Convoluting an Invisible World : Observation and Modeling in Solid Earth Physics

Abstract

Since scientific activities are understood historically and practically, our conception of science has been radically changing. While Ian Hacking has emphasized the importance of the practice of intervening in materiality as well as representation, Andrew Pickering, a sociologist of science, has suggested a new pragmatic approach for understanding scientific practices to replace the traditional representation theory. This paper focuses on arguments of scientific realism, acknowledging that recent studies of scientific practice emphasize the historicity and locality of science. It criticizes the way that they reduce the process of making reality to the scheme of stabilization between representations and materiality. It emphasizes the heterogeneity of the components in scientific practice through a description of the practices of observation and modeling in earth physics, and shows that the components call up various modalities of making reality. This paper examines the practice of geodesy, a sub-discipline of earth planetary physics based on the paradigm of plate tectonics theory. Geodesists use various methodologies of observation to explore the state of the earth. Each of those methodologies has a unique data format, as well as unique methods of data processing and data analysis, forming expertise. The paper comprises three chapters, the first on observation, the second on modeling, and third containing a discussion. The first chapter explains two representative observation methodologies: gravimetry and GPS observation. The second chapter presents examples of analyzing each observation datum to explain various techniques of realizing entities in geodesy. Gravimetry, a traditional method for exploring the condition of the earth, utilizes Newton's theory of gravitation. Gravity data are expected to contain information about all kinds of mass distribution, and are used for such purposes as underground resource exploration, the detection of ice-sheet mass variation caused by global warming, and the detection of crustal deformation. As gravity data reflect all those kinds of phenomena, geodesists need to separate each signal of the phenomena from the surrounding noise in a process called "correction." GPS observation, a rather new geodetic method, is a widespread technology to determine the position of observation points; it is used for detecting crustal deformation in solid-earth physics. In the late 1990s, GEONET, a 1,200-point GPS observation system, was constructed by the Geospatial Information Authority of Japan, and is now the standard observation tool for detecting crustal deformation in this country. In the context of geodesy, people cannot perceive the "world" without those observations. Observations provide an image of the world through figures, which geodesists always work with. Those figures represent the geodetic "visibility" within the territory of "installability," that is, the area where geodesists can put their observation tools, and define the territory of "fact" in geodesy. Based on those figures, geodesists often infer the condition of the earth's interior, which is regarded as a territory both invisible and hypothetical. That activity is called "modeling." The second chapter presents several examples of modeling based on data collected through the aforementioned observational methods. The first case involves modeling of plate motion in southwest Japan, based on GEONET crustal deformation data. In that case, a geodesist first shows the GEONET data as arrows on a map, then constructs a model of plate motion in that region of Japan, finally deducing the calculated value from the observational data on the map. As a result, the observation arrows are almost entirely eliminated. That case demonstrates how geodesists use observational data, figures, and models to show that the hypothetical model

(View PDF for the rest of the abstract.)