Transcript Environmental Cyberinfrastructure: Turning Data into Knowledge

S
I
N
E
2
0
0
1
Environmental
Cyberinfrastructure:
Turning Data into Knowledge
Margaret A. Cavanaugh
National Science Foundation
SINE Workshop - SDSC
October 29, 2001
S
I
N
E
2
0
0
1
Context for Environmental
Research and Education
• Explosion of high tech tools: Information Technology,
genomics, telecommunication, GPS, EOS, advanced
sensor systems
• Demand for integration of human and societal factors
into studies of natural systems
• Expansion of disciplinary knowledge and expertise
• Public demand for improved math & science
education
S
I
N
E
2
0
0
1
Some Special Challenges for
Environmental Research
•
•
•
•
•
Global scope
Complexity of systems
Disciplinary integration
Volume of data and improved spatial resolution
Need for long-term data collection and
simulations
• Predictive power for decision-making in
uncertainty
S
I
N
E
2
0
0
1
S
I
N
E
2
0
0
1
Needs in Climate Modeling
• Centralized computational operations
• Open access policy for high-end computing facilities
• Common modeling infrastructure that links various
modeling efforts
• Institutional arrangement for delivery of climate
information for the public and private sectors
• Skilled technical and scientific workforce
S
I
N
E
2
0
0
1
Needs in the Ocean Sciences
• Hardware
• Memory
• Mass-storage capacity
• Network bandwidth
• Software
• Models, data analysis and assimilation packages
for massively parallel computers
• Visualization techniques for expanding data sets
• Well-designed and tested community models
• Skilled technical and scientific workforce
S
I
N
E
2
0
0
1
Needs in the Earth Sciences
• Access to and integration of 4 D geospatial data from
distributed, heterogeneous data formats, storage
systems, and computing platforms
• Large scale, dynamical modeling and pattern
recognition tools
• Visualization and simulation that permits extensive
data overlays, including uncertainties
• Incorporation of telemetered data from remote or
mobile sensor systems into integrated digital database
with seamless scaling across wide spatial range
S
I
N
E
2
0
0
1
Needs in the Earth Sciences (2)
• Collaboratories
• Support wide-area sensing and observing systems
• Enable data and program exchange
• Maintain and update data and models
• Support development of common framework
• Support hazard analysis and forecasts
S
I
N
E
2
0
0
1
Needs in Environmental Biology
• Bioinformatics
• Genome information analysis and archiving
• New database structures (hardware and software)
to support mining dense genetic sequence and
genomic data
• Algorithms for relating genes to biological function
and doing phylogenic reconstruction
S
I
N
E
2
0
0
1
Needs in Environmental Biology (2)
• Computational Biology
• Expanding virtual collections
• Modeling and simulating interactive multiscale
systems, such as ecosystems or species
• Networks for linking centers for scientific
collaboration
S
I
N
E
2
0
0
1
Changes in Environmental
Education
• New demands for curricular materials
• Connect data on Earth’s physical and biotic
components
• Include recent research data and results
• Discovery-based learning
• Hands-on activities
• Consideration of social implications in decisionmaking exercises
S
I
N
E
2
0
0
1
Needs in Environmental Education
• Digital Libraries for data collections and images
• Tools for manipulation of data and discovery activities
• Communications networks for access and
collaborations
• Support services for contributors, teachers, and
students
S
I
N
E
2
0
0
1
Synthesis of Needs of Various
Environmental Communities
•
•
•
•
•
•
•
•
Management of distributed observing systems
Standards for data integrity and accuracy
Capacity to archive large data sets
Rapid access to databases and computational power
through high bandwidth networks
Common modeling frameworks
Computing power for modeling, visualization, and
prediction
Long-term maintenance, operations, and data
management
Need for skilled workforce
S
I
N
E
2
0
0
1
Integrated Environmental
Challenges
• Challenges that have not been identified because
they are beyond the needs of any one community
• Examples:
• Security and privacy issues
• Combining spatial images and digital data
• Modeling frameworks that integrate earth,
atmospheric and ocean systems
• Modeling frameworks that integrate physical
systems with biological and engineering systems
• Knowledge or decision-support systems that
combine geospatial information with
demographic, economic and other data for land
use and disaster management
S
I
N
E
2
0
0
1
Conclusions
• Environmental research, education, and
management is highly interwoven
• Creative expertise from all relevant disciplines must
be engaged to solve the common challenges facing
environmental research and education
• Growth requires extensive cyber-infrastructure
resources, so need for broad community involvement
in planning the most effective development path