can reasonably expect to study. The following paragraphs provide more information on the SO, and indicate as well a set of scientific steps that will lead to a detailed SO field program toward the end of the decade. Satellite measurement capabilities will be critical to many aspects of the SO scientific program.

2.3.11.1 Definition phase

If one asks, "What is the SO?" the answer generally comes back as something to the effect that it is the oscillation of the sea-level-pressure difference between the Indonesian region and the southeastern Pacific Ocean. This pressure fluctuation arises from a shifting of atmospheric mass between the Eastern and Western Hemispheres. Others view the phenomenon as an interaction between the Pacific trade wind fields and the monsoonal systems. In any event, none of these descriptions is really satisfactory for they only refer to indices of a process that has global ramifications.

The first job for the climatologist, then, is to define adequately the global couplings in the ocean and atmosphere that attend the thing we now simply call the SO. We will have to answer such questions as: What fields and regions of the globe participate in the SO? What are the space and time scales of variability in these fields and how can we optimally measure SO phenomena? Why is it that many SO phenomena (e.g., El Niño) appear tightly locked to the seasonal cycle? In short, we will have to do a comprehensive, quantitative job of describing what we mean by the term SO. A strong start has been made in this direction. Additional field measurements (see below) will be required to complete the process.

2.3.11.2 Hypotheses

The description called for above should give rise to a series of hypotheses regarding the mechanisms associated with SO-type phenomena. For instance, numerous spectral analyses indicate that the SO phenomena most often lie in the frequency range between about one cycle per three years and one cycle per six years. We currently have no hypotheses to explain this range of preferred time scales. It does not appear possible that their origin is in the atmosphere, for the atmosphere's "memory" is simply too short. However, the ocean is an attractive candidate for these time scales since they are comparable with the circulation time around major ocean gyres.

If the oceans do play an important role in the SO, then we must be prepared to undertake a study of ocean heat budgets since it is the oceanto-atmosphere heat flux that will influence the atmosphere and, perhaps, partially force the SO. This means we will need ways to estimate the air-sea heat exchange as well as the advection of heat by ocean currents.

2.3.11.3 Field programs; Phase I-definition

Almost all current knowledge about the SO comes from historical data measured at various land stations or islands or from ships. This type of historical data is and will continue to be a tremendous aid in conducting the "definition" stage indicated above. However, there are huge regions in the Southern Hemisphere, particularly the Southern Ocean, where there is no information: ships do not go there, there are no islands, etc. These areas, of necessity, will remain question marks in the definition phase. This will make it difficult to construct hypotheses regarding the SO. In fact, the SO appears to require an understanding of how the major atmospheric fields interact with each other and the ocean. From the oceanic point of view, we will have to understand the interaction of entire current systems. Field interactions of the type envisioned are not well described, in general, by a few spot measurements.

It is clear that a study of the SO will require a near-global perspective. The only realistic way to obtain the type of information for this view is via satellite. However, the satellites themselves will not provide all the information we need. A strong requirement exists for complementary measurements from drifting buoys, instrumented islands, ship-of-opportunity programs, etc. In context then, the satellite data will provide the "glue" necessary to tie together current conventional measurements made on a sparse spatial grid. Satellite information will also help us understand how to interpret the (spot) historical record in terms of field properties.

In summary, a preliminary measurement program using both conventional and satellite systems is required to even delineate SO phenomena, particularly in the Southern Hemisphere. Climatologists are in a position now to begin to specify the required data to outline such a measurement program.

2.3.11.4 Field programs; Phase II—process experiment

Given a reasonable set of SO hypotheses, plus an idea of the type of sampling scheme needed to resolve the SO signals, meteorologists and oceanographers will be able to design a reasonable field program to investigate the origin of the phenomenon. Such a program will be highly compatible with proposed large-scale oceanographic programs of the 1980s (e.g., the WOED discussed above) and with the instrument development required by these programs (e.g., drifting buoys, remote data transmission, etc.). The greatest potential for disagreement appears to be the siting of such experiments (e.g., tropical and South Pacific versus North Atlantic).

2.3.11.5 Data problem

A major difficulty associated with the SO project is the huge data

management specter that it raises. The climatologists will have to find a method to easily synthesize disparate types of information from numerous sources. A giant stride towards this synthesis can be achieved by developing a data system that is common to as many sources as possible. In this case the role of the satellite again appears exceedingly important, for the satellite can act as a data link that will collect and relay both remotely sensed information and information from a variety of surface and subsurface platforms. Having many types of information in a single data stream greatly simplifies many of the coordination and collation problems that would otherwise arise.

2.3.11.6 The role of satellites

The role of the satellite in future climatological research, such as the SO project, is twofold. First, it must provide estimates of remotely sensed fields that can be related to conventional historical measurements taken from platforms at sea or on land. Such variables include, but are not limited to, sea-surface temperature, wind velocity, total precipitable water, sea-level wind velocities, temperatures, etc. Much has been written about these measurement capabilities. In summary, prudent combination of conventional measurements with remotely sensed data now appears to offer a capability accurate enough to initate a SO experiment.

The second capability we need from the satellite is the ability to position and collect data from remote stations, e.g., drifting buoys, islands, ships of opportunity. This capability currently exists but has not been applied to the climatic experiment envisioned here.

2.3.11.7 Special instrument problems

The study of climatic variability introduces a special problem both for conventional and, particularly, for satellite instrumentation. Relatively small amounts of instrument drift will, over a particular sensor's lifetime, look much like the signals that we expect to study.

It is absolutely vital that instrument stability and long-term calibration be matters of key concern in any future conventional or satellite system that will be used for climate research. The stability and drift problem also suggests that the satellite products must constantly be referenced to conventional measurement sets with which we are familiar. We will be able to find regions of the ocean where the remotely sensed fields can be compared with conventional estimates of the fields. This field-field comparison may avoid the problem of trying to determine the representativeness of individual calibration sites.

2.3.12 Biological Studies

The oceanic physical processes relate closely to many biological pro

cesses of importance. Some important applications of the color scanner observations have already been mentioned above. Following are two additional brief examples of the use of satellite-related information.

Figure 8 illustrates trajectories derived from drifting buoys deployed during the Global Weather Experiment, 1979, and tracked by the Argos DCLS. As a by-product of these results, it proved possible to trace the origins of lobster larvae that developed into very significant fishery resources on some sea mounts in the eastern Atlantic. Figure 9 illustrates generalized current systems derived from theoretical studies and supported by analysis of drifting buoy tracks. The fishery resources of this area are important, and understanding the ocean dynamics that support the fishery resources is an example of the useful application of satellitesupported observations.

Turning to quite another kind of biological application, the migration of whales from the Caribbean to the Gulf of Maine has been studied for a number of years. There are, however, very little environmental data on the scale required to piece together a meaningful picture of the relationship between the migration route and the environment. Such a data base is now becoming available with Coastal Zone Color Scanner (CZCS) (visible and thermal infrared) and AVHRR data. Unfortunately, the Nimbus 7 CZCS will certainly not last much longer; hence, these data will only cover a limited two- to three-year span with many gaps. A follow-on color scanner effort would contribute to an extension of the visible data (with collocated thermal infrared data). The idea of the ongoing research is to seek a correlation between the whale migration routes over a number of years and ocean temperature and color.

Figure 8. Drift tracks of three South African weather buoys deployed in the southeast Atlantic Ocean during the Global Weather Experiment, 1979. Bottom topography features shallower than 3,000 m are shown.

Figure 9a. Composite of drifting buoy tracks in the region off southwest Africa during the Global Weather Experiment, 1979. The observations during the GWE provided for the first time essentially synoptic views of ocean circulations over extensive regions; such information is essential for describing oceanic circulation and verifying models.

Figure 9b. Conceptual depiction of characteristic large-scale flow regimes off southeast Africa. Dark arrows depict narrow, intense currents; open arrows depict flow believed to show little directional variability. The conjectured western boundary current nature of the East Madagascar Current shows similarity to features of the Agulhas Current, which is related to important fishery activity.

20'S