4
The Six Elements of GOALS
GOALS is a complex program calling for investigations of interaction between several major components of the climate system, possibly for the first time. Importantly, this interaction also implies the improvement of interfaces between the specialized scientific disciplines that previously concentrated on individual aspects of the global climate system. The program also demands a seamless interface between observing systems, models (and time integrated predictions), applications, and so forth. In recognition of the complexity of GOALS, the panel proposes the following six key elements or activities that need to be supported in order to achieve the objectives of the program:
- long-term observations and analyses;
- process studies;
- empirical and diagnostic studies;
- modeling;
- applications and human dimensions; and
- data management.
The first four elements, along with the last, were identified in the GOALS Science Plan (NRC, 1995). Because of the direct impact of seasonal-to-interannual climate prediction on society, applications and human dimensions have been added as an element of GOALS. The specific reason for this addition by the GOALS Panel is to support the concept of an end-to-end predictive capa-
bility that provides an interactive interface between the physical scientists who produce the forecasts and the user community.
The six elements of GOALS are highly interrelated, and each element has the potential to make an impact on the program's full geographical and phenomenological range. Thus, the program needs to be balanced with strong efforts in all six areas. Prioritization in each of these elements will be based on the objectives outlined at the end of Section 2.
Sustained long-term observations of key variables of the climate system form the backbone of GOALS. They are required in order to provide both a robust description of seasonal-to-interannual variability and its context relative to longer and shorter time scales. The data requirements for model initialization and verification often help in defining the observations and observing systems needed. Comprehensive analyses of these global data are required with the aim of producing global syntheses of climate and its variations in useful gridded format.
Empirical and diagnostic studies often rely on the long-term observations program element and analyzed data sets. They are invaluable in defining the areas in which model improvements are required. They also help assess which process studies are needed and help identify physical relationships which suggest aspects of the climate system that are potentially predictable.
Process studies are required to improve the predictive capability of coupled models by obtaining the data necessary for analyses, which can lead to explanations of the physical processes that need to be represented by new parameterization schemes. They need to have a sampling density sufficient to resolve time and space scales of variability in the region of interest. They are expected to lead to an explanation of the physical links that can extend predictable signals to remote locations. They are necessary to test hypotheses regarding the dynamics of interactive sub-components of complex systems. Process studies can also provide a physical linkage between the more widely spaced observations obtained in new long-term data sampling arrays and networks.
Modeling involves the development and application of improved complex coupled ocean—atmosphere and ocean—atmosphere—sea—ice—land—surface models. Modeling is considered the unifying theme underlying GOALS because of the ability of models to integrate or encapsulate knowledge derived from process studies, observations, and so forth, and to generate the products required to support the applications of relevance to societal problems. The results of studies with less complex models can suggest physical processes that should be incorporated into comprehensive coupled models; they should help the development of parameterization schemes for processes not explicitly resolved in complex models. Simplified models are also necessary to better understand the behavior of complex models. They can suggest regions and phenomena where predictability exists.
Applications and human dimensions studies enhance the utility of forecasts by taking into account user needs.
Data management is considered a critical element that cuts across the totality of GOALS. It should be designed to provide a fluid interaction with the scientific community at large. A carefully configured data management plan will provide a legacy for GOALS. The coordination of a GOALS data management plan with the international CLIVAR data management plan is essential.
Figure 4-1 provides a composite description of the GOALS program. The six elements are grouped into four components, to simplify their schematic repre
sentation. The first is associated with observation systems and their iterative development. The second refers to modeling and prediction, while the third concentrates on applications and human dimensions. The fourth, data management, is seen as an overarching element. Within each class of elements, a procedure for research and development is described. Depicted between the classes of elements are a series of interactions that signify the feedbacks considered necessary for the development of each component of GOALS.