5 Tectonics I

Sunday, 16 October 2005
8:00 AM - 12:00 PM

THE USE OF HIGH RESOLUTION VIRTUAL OUTCROPS TO VALIDATE NUMERICAL MODELS OF GEOLOGICAL ARCHITECTURES

Paper number: 5-1
Program: Technical Sessions
Time: 8:00 AM – 8:15 AM

Authors

Abstract

Geological modelling is an important tool to study Earth architectures. We utilise various numerical modelling methods to study upper-crustal deformation process in three and four dimensions. All modelling requires validation to ensure that there is sufficient match between predictions made by the model and “real world” observations. Here we show how new methods of digital geological mapping at a range of scales from 10^-3 to 10^5 m, may be used to acquire geometric and kinematic data from well exposed examples of linked fault arrays and folded beds. These data are synthesised and interpreted using our “GAVA” workflow (Geospatial Acquisition, Visualisation & Analysis). The Durham group have aquired field data utilising a range of GPS, laser-scanning and digital photogrammetry methodologies, depending on the overall nature of the outcrop and the geospatial precision required. Differential GPS and Real-Time Kinematic GPS equipment are able to provide decimetre to millimetre precision, respectively, and allow rapid data collection of fault networks exposed in beach sections, wave-cut platforms, river-bed outcrops, etc. Data from quarry faces, and steep, vertical or overhanging cliff sections are captured using terrestrial laser scanners capable of measuring the 3D position of a tight grid of many points (typically 10^5 - 10^7), so that the surface topography of an outcrop can be captured in just a few minutes. Point-spacing is sufficiently dense for individual geological surfaces to be easily identified and picked within the point-cloud using an interactive on-screen process comparable with interpretation of seismic data.

Spatial data of structural geometries measured with GPS or picked from laser scanner point-clouds tend to be “2.5D”, rather than fully three-dimensional in character. Errors associated with the interpolation of the point data to produce 3D surfaces can be quantified. A variety of stochastic and geometrical methods can be used to extrapolate individual fault planes into (and out of) the virtual outcrop surface, to link fractures together to form a connectivity network in 3D, and to quantify the uncertainty associated with the interpolation process. Thus, the output of the GAVA analysis is a quantitative range of realistic values that can be compared directly with results of numerical modelling.