gram, under the direction of the World Climate Research Program was begun. Called CLImate VARiability, (CLIVAR), it is now viewed as the next major atmosphere-ocean program, with the goal of greatly improving our ability to forecast variations in climate on very long time scales. CLIVAR field programs are expected to last at least through the next decade and perhaps provide at last scientific motivation for the long-awaited Global Ocean Observing System, perhaps even a Global Climate Observing System.

In looking back at the IDOE, recall that one of the major characteristics of programs during this time was the initiation of major programs and then their dissection into smaller but still collaborative programs for the sake, primarily, of simplifying the management required. Recall the four major components of IDOE: Environmental Forecasting (EF), Environmental Quality (EQ), Living Resources (LR), and Non-living Resources, or Sea Bed Assessment (SBA). Consider for a moment the breakdown, or dissection, of the EF program, which consisted largely of physical oceanography programs. The major examples are MODE (Mid-Ocean Dynamics Experiment), POLYMODE (the U.S.-Soviet follow-on to MODE), NORPAX (North Pacific Experiment), ISOS (International Southern Ocean Studies), CLIMAP (Climate Long-range Investigation, Mapping, and Prediction Study), and CUEA (Coastal Upwelling Ecosystems Analysis). The latter two showed the way to truly interdisciplinary work, with CLIMAP studying physical phenomena in the distant past using paleoceanographic techniques and CUEA showing the way to investigating the physical impacts on fisheries, or "living resources."

Another legacy of the IDOE was the start-up of a number of midsize programs during the last year of the decade. These were clearly multiyear projects, with a requirement that funding continue in order to maintain them. Whether they were started in order to guarantee funding continuity or whether the continuity was already planned is not clear to me. In any case, they were logical follow-ons, but also led the way into the large global programs to follow. Some examples are the Coastal Ocean Dynamics Experiment (CODE); Tropic Heat (TH), a study of the Eastern Pacific Cold Tongue; Pacific Equatorial Ocean Dynamics Experiment (PEQUOD); the Western Equatorial Pacific Ocean Circulation Study (WEPOCS); and Transient Tracers in the Ocean (TTO), which was a follow on to the Geochemical Ocean Sections (GEOSECS) study, and a precursor to the tracer work to be done in WOCE and other survey experiments. This is another example of two disciplines coming together to study common problems.

Perhaps a better way of looking at the transition is shown in Table 1. MODE, which Walter Munk describes briefly, was the first comprehensive look at the mesoscale eddy field. Followed by the joint Russian-U.S. POLYMODE, it paved the way for the World Ocean Circulation Experiment. In a similar fashion, the GEOSECS program, leading into the study of Transient Tracers, also paved the way for the high-precision tracer work during WOCE.

Similarly, NORPAX, expanding its range with the Hawaii to Tahiti Shuttle, largely a survey using expendable bathythermographs with frequent crossings of the equator, was one of many midsize programs that paved the way for TOGA. Others include Tropic Heat, PEQUOD, WEPOCS, the National Oceanic and Atmospheric Administration's EPOCS (Eastern Pacific Ocean Climate Study), and others.

Major characteristics of these midsize programs include the following:

1. They are needed to address problems too big for one or two principal investigators (PIs).

2. They are usually regional, not basin-wide or global.

3. They require several, but usually a s mall number of PIs.

4. They usually involve coordinated field work.

5. They are usually fully collaborative.

6. The cost averages approximately $1 million to $3 million per year.

7. There is little need for international or interagency coordination.

Some of the parallel characteristics of global change programs are the following:

1. The studies are usually long-term (several years) and large-scale (global or at least basin-wide).

2. They require a large number of PIs, although funding may be accomplished through individual grants.

3. They require a collective review process that may differ from the normal review of individual proposals.

4. The cost may average $5 million to $10 million per year or more for any given program.

5. They are usually fully inter-agency (national programs).