Geophysical methods are widely used in the earth sciences to obtain information about regions of the subsurface that we cannot directly sample. Determining the way in which the measured geophysical parameters are related to the material properties of interest (e.g. lithology, porosity, fluid content) is the objective of rock physics research. Much of the research to date has focused on understanding the relationship between geophysical properties and material properties in relatively small, homogeneous laboratory samples, under controlled laboratory conditions. The critical question is how best to apply these results to the interpretation of geophysical data from large regions of the subsurface – the "real world" – where there are numerous additional complexities associated with the spatial heterogeneity of geologic systems.
Of specific interest in our research are the relationships between elastic wave velocities and electrical properties of porous geological materials, and the level of water saturation. Studies based on laboratory measurements have clearly shown that the lab-scale relationship between elastic or electrical properties and water saturation is highly sensitive to pore-scale and sample-scale fluid distribution. When we consider geophysical field measurements, with a frequency of measurement different from the laboratory and a significantly increased "sample size", we suggest that the spatial heterogeneity of the subsurface can act in an analogous way to affect the field-scale dependence of our measured parameters on saturation. That is, in order to extract information about fluid content from geophysical data, we first need to account for the heterogeneity within the sampled region. While we can numerically model simple representations of a heterogeneous region of the earth, what we really need is an imp roved quantitative description of the heterogeneity that actually exists in various geological environments. In parallel with our rock physics research, we are analyzing ground penetrating radar (GPR) images of sediments from various depositional environments. Using geostatistical methods for analysis we find that these GPR images have well-defined correlation structure, that can be treated as representative of the spatia l heterogeneity of the different depositional environments. This provides a way to start incorporating the spatial complexities of the "real world" into our rock physics.
About the Author(s)
Rosemary Knight received her BSc. and MSc. in Geological Sciences from Queen's University and a Ph.D. in Geophysics from Stanford University in 1985. She is presently a professor in the Department of Earth and Ocean Sciences at the University of British Columbia. Her principal area of research is rock physics with particular interest in the effect of pore fluids on the electrical and seismic properties of rocks. Another current area of research is the use of ground penetrating radar as a means of characterizing the subsurface. Rosemary was the Distinguished Lecturer for the Canadian Geophysical Union in 1994-95, and with her students has received Honorable Mention for the Best Paper Award at SEG's International Exposition and Annual Meeting for the past two years. In 1995 she was Vice President of the Near-Surface Geophysics Section of SEG and is now a member of the Editorial Board for The LEADING EDGE.