Monterey Bay Aquarium Research Institute
Monterey Bay 2006 Field Experiment
The Collaborative Ocean Observatory Portal (COOP)

A Collaborative Portal for Ocean Observatories

Links:

* MTS/IEEE Oceans 2006 Meeting Paper
* Page from 08/11/2006, summary & discussion view
* Page from 08/11/2006, proposal & voting view

Abstract

A Collaborative Ocean Observatory Portal (COOP) is being developed to enable distributed investigators to collaboratively operate ocean observatory systems. COOP is being created within the Autonomous Ocean Sampling Network program to support the Adaptive Sampling and Prediction (ASAP) field experiment, occurring in Monterey Bay in the summer of 2006. ASAP will involve the day-to-day participation of a large group of researchers with ties to geographically diverse institutions throughout North America. These investigators will have to interact on a continual basis to optimize data collection and analysis. While some investigators may need to be physically present to launch and retrieve their assets, the long duration of these observatories will make sustained co-location of researchers difficult. Likewise, future ocean observatories and observing systems (such as moored arrays and cabled observatories) will operate 24 hours a day, 7 days a week over many years or even decades. Sustained co-location of researchers for situational awareness and decision-making will be impractical. Hence, there is a need for collaborative data distribution and situational awareness tools appropriate for the ASAP experiment, and eventually, for all ocean observatories.

Distributed operations which enable real-time situational awareness are a reality in industry (refinery operations for large conglomerates, airline operations, utility companies) defense as also civilian government agencies (NOAA, FEMA and NASA). Even academic and quasi-academic institutions (e.g. NASA’s Virtual Astro-Biology Institute and National Virtual Observatory) have made substantial investments in allowing for geographically distant collaboration and mission operations.

The needs of an ocean observatory differ from those of other collaborative ventures. The primary differences are the heterogeneity and time-varying nature of the data sources involved in an ocean observatory, and the lack of constraints in the decision space. For example, while most of the existing collaborative efforts manage a variety of semi-independent components, the set of components is well defined and varies slowly (if at all) over the life of the program, and decisions that affect these components are well defined. Conversely, ocean observing systems will consist of a wide range of mobile platforms including drifters, autonomous underwater vehicles and ships, fixed measurement assets such as moorings and radar, and remote measurements from satellites and aircraft, all of which may be reconfigured at will to respond to observational opportunities and changing objectives.

An important step in designing COOP was to examine how co-located researchers make collaborative, operational decisions. The August 2003 Autonomous Ocean Sampling Network (AOSN-II) field program made decisions in real-time, albeit with co-located researchers. Observational data was captured throughout the field experiment and was assimilated in near real time into ocean models. Every 24 to 48 hours, investigators held “Real-Time Operations Committee” (RTOC) meetings to review the previous day's observational data, the synoptic ocean forecast for the next day, forecast skill estimates, and weather reports. The RTOC meetings were held in a common meeting room where by researchers orally and visually presented their reports. It was during these meetings that results were discussed, model needs were recognized, and adaptive sampling plans were made. Observational assets were then redeployed based on decisions made during the RTOC meeting. To be successful, the COOP design would have to allow users to have an experience like attending an RTOC meeting, but from their computers wherever they happen to be located.

In designing COOP, researchers defined desirable functionality and experimented with prototype versions of COOP. A survey of ASAP Principal Investigators identified a hierarchy of needs, revolving around four central requirements: access to observational data and model outputs, data-sharing tools, on-line meeting spaces, and tools to facilitate decision making processes. In the winter and spring of 2006, the ASAP participants held a series of "Virtual Pilot Experiments" (VPE) to test their data management and modeling systems, using model-generated data in lieu of actual field data. During each VPE, the ASAP Participants used the evolving collaborative tool, providing direct experience with the system allowing the evolution of a more effective system architecture.

The resulting COOP tool set consists of several components, starting with a publicly viewable, web-based tool for reviewing the day's progress and proposed actions. Registered scientists are able to discuss the day's progress (and attach illustrative data), propose actions (and back-up those proposed actions with supporting data), and discuss and vote upon proposed actions. The tool provides links to other system components, such as a database of data collections in both original and common formats, interactive data access and manipulation tools, and pages of automatically generated graphical summaries of observational results and model forecasts.

Site maintained by Michael Godin

Last updated: May. 18, 2010