1997
• Total by compound
• Total by crop type
• By crop-compound combinations
• 32,000 data files as initial output

8
Image-to-Intelligence Archive (I2IA)Assisting
Image Analysts in Building and Using Archives

• The amount of data available to an image analysts
  today is enormous
• We are developing an agent-based system
  architecture that
• autonomously manages a massive but dynamic image
  data archive
• transforms that to an intelligence or information
  archive to aid national security needs

9
Geoconform Images to Create Dynamic Data Archive
Monitor Distributed Data Sources
Create Intelligence or Information Archive
USGS
Feature Search, Filtering, Retrieval
Where are the power plants?
Features
NIMA
Data Search Retrieval Agents
Modeling and Analytical Applications
Visualization for Decision Support Systems
Intelligence Archive Image Analysis, Indexing,
and Characterization
Geospatial Data
NASA
Image Data Archive
Data Processing Integration Agents
NOAA
Index Power plants Airports Hospitals
Spatial Search Filtering of Image Data
EPA
Image Analysis
Intelligent Agents
Geospatial Analysis
10
Transportation Routing Analysis GISWeb-TRAGIS

• Transportation Routing Analysis GIS (TRAGIS)
  model is used to examine highway, rail, and
  waterway routes for route campaigns or future
  shipments
• Transportation routing model used by DOE
  community for routing and risk assessment studies

11
TRAGIS Rail Network

• 1100K rail network development sponsored by
  Dahlgren NSWC Joint Program Office
• Uniquely value added and topologically refined
• Used for infrastructure assurance analysis
• Also incorporated into Transportation Routing
  Analysis GIS model, funded by DOE/NTP

12
TRAGIS Routing Model

• Transportation Routing Analysis GIS (TRAGIS)
  model is used to examine highway, rail, and
  waterway routes for route campaigns or future
  shipments
• Output includes population data along routes for
  risk assessment models

13
LandScan Global Population
Innovative approach using GIS and RS techniques
to produce worlds finest population distribution
model, database, and tool

• Accepted standard for estimating population at
  risk by the DoD and DOS
• Federal agencies including DoD (DTRA, NGA), DHS
  (FEMA, TSA), DOE, USGS, NASA, EPA, and HHS (CDC,
  NIH) are current users of the data for research
  and development and routine exercises
• Over 1200 non-defense registered users worldwide
  including WHO and the UN agencies
• Used in Rand McNallys World Goodes Atlas and
  National Geographic Maps

• Finest spatial resolution (lt1km) global
  population distribution ever produced
• Global coverage in consistent raster (GIS) format
• Regular (yearly) updates
• First to employ satellite imagery worldwide
• Allows quick and easy assessment, estimation, and
  visualization of population at risk
• Integrated with transportation, socioeconomic,
  and consequence assessment models (HPAC)

14
LandScan USA

• Very High Spatial Resolution
• 3 arc second resolution (about 90 m. or 300 ft.)
  or finer
• Finer Temporal Resolution
• Nighttime (residential) and daytime population
  distributions
• High Currency
• Modeled from best available Census 2000 block
  data
• Value Added
• Demographic and socio economic characteristics
  for easy integration with risk and impact
  assessment models
• Compared to Census Data
• Better than census resolution for most city blocks

15
Space-Time Visualization
16
Intelligent Consequence Management

• Real time consequence analysis
• Compounding effects from disasters
• Dynamic traffic assignment
• GIS-data integration
• Sensitivity of lead time

17
(No Transcript)
18
Near-Real Time Population Model

• Combining
• Image processing
• GIS modeling
• High-performance computing