A Scientific Strategy for U.S. Participation in the GOALS (Global Ocean-Atmosphere-Land System) Component of the CLIVAR (Climate Variability and Predictability) Programme (1998)

Process studies involve highly focused special experiments conducted to advance the understanding of specific physical processes. GOALS process studies should be enunciated or designed with the practical aim of improving seasonal-to-interannual climate predictions.

These special experiments often comprise a set of intensive measurements that are limited in duration and geographic extent. Usually, process studies employ an ensemble of techniques involving observations, empirical studies, and modeling to further the understanding of a process or processes. They may be in the form of experiments designed to test particular hypotheses or exploratory efforts to gather statistical information on specific processes or phenomena for use in developing and refining theoretical, mathematical, or conceptual constructions of the dynamics of various interactions between Earth system sub-components. Process studies can serve as test-beds for new observing platforms that may eventually contribute to the long-term observing system. They can improve our understanding of physical processes and help in the development of more realistic parameterization schemes in models, thereby leading to improved prediction skill. They can also provide critical tests of these models.

GOALS places priority on the improved understanding of processes leading to improvements in their representation or parameterization in coupled ocean—atmosphere—land models. Suggestions for specific process studies may emerge from the modeling community in order to address specific problems of model parameterization. The context of process studies in the GOALS program is shown in Figure 4-1.

Two basic categories of observational process studies can be defined:

Single process studies look at a particular process such as vertical mixing in the equatorial ocean, or near-equatorial atmospheric convection (e.g., as in EMEX: the Equatorial Mesoscale Experiment), or cloud-radiation feedback, and so on. They can, however, be quite complex in terms of the research and observations required for their investigation. Single process studies are relatively easy to plan, coordinate, and conduct because they often (but not always) fall within a single scientific discipline. Drawbacks of single process studies often include a lack of concurrent data and information regarding other processes and events occurring within the total system and thereby a lack of context relative to larger scales and processes.

Combined observational process studies are essentially two or more single process studies that are collocated or adjacent and are conducted contemporaneously. Combined process studies can provide a multivariate data set that has the critical mass needed to evaluate coupled model and satellite products. Such field programs can provide a facilitating framework for interdisciplinary studies, as well as a better context and supporting information than single process studies. Combined process studies are, however, usually more difficult to fund, organize, and conduct. They also require substantial collaboration among scientists from the different disciplines involved and, correspondingly, collaboration between the diverse funding sources, which are to this day oriented generally along disciplinary lines. The situation is, admittedly, improving rapidly with the USGCRP and federal agency emphasis on multi-and inter-disciplinary activities. The more complete picture that combined process studies provides makes them preferable to several isolated single process studies and experiments. Often, maximum benefit is obtained when single process studies are embedded within a larger more comprehensive combined process study or experiment. This strategy, endorsed by the panel, is also in accord with the thinking of other scientific groups and federal agencies in order to maximize the benefits from research experiments and their applications.

Process studies are expected to play an important role in formulating and testing hypotheses during GOALS, just as they have in earlier programs. For example, TOGA-COARE, the Coupled Ocean—Atmosphere Response Experiment (COARE), was developed as a part of TOGA to address the general problem of accurately quantifying air—sea fluxes of heat, moisture, and momentum in the warm-pool region of the western Pacific. Experimental modifications made to models covering the western tropical Pacific have led to the improved prediction of the global scale impacts of ENSO, illustrating the very tight coupling of the ocean and atmosphere in this region through the exceptional sensitivity of both the atmosphere and the ocean to changes in the other. COARE addressed a suite of processes that are important for coupled models but went further in conducting its observations in a key region for climate variations associated with ENSO and the global climate. The observations obtained in the experiment not

only characterized the individual processes but also resulted in a benchmark data set in a critical region for testing coupled models and remote sensing algorithms. Thus, COARE was a combined process study that addressed both the prediction and the monitoring goals of TOGA while providing data for a wide range of physical processes relating to the heat balance of the tropical Pacific warm pool.

However, insufficient progress has been made during TOGA on achieving a better understanding of the more complex processes that affect and control the atmosphere—ocean exchange fluxes, which are intimately involved in the coupling of the ocean and atmosphere on larger space and time scales than the experimental region of TOGA-COARE. The mean state and the variability within the warm-pool region of the Pacific have been especially problematic for coupled models (and even stand-alone models) to simulate correctly. Thus, it is expected that new process studies will be proposed in GOALS.

Learning from the experience of TOGA-COARE, the panel recommends that GOALS coordinate the design and implementation of essential process studies to advance our knowledge of several facets of GOALS simultaneously.

At present, proposals for particular process studies in GOALS are at an early stage of development. However, under the auspices of the PACS program, a number of process studies in the eastern Pacific Ocean dealing with ocean—atmosphere interaction, precipitation processes, and the equatorial boundary layer are already funded and currently under way. Ocean cruises took place in Spring 1997 to study the local modification of the equatorial ocean cold-tongue during the evolution of El Niño. An ocean cruise with weather radar instrumentation is occurring in Winter 1997/1998 and investigating changes in El Niño-related convection in the eastern Pacific Ocean. The Pilot Research Moored Array in the Tropical Atlantic (PIRATA) is being developed to monitor ocean—atmosphere interactions in the region of the Atlantic tropical SST anomaly dipole. PIRATA consists of an array of moored buoys that, among other objectives, will help determine a strategy for long-term observations in the equatorial Atlantic Ocean.

Several other international activities under the auspices of GOALS are also under way. The planning for the Asian—Australian monsoon system experiment (a GOALS regional focus) is proceeding with the implementation of the South China Sea Monsoon Experiment (SCSMEX). SCSMEX plans to study ocean—atmosphere interaction during the onset of the East Asian summer monsoon. A significant extension of the TOGA-TAO array into the equatorial Indian Ocean and northwest Pacific is being implemented by Japan. Process studies are in the planning stage to observe ocean—atmosphere interactions in the eastern Indian Ocean during intraseasonal variations of the monsoon. The expansion consists primarily of deep ocean moorings called the Triangle Trans-Ocean buoy network (TRITON) array. The TRITON moorings are modeled after the NOAA TAO

moorings and contribute seamlessly to the ENSO observing system data. In order to improve the understanding of Atlantic modes of variability, NOAA, together with international partners, is expanding ocean/atmosphere observations into the tropical Atlantic in pilot mode. The Atlantic observing system is built from the same proven technologies used in the ENSO observing system, namely, deep ocean moorings (e.g., PIRATA), drifting buoys, tide gauges, and VOS. Additionally, a fifth network of autonomous profiling floats will be added to the system in order to measure better subsurface ocean currents, temperature, and salinity. The Atlantic observing system, together with the ENSO observing system, and the Indo-Pacific TRITON expansion form a significant contribution to the climate module of the GOOS. A Joint Air—Sea Monsoon Interaction Experiment (JASMINE) is being proposed for the Indian Ocean. JASMINE aims at studying ocean—atmosphere interaction in the eastern Indian Ocean during the intraseasonal transition of the South Asian Monsoon. In addition, extended observations of the ocean—atmosphere interaction within the Indonesian Archipelago and the Indonesian Throughflow are planned.

Complementary land process studies are being proposed under the framework of GOALS, but their implementation will be coordinated with and undertaken by GEWEX. An example is the GEWEX Asian Monsoon Experiment (GAME) under which land surface processes are being investigated in a variety of climatic regimes ranging from tropical forests to tundra.

For the further development of process studies beyond those mentioned above, the panel suggests that the following criteria and activities be kept in mind:

1. Whenever feasible, single process experiments needed for GOALS should be coordinated and combined, although it is realized that individual pilot studies may be necessary at first.
2. Domains for GOALS process experiments should be chosen to include oceanic regions where SST or heat content variations are likely to provide a relevant atmospheric predictive capability.
3. Process experiments for GOALS should be designed to provide a critical mass of ocean and atmosphere multivariate data from in situ and remotely sensed observations to support adequate future model experimentation. They should also provide a basis for the development and evaluation of relevant remote sensing techniques.
4. Combined process experiments should be undertaken in key regions for seasonal-to-interannual climate variability, as demonstrated, for example, by model sensitivity studies.
5. Process studies should be designed for use as test-beds for new observing

• techniques that provide enhanced monitoring of priority variables in limited but important regions.
• Long-term observations should be used to provide the context for limited-duration intensive process studies.
• Workshops should be held to enable scientists to develop consortium proposals for necessary process studies.

The panel recommends that GOALS process experiments should be coordinated either with GEWEX process studies, such as GAME and the GEWEX Continental-Scale International Project (GCIP), or with DecCen process studies. Close coordination is considered necessary with climate-related U.S. and international programs to achieve mutual objectives. This is important for producing a critical mass of observations as well as for sharing the funding of process studies.

The TOGA investment in COARE and other relevant process studies should be leveraged by continuing support of analysis and modeling activities that use observations from associated field programs.