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.
4 Expansions and Extensions GEOGRAPHICAL EXPANSION As discussed above, top priority for the future of observations related to climate modeling and prediction on seasonal to interannual scales is to sustain the (heavily Pacific) Tropical Ocean and Global Atmosphere (TOGA) program observations now in place or soon to be so [full Tropical Atmo- sphere Ocean (TAO) array] and to arrest the degradation of the operational meteorological network, particularly in tropical regions. But these steps alone are only first steps. We must expect that in the future, as in the past, an evolutionary, research-based process will provide strong indications that certain observations should be extended or expanded in certain directions, and the scientific and programmatic machinery must be capable of dealing with these indications as they arise. Already one can argue in favor of bringing observing systems in the other two tropical oceans up to the intensity and extent now nearly realized in the Pacific Ocean. The Pacific is the seat of the strongest El Niiio/ Southern Oscillation (ENSO) variations, and thus the logical site of initial focused effort. But both of the other oceans, and especially the Indian Ocean, have demonstrable interannual variations that exhibit correlation with climatically important fluctuations in the Pacific. For example, Yasunari (1987) and Rasmusson et al. (1990) described the propagation of both bien- nial and 3- to 4-year atmospheric signals from the Indian to the Pacific basin. Yasunari (1987) argues "that the ENSO should be understood as a 37
38 OCEAN-ATMOSPHERE OBSERVATIONS global-scale land (cryosphere)-atmosphere-ocean coupled system rather than an atmosphere-ocean coupled system over the equatorial Pacific." Rasmusson et al. (1990) state that "the ENSO cycle cannot be fully described and understood without consideration of the entire Indian Ocean-Pacific Ocean sector, i.e., monsoon-tradewind systems, as a unit." In the Atlantic sector, interannual variations in northeastern Brazil precipitation show relation- ships not only with ENSO but also with Atlantic sea-surface temperature (SST) (Hastenrath and Heller, 1977; Moura and Shukla, 1981; Kousky et al., 1984). In addition, many of the important climatic anomalies affecting regions of Africa are closely associated with tropical Atlantic SST anoma- lies (Lamb, 1978; Servain, 1991; Lamb and Peppler, 1991; Hirst and Hastenrath, 1983; Wagner and da Silva, 1993). A recent study by Zebiak (1993) sug- gests the existence of a coupled model of variability in the tropical Atlantic that is dynamically similar to ENSO and perhaps related to interannual climate variability in surrounding regions. Both the tropical Indian and Atlantic oceans are strongly wind driven, and therefore determination of wind forcing faces the same difficulty that gave rise to the Pacific TAO array: lack of other means to determine the surface wind accurately and with appropriate spatial resolution in the equa- torial waveguide. Basic considerations like these point toward the deliberate extension of TAO-like arrays and toward increased expendable bathythermograph (XBT) and drifter programs. Assuming that neither the energy in the community nor the resources in the agencies will suffice to start such expansions in both oceans now, on balance we favor beginning the expansion in the In- dian Ocean, for the intrinsic scientific importance noted above and for the practical effect of programmatic synergy with major World Ocean Circula- tion Experiment (WOCE) and Joint Global Ocean Flux Study oceanographic efforts scheduled for that ocean in 1994-1995. This reasoning underlies our Conclusion 3. To make this sequencing recommendation is not to say that similar plans for the Atlantic should not also be brought to maturity and then set in motion on as rapid a timetable as community interest and re- sources permit. Providing a more detailed road map for this expansion is not something that this panel alone can confidently do. Scientists directly interested in making long-term measurements in the two oceans and scien- tists interested in developing models or prediction schemes that would uti- lize such measurements should be the ones to put forward cases for particu- lar programs as contributions to enhancing climate predictions. We expect that a good deal of this rationale will be developed as part of the emerging Global Ocean-Atmosphere-Land System program. When such cases are made, agencies with responsibility for measurements and predictions of climate variations should respond positively to them. For the mid- and high-latitude oceans, the need for global drifter de-
EXPANSIONS AND EXTENSIONS 39 ployments and for further exploitation of voluntary observing ships (VOSs) potential in order to obtain an improved global SST mapping capability has been noted above. Beyond these improvements the expansion situation is less clear. There does not seem to be a strong rationale for extending spatially fine-grained buoy arrays to higher latitudes on grounds of improv- ing wind fields there. The details of the forced ocean circulation response are presumably less critical for feedback to the atmosphere at middle lati- tudes, scatterometer data may be reliable indicators of stress at these higher wind speeds, and the operational meteorological models are more capable of producing reasonable midlatitude surface wind analyses. There is a rationale for establishing a limited set of open-ocean sites, in areas of dif- fering climatology, for accurate determination of all the important air-sea flux parameters to the highest-available accuracy. Long-term operation of such sites will provide important checks on flux fields derived from present and future spacebome sensors. Because the chain of steps from spacebome sensor to geophysical field is generally long and complex and because the particulars of this chain tend to vary from spacecraft to spacecraft, from region to region, and from time to time, the ongoing calibration afforded by such sites will remain important. In addition, certain key variables like surface air temperature and hu- midity, and thus the components of the air-sea fluxes that depend on them, cannot be measured from space by any techniques now known or envi- sioned. Thus, long-term in situ measurement sites are required to provide the complete records of SST, near-surface meteorology, and air-sea fluxes that are necessary to validate the global fields available from climatologies and from model-based analyses. EXPANSION OF TECHNIQUES AND OF FIELDS OBSERVED Beyond the ones discussed above, a number of other observables and measurement techniques are best described as being in the research arena now but that in the near future hold distinct potential for extensive long- term monitoring with real impact on climate predictions. Assuming that the TOGA TAO array is continued, the buoys of that array afford ready platforms for the collection and transmission of a number of parameters beyond wind, SST, and upper-ocean thermal structure. Rain gauges already have been deployed on TAO moorings with success. Moored salinity (conductivity) sensors have achieved much better reliability and stability in recent years. Robust and semiunattended instrumentation for accurate determination of surface heat flux via bulk formulas on both buoys and VOSs is becoming a reality; this includes the promise of VOS SST measurements that are not contaminated by ship seawater intake effects. One of the most intriguing prospects is that it may soon be possible to
