HR: 1340h
AN: IN33A-1171 [Abstracts]
TI: Federated Observatory "Sense and Response" Framework
AU: Chave, A
EM: alan@whoi.edu
AF: Woods Hole Oceangraphic Institution, Woods Hole Oceanographic Institution, Woods Hole, Ma 02543-1050
United States
AU: Arrott, M
EM: marrott@ucsd.edu
AF: Calit2, University of California, San Diego
9500 Gilman Drive, La Jolla, Ca 92093-0436
United States
AU: * Orcutt, J
EM: jorcutt@ucsd.edu
AF: Scripps Institution of Oceanography, 9500 Gilman Drive, 0233, La Jolla, Ca 92093-0233
United States
AB:
Ocean Science is at the beginning of the epoch of the "Instrumented Earth". From here on, there will be a continuous
accessible digital "signal" that represents the "Now" state of the Earth systems. For the foreseeable future, this signal
will grow in multidimensional and multivariate resolution and complexity. The primary goal of this activity is to provide
scientists with the means to characterize and respond to emergent phenomena though continuous real-time monitoring and
analysis of selected portions of Earth's signature.
The overarching focus of the Ocean Observatory Services activity is the development of a "Sense and Response" framework that
operates on a continuous basis in real-time in a controlled and comprehensive manner. This presentation addresses the designs
and prototypes of the basic components an Ocean Observatory taking place within LOOKING using WS Resource Framework based
SOA:
Data Stream Representation
Real-time Data Stream Processing
Instrument Management
Data-Stream Representation
The representations of data "sets" (data files) are often the focus of data management and archive systems, and only recently
have efforts turned their attention to the unique characteristics and requirements of streaming data. In order for broad
adoption of this essential data type by federated systems of ocean observatories enabling both producers and consumers of
data streams to interoperate with technology independence the characteristics of data streams must defined and modeled
Data-Stream Processing (DSP) Grid
This activity is researching and prototyping a Data Stream processing capability that expands and contracts its capacity to
meet the demands of dynamic and continuous real-time monitoring and analysis. The investigation is based on the hypothesis
that workflow statements are the most likely expression of scientific inquiry in a continuous real-time observatory context.
To this end, the project is focused on the integration of data stream, workflow and Grid technologies into a Data Stream
Processing (DSP) Grid. The DSP Grid is conceptualized as a workflow engine with the capability to dynamically:
Marshal the required data stream resources,
Plan the decomposition of the process graph and
Dispatch the component work across Grid resources
to maintain data stream processing in real-time.
Instrument Management
Within the context of a federated system of Ocean Observatories, the concept of an Instrument is central. Instruments
participant in a broad range of relationships spanning numerous application domains, and are required to satisfy a complex
set of interactions across a dynamic range of computational contexts. This encompasses deployment, configuration,
calibration, integration into observing system, monitoring (e.g., determining instrument status) and remote interaction
(e.g., changing sample rates, requesting data). Instrument management includes both coupled computational components (e.g.,
the bidirectional symbiotic relationship between simulation models and instrument operations) as well as traditional
operator-directed instrument control. Instrument management is a central task in observing systems and cuts across all
aspects of a observing system.
DE: 1600 GLOBAL CHANGE
DE: 9800 GENERAL OR MISCELLANEOUS
DE: 9820 Techniques applicable in three or more fields
SC: Earth and Space Science Informatics [IN]
MN: Fall Meeting 2005