PPT - IGOS Cryosphere

Download Report

Transcript PPT - IGOS Cryosphere

Data Management
Considerations for IGOS
Mark Parson, NSIDC
Some Basics to Consider
• Look at what’s been done before.
• Of note: NRC 95. Finding the Forest for the Trees
• Fundamentally it is a question of data
preservation and access—two sides of the
same coin
• Data management is about people—involve
professionals early
• People need to understand best practices—OAIS
is great and should be adopted
Data Management Considerations for IGOS, 2 March 2005
More Basics to Consider
• Know your users--and recognize they will
change in unanticipated ways
• Document well - especially data uncertainty
• Scientists and data managers need to work
together on documentation
• Documentation always takes more time and effort
than you think it will
• Data management costs money--there is no
good business model but NRC 95 estimates
10% of total project cost.
Data Management Considerations for IGOS, 2 March 2005
Facilitating Integration - Issues for Observing systems
• Define and enforce data standards and
methods, e.g:
•
•
•
•
•
Data and metadata formats
Spatial and temporal scales and grids
Data collection methods
Data transfer methods and interfaces
Data quality and accountability procedures
• Put someone in charge of keeping track of all
this — herding the cats
• More and better data integration and
information mining tools will be available but
they require good data.
Data Management Considerations for IGOS, 2 March 2005
The equatorial bias—a problem in the cryosphere
Many view the world as a flat earth:
“A manner of representing the distances which gives the worst
results of all”.
- Claudius Ptolemy (c. 100 AD)
Data Management Considerations for IGOS, 2 March 2005
The equatorial bias—a problem in the cryosphere
We see a round (spherical) Earth
This informs grids,
projections,search tools,
subsetting methods, etc.
If you consider the poles
it is easy to handle the
equator, but not the other
way around.
Data Management Considerations for IGOS, 2 March 2005
An example—85% false positives
Orbit track
Eighth orbit granule
Coverage area as lat/lon extremes
Users area of interest
Data Management Considerations for IGOS, 2 March 2005