Needs for Integration/Synthesis of System Studies in Earth Sciences

Syun-Ichi Akasofu

Founding Director, IARC

In a scientific field, it is common to deal with a system that consists of many elements or subsystems. All of us engage in a study of system components and their parts. However, at a certain point in time, integration/synthesis becomes necessary. Indeed, when a field succeeds in I/S, it makes a distinct, often epoch-making, progress. In most fields of geophysics, considerable progress has been made in system component studies during the last few decades. Thus, it is worthwhile to make a little more effort to achieve I/S.

A study of individual system elements may be called a discipline here. In the early stage of development of a field, we deal with system components (disciplines). As the field begins to develop and seeks to achieve broader relevance, interdisciplinary and multi-disciplinary research begins. As the field begins to mature, I/S of multi-disciplinary research efforts becomes necessary in order to understand the system. In fact, a system study cannot be accomplished without I/S. In other words, I/S is the main tool for a system study. Here, integration means to assemble parts (some known and some unknown) in a way similar to constructing a rocket or satellite, while synthesis implies that the assembly may result in something new and unknown, namely an unexpected product, even a paradigm change. ⁽¹⁾ The simplest example of the need for I/S can be explained by using a pencil. One group of researchers finds that a pencil is sharp and hard, but another group finds later that it is round and soft. Thus, a controversy naturally erupts. However, I/S can tell both groups that a pencil has two incompatible ends--the point and the eraser--elevating the understanding of the object.

Using the study of the human body as an analogy, individual disciplines may be equivalent to the study of the brain, heart, stomach, etc. A multi-disciplinary study may be equivalent to the study of the nervous system, the circulatory system, or the digestive system. However, if one plans to examine how the body responds to overeating, it is necessary to study the human body as a whole. In the study of global warming, the earth corresponds to the human body and overeating corresponds to the release of carbon dioxide into the atmosphere. The resulting obesity (global warming) should be dealt with in a different dimension of I/S, for instance the Human Dimension study, or others.

I/S is a very difficult task in a great variety of ways when one faces the concrete steps that need to be considered in accomplishing it and succeeding at it. I/S often tends to slow down, stall, or even disintegrate into system component studies. Some of the reasons are given in the following pages and should be considered carefully in planning I/S.

First of all, in pursuing I/S, there is a fundamental contradiction in the way we conduct research in the present trend of specialization. Although everyone may be aware that I/S is essential for making progress in any field, everyone has no choice but to excel in a narrowly focused subject, in order to get recognized as an expert by the community. This tendency to study a narrowly focused subject may also be basic human nature, as one tends to be more interested in one aspect of a particular tree (or one end of a pencil) rather than the forest (or both ends of a pencil), and thus one tends to forget the health of the forest as a whole. This is one of the reasons why I/S studies tend to disintegrate into component studies. Further, it is often the case that researchers cannot get grants easily unless their proposals are clearly focused. I/S is often treated as an unfocused subject by the referees. In some cases, I/S becomes nothing but a large umbrella under which individual researchers conduct system component studies; they may result in fascinating discoveries (which are sometimes pronounced to be the success of I/S), but they are not a system study.

There are always opinions in any scientific field that conducting I/S is impossible, because many system components are poorly understood. Although this is true, scientists often use the statement as an excuse to concentrate on their own specialties of interest. However, a system study must proceed with the available knowledge of its components at a certain stage. Otherwise, it can never be done and progress in the field may stall. In fact, there may be feedback products, objectively revealing which system components are poorly known and how to prioritize them for study. Thus, if I/S is properly conducted, this is an objective way for the prioritization, which is better than an instant committee decision. The healthy advance of a scientific field requires I/S.

Secondly, there is always considerable confusion about the definition of I/S. The definition of I/S is different for different fields and depends on the stage of progress and also on the purpose of the research.

In some fields, I/S is to assemble and standardize data acquired by different people. In other fields, I/S is to bring observers and modelers together or to compare different models. In still other fields, I/S is to establish a new field or project of research by combining several different, often well-developed fields. Thus, it is not possible to avoid the confusion, in which different researchers even in the same field have very different definitions of I/S, depending on their individual understanding of the progress of their field. For this reason, in reality, it may be difficult to accomplish I/S in a concrete way with a large number of leading researchers who tend to be naturally independent.

I/S cannot succeed unless it has a sharply defined goal, which is shared by all the small core team members (PIs) who lead participants. All the PIs have to agree on the final product . Using the earlier analogy, the goal of one group must be limited to how the human body responds to overeating (how the earth responds to a large release of carbon dioxide to the atmosphere). In dealing with the resulting obesity (global warming), we must define a new I/S with a different dimension including medical, psychology, and social science specialists.

It is not possible to succeed in I/S by having many goals. There are now enough projects being undertaken by scientific organizations and state and federal agencies that cover practically all aspects of the world's problems. We need to deermine a project under a sharply defined goal, rather than trying to loosely coordinate many or all the relevant projects. Without a sharply defined goal, I/S will disintegrate immediately into existing individual projects.

What I have learned about I/S during my research career is that to reach a sharply defined goal, I/S is best accomplished with a small core group of PIs. This is partly because I/S cannot succeed unless the PIs together can agree on defining and planning the concrete steps to take in accomplishing I/S. Each PI should have his or her own "road map," as well as milestones for each year, which are understood and accepted by the other PIs. They have to construct a common "I/S chart" to make sure that each knows precisely what the others are working on. Even with this focused approach, strong leadership is needed to push forward I/S.

A study of subsystems is in most cases incomplete. Thus, in order to respond to the opinion that the understanding of the subsystems is too incomplete to conduct I/S, one solution is to select several (at most 6) key subsystems to study as a first step, but more important is to plan carefully how to build up, systematically, the I/S study based on the first step. In many cases, a second and third step may be necessary for a higher level of I/S. An important point to make here is the plan for this systematic approach to get to the second and third stages. Otherwise, the I/S study will disintegrate into subsystem studies.

Needless to say, I/S is a long-term project . In each stage, the balanced development within and among the four themes is crucial. If each theme or its crucial parts has to rely on a chancy proposal process in accomplishing the specific goal, it is impossible to carry out I/S. Stable block funding, based on a well-planned scheme, is an efficient and effective way to develop I/S.

Above all, one of the most important aspects of I/S is education. I/S is not something only senior researchers should do, nor should they do it alone. When new students and post-docs arrive, they have to face the problem of choosing what research project they should undertake. Students have to be ultimately responsible for finding a niche that matches their interest. Finding a significant (not trivial) niche requires some sort of I/S, intentionally or unintentionally. One of the best ways to find a niche is to join I/S projects. Good students should be able to find a significant niche (a specialty) by themselves, while they are participating in I/S and pursue their own specialty, if they wish. In this way, they can also learn the lesson that they are dealing with a part of the system of their concern, and, thus, that they have to initiate or participate in I/S at a certain stage of their career in understanding the system of their ultimate concern. Otherwise, they may never learn about the forest, where their tree stands. Actually, anyone who makes a great contribution in observations, for example, must have intentionally or unintentionally performed I/S to examine carefully what kind of observation would have led to a better understanding of the subject; the result may be a breakthrough. It is often the case that talented researchers tend to dig a narrow hole, sometimes needlessly deep, and they lose sight of their surroundings.

In summary, a field cannot make healthy progress without I/S; it has to be the main tool of a system study.

We need to have some of the following in order to achieve I/S:

• final products must be agreed on by all the PIs
• a sharply defined goal must be shared by all PIs
• a small core team of PIs is required
• concrete steps must be taken together by all the PIs
• road maps, milestones, and an I/S chart must be agreed upon by all the PIs
• a balanced development of system component studies must occur
• strong leadership is required
• a long-term project must be developed
• education is crucial, as is finding a niche

⁽¹⁾ Footnote: Climate research has made a considerable advance during the last few decades by combining observations and modelings (Trenberth, 1992; Raddiman, 2001). The theory of plate tectonics resulted from the synthesis of many geophysical and geological facts (Stewart, 1990). The concept of magnetospheric physics was born by combining a great variety of auroral and satellite observations (Akasofu, 1968).

References

Akasofu, S.I., Polar and Magnetospheric Substorms , D. Reidel Publ Co., Dordrecht, Holland, 1968.

Ruddiman, W.F., Earth's Climate ; Past and Future, W. H. Freeman and Company, New York, 2001.

Stewart, J.A., Drifting Continents & Colliding Paradigms , Indiana University Press,1990.

Trenherth, K.E. (ed), Climate System Modeling , Cambridge University Press, 1992.