Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
OCEAN CLIMATE MONITORING Monitoring, that is the collection of regular observations of the ocean and atmosphere over large regions for long periods of time, has been frequently cited as a necessary element for progress in understanding climate variability. Yet, as we shall review, advocates have been unable so far to obtain a commitment to establishing large- scale ocean climate monitoring programs, particularly in the ocean. In this chapter, the case for ocean climate monitoring is reviewed. MONITORING: BUILDING A DESCRIPTION FOR UNDERSTANDING A description of oceanic processes, particularly of ocean heat transport and storage, is needed to evaluate the role of the ocean in climate variability. The long time scale of oceanic heat anomalies may be an important factor for climate forecasting, but the length of time needed to describe and understand these anomalies presents a problem in experimental design. Furthermore, events like the Southern Oscillation occur sporadically (typically at 2- to 10-year intervals) and have a cycle length of about 2 years. Such large-scale ocean- atmosphere interactions must be described over several events because of their complex nature. A description of a single event would not be sufficient to understand the phenomenon because each occurrence is different. An ensemble of descriptions is needed to separate out possible overlapping events and to define a characteristic occurrence of the phenomenon. (See, for example, the compositing technique used by Rasmussen and Carpenter (1982) to study the Southern Oscillation/El Nino.) The time needed to describe and understand the climatic influence of the Southern Oscillation is long. It is thus important to have some means for ocean climate monitoring that can give regular, reliable, and repeated oceanic and atmospheric observations over the course of many years. 37
38 Our challenge in experiment design is to decide what set of observations will suffice. We do not need to instrument the oceans on a fine grid to test our oceanic general circulation models. The cost of establishing a long time series, even at one point, can be great. Long series of measurements that are now being taken should be continued (WCRP, 1981). Because of cost, however, we should be sure of our need before establishing new stations. Before we can define the optimal set of measurements, we need to have at least a crude estimate of the variability of the ocean. Therefore exploratory time series of limited duration should be taken wherever feasible, particularly in conjunction with major oceanographic experiments. As our description of ocean variability improves, so will our ability to define the optimal system for ocean climate monitoring. The optimal system in this sense has been defined as the minimal set of observations required to define the critical ocean climate indices (NOAA Office of Ocean Technology and Engineering Services, 1981). In optimal design the scale of observations should match the complexity of the phenomenon we want to observe. The problem we face is that for most oceanic climate processes, our knowledge is so rudimentary that we cannot specify the optimum set of observations. Before we can, we must explore. Then, armed with a preliminary description of the phenomenon, we can begin to define what is needed for effective monitoring. The next steps in extending our monitoring of the ocean for climate need not be elaborate. Many proposals for doing this have been made. The Ocean Science Committee (1974), in a series of workshops led by Henry Stommel, recommended the establishment of "phantom weather ships." In this program, commercial ships would collect measurements as they passed certain designated points in the ocean. The resultant time series of observations would provide regular samples at fixed locations and function much as the Ocean Weather Stations did but without the great expense of maintenance. The phantom weather ship idea was reiterated (Scientific Committee on Oceanic Research, 1977) by an international panel led by R.R. Dickson. Since that time there has been no move toward implementation. The obstacle seems more to be a lack of coordination than a lack of money. Particularly internationally, we seem to have no effective means for getting this program under way. We will discuss this problem below under "Pilot Ocean Monitoring Studies."
39 Yet another relatively inexpensive source of climate information could be gained by extending the global network of sea-level observation. This would be particularly effective if extended to isolated islands. Such stations can be set up relatively inexpensively. A long-term commitment to their maintenance needs to be made. The scientific value of long-term global sea-level measurements could be high in comparison to cost. For example, much of our knowledge of the processes involved in the El Nino phenomenon (Wyrtki, 1977) comes from sea-level measurements in the western Pacific. Sea-level measurements can also serve as an indicator of the effects of increasing atmospheric carbon dioxide (Baker and Barnett, 1982). Increases in ocean temperature will presumably accompany increases in atmospheric temperature. The resulting increase in the heat content of the upper layers of the ocean will cause sea level to rise. The effects could be global, and detection may be enhanced by the coherence of the signal. If significant information on world climate variability can be gained through relatively inexpensive means (e.g., compared with satellites), what is holding us back? Proposals for sea-level observations go back many years; Stommel's workshops (OSC, 1974) recommended that they be made. Again, international coordination is an issue. In addition, the collection of simple sea-level measurements over many years is not perceived as an attractive activity: the payoff is distant, the technology is not glamorous, and the program is difficult to defend from the budget trimmers, who often mistrust long-term commitments. Nevertheless, in advancing our knowledge of the ocean's role in the global climate system, sea-level measurements remain important and effective. Though the question of long-term climate change has been considered to be outside the scope of this study, a parallel need for ocean monitoring for this purpose as well is worth noting. In a recent study on the effects of atmospheric carbon dioxide chaired by Joseph Smagorinsky (Climate Board, 1982), an ocean monitoring system for early detection of climate change was recommended. The report states, "The operational monitoring of the ocean's response to climatic change may provide an early indication of climate change. Of particular value appear to be such indices as potential temperature and salinity changes on isopycnals in the wind-driven gyres."
40 EXPLORATORY TIME SERIES To develop an effective ocean climate monitoring methodology, we first need estimates of the space and time spectrum of oceanic variability for many regions of the ocean. Further, trial time series can explore the possible benefits of and practical means for monitoring various regions of the ocean. Following the Tokyo Study Conference (CCGO, 1983), we call such short-term observational programs "exploratory time series." An exploratory time series is intended to be of limited duration and to provide a first estimate of the climate- relevant oceanic variability. If appropriate in special cases, an exploratory time series may evolve into climate monitoring. An inhibiting factor in the establishment of ocean monitoring has been the tendency to argue that once established, an ocean-monitoring time series should not be terminated. Because of the seemingly endless advance commitment this demands, there has been a reluctance to accept ocean monitoring. As a consequence, we may have been too cautious in collecting climate time series. We need a means to extend our knowledge of oceanic variability through time series measurements without feeling guilty about terminating the series. The answer: exploratory time series. Exploratory time series should be designed to assure that the spectrum of variability is resolved, to examine the feasibility of observational techniques, and to assess the benefits that might be obtained from future monitoring. They should be geographically dispersed, incorporated into large-scale oceanographic experiments, and used as a preliminary to ocean climate monitoring. Research scientists will normally design and establish the exploratory time series and analyze and review the results. Ocean climate monitoring, on the other hand, will normally be an operational activity, just as monitoring now is in the atmosphere. Although exploratory time series are a useful preliminary step, ocean climate research programs will need a reliable source of routine global data. There thus ultimately must be a commitment to ocean climate monitoring.
41 A MONITORING STRATEGY FOR NSF The need for long time series of oceanic and atmospheric measurements to describe long-time-scale and intermittent climate anomalies poses a particular problem for NSF. The collection of long time series (monitoring) of ocean variability is hard for NSF to support in the face of competing proposals for short-term focused science. Furthermore, most oceanographers believe that NSF should not be supporting climate monitoring; that is seen as NOAA's role. Monitoring, as such, probably cannot be defended as an activity for creative scientists. Yet a commitment to monitoring is necessary if we are to obtain the description of the ocean that is needed for understanding and prediction. Such a comnitment is likely to be made only by agencies other than NSF, whose mission involves environmental observation. To encourage ocean climate monitoring, NSF should work closely with the other agencies (NOAA and NASA) whose responsibility it will be to carry it out. A possible first step would be to draw up a formal interagency plan. The research sponsored by NSF will need the base of observation that can be provided by monitoring. If NOAA and NASA are to commit to the development and operation of a monitoring system, they need assurance that it is needed and encouragement to act. Recent experience indicates that a commitment to ocean climate monitoring may not be easy to achieve. The attempt should be made, however, since the need is great. The issue is broad enough that perhaps NSF should seek the involvement of the National Climate Program Office. Without the identification of ocean climate monitoring as a national need, it may be impossible to marshal the needed resources and to obtain the long-term conmitment. Some support should go as well to the development of techniques and instruments for ocean monitoring. Techniques should include means for measuring the annual cycle and interannual variability of the upper-ocean heat budget, for volumetric surveys of water-mass properties as a means for determining trends, and for measuring the fluxes of mass, heat, salt, and tracers associated with the circulation of water and its exchange with the surface layers. These techniques should be considered for development and use in ocean climate research experiments. As such, they could be supported by NSF. As new techniques develop for obtaining consistent observations over many years, NSF should work
42 with other agencies, particularly NOAA, to encourage their operational implementation. This might be done in part by joint support of the operational use of prototype techniques in the final phases of development programs. PILOT OCEAN MONITORING STUDIES Pilot Ocean Monitoring Studies (POMS) is today more an acronym than a program. The idea came from a 1978 meeting (Global Atmospheric Research Programme (GARP), 1979) chaired by R.W. Stewart of Canada. A follow-up meeting on POMS a year later (GARP, 1980) considered a comprehensive range of topics: existing data bases, improving ocean circulation models, network design, large-scale experiments, altimetric and hydrographic surveys, national and institutional programs, instruments and methods, and data management. Subsequent to that meeting, however, significant additional steps have not been taken. Although ocean monitoring is generally conceded to be essential to obtaining a description of the ocean's role in climate variability, we are not moving toward the creation of an effective program, either nationally or internationally. Internationally, the Integrated Global Ocean Service System (IGOSS), jointly sponsored by the Intergovernmental Oceanographic Commission (IOC) and the World Meteorological Organization (WMO), would appear to be the appropriate operational mechanism. IGOSS has worked to coordinate existing national ocean-observing programs. The progress here has been slow but real. IGOSS has not, however, sparked the development or even the planning of an ocean climate monitoring program. This may be because IGOSS still does not have a clear identity and objective, either in the minds of the sponsoring international agencies or in those of the member states. Some frustrated scientists have suggested setting up an ocean climate observing network independent of IGOSS, perhaps through bilateral arrangements or through the Scientific Committee on Oceanic Research (SCOR). Nevertheless, an intergovernmental framework will ultimately be needed for global climate monitoring, and this points us back to the IOC, the WMO, and IGOSS. The United States has energetically supported IGOSS through the IOC and the WMO. Further increasing the level of American support would make our contribution even more lopsided and could be ineffective. Perhaps IGOSS could be invigorated to support ocean climate monitoring through a CCCO initiative. CCCO could sponsor a definition of the
43 need for ocean climate monitoring, the implementation of a pilot network, and the use of the system to develop prototype products. If this is done in cooperation with IGOSS, it might encourage IGOSS to take over and run whatever is developed.
