by Henry Lansford
For ten hot days last August, 276 atmospheric scientists, oceanographers, data-management specialists, and computer experts from 68 institutions in 13 countries gathered at the Meteo France International Conference Center in Toulouse for a scientific meeting that was the first of its kind.
Parts of the TOGA COARE International Data Workshop were in familiar formats. Speakers used overhead transparencies to illustrate presentations to audiences seated in an auditorium, and workshop participants scanned panels of text and graphics displayed in poster sessions. But the rest of the workshop was something completely different. In between the traditional sessions, workshop participants used a battery of computer workstations to juggle and massage data and gathered in small working groups to assemble and analyze data sets and produce recommendations and time lines. These products were promptly presented and dissected in plenary sessions, then taken back to the working groups for revamping. All of the workshop sessions were documented on line on the World Wide Web using Mosaic.
COARE data sets were available to workshop participants on 40 workstations with full Internet capabilities. (Photo by Henry Lansford.)
The data that were subjected to this intense process were gathered in the western tropical Pacific Ocean and the atmosphere above it during the 1992-93 Coupled Ocean-Atmosphere Response Experiment (COARE), a highly focused field research component of the Tropical Ocean and Global Atmosphere Program (TOGA). TOGA is part of a long-term, worldwide scientific effort to learn how the ocean and atmosphere interact in a global system that produces climate variations on time scales of seasons to years. The goal is to understand this coupled system well enough to predict regional climate variations.
The workshop was managed by the TOGA COARE International Project Office (TCIPO), which is based at UCAR. TCIPO director Richard Chinman says that, to his knowledge, this is the first scientific meeting that has combined face-to-face discussions of scientific data with powerful computers that provide real-time access to the data and the capability to subject it to intensive analysis and display. "Most of the COARE data sets were available on machines set up in two adjacent rooms," he says. "The data sets were in a common data format that was compatible with many of the applications that the COARE principal investigators were used to."
To facilitate data analysis and exchange, TCIPO worked closely with the Research Data Program of the NCAR Atmospheric Technology Division to provide a computing environment that included
According to Chinman, progress toward research goals, integrated data analyses, quality control, and results at the Toulouse workshop was so significant that participants encouraged TCIPO to develop a "virtual workshop" through which investigators at their home institutions can access the same computing environment, tools, and data sets that were available at the physical workshop.
Genesis of TOGA COARE
TOGA was launched in 1985 by the World Meteorological Organization (WMO) with contributions from 16 nations. The goal was to test the hypothesis that the warm tropical ocean is tightly coupled to the global atmosphere and that this coupling is a major influence on interannual climate variability. TOGA launched systematic observations of the tropical oceans by deploying instrumented drifting and moored buoys. The data from the buoys were analyzed at weather analysis and forecasting centers around the world.
This first phase of TOGA significantly improved scientific understanding of climate variability, especially the recurring El Ni–oÐSouthern Oscillation (ENSO) phenomenon. El Ni–o is a warming of the surface of the Pacific Ocean along the west coast of South America that appears to disrupt climate in distant places when it occurs about every four to five years. It is closely associated with the Southern Oscillation, a fluctuation of the trade winds that is linked with changes in the sea-level pressure field between the Indian and Pacific oceans. ENSO events have been accompanied by droughts in Australia, India, and Africa; floods in South America; and severe winter storms in the United States. Climate variations such as ENSO have serious impacts on human affairs, including loss of life, crop failures, and depletion of fisheries. However, they cannot be predicted far enough ahead to allow government officials and other decision-makers to take effective action to mitigate their impacts.
The best hope for predicting such climate variations lies with mathematical models of ocean and atmosphere processes that simulate the behavior of the coupled system. The TOGA Scientific Steering Group (SSG) concluded that progress in modeling climate variability was limited by lack of detailed knowledge of the ocean and the atmosphere in the regions of warmest sea surface temperature. A critical area is the western Pacific warm pool, northeast of Australia, which covers more of the earth's surface than the continental United States. With water temperatures consistently higher than 28¡C, it is the largest single expanse of warm water on the planet.
The chairman of the TOGA SSG was Peter Webster, then a professor of meteorology at Pennsylvania State University and now with the University of Colorado. Webster and Roger Lukas, a professor of oceanography at the University of Hawaii, took the lead in proposing a major U.S.-sponsored process study in the western equatorial Pacific that would involve both oceanographers and meteorologists. COARE was conceived and planned primarily by Webster, Lukas, and a cadre of other university scientists, many from UCAR member institutions, who later participated in the field work. Webster says that the experiment evolved from a grassroots effort in the scientific community. "COARE was driven by the PIs from the start," he says.
In 1987, Webster and Lukas met with Kenneth Mooney, director of the U.S. TOGA Project Office at NOAA, and Richard Lambert of NSF. They defined the basic elements of a hypothetical coupled ocean-atmosphere experiment that soon became known as COARE. After a 1989 planning workshop in Noumea, New Caledonia, COARE rapidly went international. Detailed plans were developed to define the roles of ships and research aircraft from Australia, China, England, France, Japan, and the United States.
The UCAR community has been active in COARE management as well as planning. COARE organization and operations were centered in the TCIPO--organized with support from NOAA, NSF, NASA, and the Office of Naval Research and directed by David Carlson, who came to UCAR from Oregon State University in 1991. TCIPO received substantial logistical and administrative support from the UCAR Joint International Climate Projects/Planning Office. (For more background on TOGA COARE, see the January-February 1993 issue of the UCAR Newsletter.)
Collecting and integrating data
The COARE data were collected during 120 days of intensive scientific operations in the western Pacific warm pool by hundreds of researchers from 21 nations. COARE field work was directed from operations centers in Townsville, Australia, and Honiara, Solomon Islands. Researchers used 7 satellites, 7 aircraft, 14 ships, 34 instrumented buoys, 37 weather stations, and a variety of other observing tools to measure ocean- atmosphere energy exchanges. Additional data came from a program of enhanced monitoring of the ocean and atmosphere in the western Pacific that began in mid-1992 and continued through mid-1993.
At the end of the fieldwork, Carlson, who served as operations director, said, "COARE has produced what undoubtedly will be the reference data set in ocean-atmosphere interaction. These data establish new standards for accuracy, resolution, reliability, and quality."
The ultimate application of the COARE data is to improve and validate coupled ocean-atmosphere models. The need for COARE was first demonstrated by tests of such models against historical observations, which identified parameterization of ocean-atmosphere exchanges in the warm pool region as the crucial hurdle in mathematical depiction of variations in the global climate system. The COARE data will help climate modelers reproduce and predict seasonal and interannual climate variations more realistically, but a major task that comes first is to integrate all those diverse field measurements of the ocean and atmosphere into a comprehensive data set. "The ultimate goal is to model coupled ocean-atmosphere phenomena more realistically, so the goal is not to have an oceanographer produce only an oceanographic data set and a meteorologist produce only an atmospheric data set," Chinman says. "There has to be integration, and the workshop was the first real opportunity for integrating data sets. It was a big step toward a totally integrated COARE data set."
Organizing the workshop
Three chairs were appointed to get the workshop together: J. Stuart Godfrey of the Commonwealth Scientific and Industrial Research Organisation (Australia), Richard Johnson of Colorado State University (United States), and Dunxin Hu of the Academia Sinica (People's Republic of China). Scientific working groups were already organized for four areas: air-sea fluxes, ocean science, large-scale atmospheric circulation and waves, and atmospheric convection.
Roger Lukas, who was deeply involved in every phase of the workshop, says that one U.S. scientist and one non-U.S. scientist were recommended by the international TOGA COARE Panel to represent each working group on a scientific program committee, which also included the workshop chairs. "For example," Lukas says, "atmospheric convection was represented by Frank Marks from NOAA and Jean-Luc Redelsperger from the French meteorological research center." Other committee members were
Products, results, and the future
Initially, the workshop planners were afraid that the variety of measurements made in COARE might be so diffuse that many investigators would choose to stay within their disciplines instead of taking up the intellectual challenges posed by the interdisciplinary aspects of the COARE data set. But after the workshop was over, the science group leaders concluded that such fears were groundless. "Provided means are found to maintain momentum," they state in the workshop summary report, "we can reasonably expect many new and deep insights into climate processes to emerge from study of the vast COARE data set over the next few years."
According to the summary report, the workshop was extremely fruitful in terms of
One valuable workshop product was a set of time lines produced by the individual science groups for development of data products and anticipated milestones in analysis and modeling. A plenary session on the final day of the workshop summarized some scientific and data issues that cut across the interests of more than one science group.
The view was expressed at the workshop that scientific oversight of the COARE data set has reached a critical phase and that it should be structured more formally within existing scientific bodies. A set of recommendations to the TOGA Scientific Steering Group was drafted, proposing the responsibilities of a newly constituted COARE Panel. This panel would carry out the detailed scientific functions of the TOGA COARE Science Group while preserving the international oversight previously undertaken by the former TOGA COARE Panel.
Future publications and meetings also were discussed at the final plenary session. There was agreement that a series of specialized meetings in 1994- 95 should be followed by a wider COARE meeting in 1996.
"Before this workshop, it was a bit unclear as to how we would bring together the power of all these data sets and accomplish the goals of TOGA COARE," Lukas says. "The participants in this workshop laid out a five- year plan for accomplishing these goals. They defined the products, the understanding, and the collaborations that we have to develop in order to achieve the goals."
From TCIPO's point of view, Chinman sees translation of the physical workshop into a virtual workshop as a logical next step. "The technology exists for this to happen," he says, "but it's a big step to make it all work. For example, voice communication in real time over the Internet is feasible. Investigators around the world can now use work-station-based microphones and public-domain software to transmit their voices over the Internet. They can all have the same display on their workstations and discuss it with the immediacy of a face-to-face meeting. Video can even be included, but the hardware is much more expensive than microphones alone."
The workshop documentation is available on line using Mosaic. Access the session abstracts and reports via the TCIPO Home Page (http://www.coare.ucar.edu under the Data Workshop entry) and tour the workshop from either a chronological or science group perspective. A hard-copy version of the workshop summary report is available for readers who lack access to the Internet or are electronically challenged. Call TCIPO: 303-497-8692.