Difference between revisions of "Talk:NSF Air Quality Observatory:AQ Observatory Proposal"
From Earth Science Information Partners (ESIP)
Line 18: | Line 18: | ||
[[User:Rhusar|Rhusar]] | [[User:Rhusar|Rhusar]] | ||
+ | |||
+ | |||
+ | With the intention of providing greater specificity re cyberinfrastructure advances (and the risk of excessive wordiness!), I'll suggest an alternate winning theme: | ||
+ | |||
+ | :Solid but separate cyberinfrastructure components (DataFed and Unidata) now exist to support (respectively) the AQ and met communities. | ||
+ | ::Within these communities, the extant infrastructure supports (for research, teaching, and decision-making) the matching of end-user tools to needed observational data, but -- | ||
+ | :::The matching process can be labor intensive, especially where the needed data involve aggregations of observed and simulated information from multiple sources | ||
+ | :::The needed matchings and aggregations are becoming more difficult as | ||
+ | ::::end-user needs increase in complexity, | ||
+ | ::::data volumes grow, and | ||
+ | ::::data types evolve (with new instruments, e.g.). | ||
+ | ::Cross-community data use is very low because the above difficulties are greatly compounded by interdisciplinary variations in tool usage and the attendant data semantics. | ||
+ | :The AQO will leverage, augment and integrate DataFed and Unidata in a prototype cyberinfrastructure component that better serves researchers, teachers, and decision-makers in AQ, met, and related fields by overcoming the listed difficulties. | ||
+ | ::The overarching benefits will include facilitating cross-community use of observed and simulated data sets and streams. | ||
+ | ::Specific outcomes will include the enabling of use cases as follows: | ||
+ | :::(insert abbreviated use cases here) | ||
+ | ::The underpinning for these advances will be an end-to-end system that exhibits advanced functional and technical design elements as outlined below: (see also my comments on NSF requirements 3 & 7) | ||
+ | :::Functional Design | ||
+ | ::::Responding to Natural Events | ||
+ | ::::Responding to Needs of Users | ||
+ | ::::Overcoming Semantic Impedance | ||
+ | :::Cyberinfrastructure Design | ||
+ | ::::Classes of Data Sources | ||
+ | ::::Classes of Events | ||
+ | ::::Classes of Data-Analysis Tools | ||
+ | ::::Classes of Data Transformers and Aggregators | ||
+ | ::::Simplicity via Polymorphism and a Common Data Model | ||
+ | |||
+ | [[User:Dave.Fulker|Dave.Fulker]] |
Revision as of 09:01, January 20, 2006
Links to: AQO Proposal Main Page > Proposal | Proposal Discussion| NSF Solicitation | NSF Solicitation Discussion | People |
|
This page contains discussion on various aspects of the proposal.
'Win-theme' for the AQ Observatory project?
Based on our discussions I would propose a common (win-) theme for the AQO project along the following line:
- There is already an infrastructure for access/use of met data (Unidata) and for AQ data (DataFed)
- Each serves their respective communities well, but not the sister communities, so:
- AQO will extend these working domain infrastructures by making them interoperable at find/access/use level
- Add new capabilities arising from the synergy of the combined resources and tools
- Thus, the prototype AQO will demonstrate operational inter-domain networking [and encourage other nodes to join?]
Variations on this theme would be stated in several sections of the proposal. Is this our key theme? Is there a better way to phrase it? Is it too little, much, just right? Enough NSF-appeal?
With the intention of providing greater specificity re cyberinfrastructure advances (and the risk of excessive wordiness!), I'll suggest an alternate winning theme:
- Solid but separate cyberinfrastructure components (DataFed and Unidata) now exist to support (respectively) the AQ and met communities.
- Within these communities, the extant infrastructure supports (for research, teaching, and decision-making) the matching of end-user tools to needed observational data, but --
- The matching process can be labor intensive, especially where the needed data involve aggregations of observed and simulated information from multiple sources
- The needed matchings and aggregations are becoming more difficult as
- end-user needs increase in complexity,
- data volumes grow, and
- data types evolve (with new instruments, e.g.).
- Cross-community data use is very low because the above difficulties are greatly compounded by interdisciplinary variations in tool usage and the attendant data semantics.
- Within these communities, the extant infrastructure supports (for research, teaching, and decision-making) the matching of end-user tools to needed observational data, but --
- The AQO will leverage, augment and integrate DataFed and Unidata in a prototype cyberinfrastructure component that better serves researchers, teachers, and decision-makers in AQ, met, and related fields by overcoming the listed difficulties.
- The overarching benefits will include facilitating cross-community use of observed and simulated data sets and streams.
- Specific outcomes will include the enabling of use cases as follows:
- (insert abbreviated use cases here)
- The underpinning for these advances will be an end-to-end system that exhibits advanced functional and technical design elements as outlined below: (see also my comments on NSF requirements 3 & 7)
- Functional Design
- Responding to Natural Events
- Responding to Needs of Users
- Overcoming Semantic Impedance
- Cyberinfrastructure Design
- Classes of Data Sources
- Classes of Events
- Classes of Data-Analysis Tools
- Classes of Data Transformers and Aggregators
- Simplicity via Polymorphism and a Common Data Model
- Functional Design