Three-dimensional ground-penetrating radar ͑GPR͒ data are routinely acquired for diverse geologic, hydrogeologic, archeological, and civil engineering purposes. Interpretations of these data are invariably based on subjective analyses of reflection patterns. Such analyses are heavily dependent on interpreter expertise and experience. Using data acquired across gravel units overlying the Alpine Fault Zone in New Zealand, we demonstrate the utility of various geometric attributes in reducing the subjectivity of 3D GPR data analysis. We use a coherence-based technique to compute the coherency, azimuth, and dip attributes and a graylevel co-occurrence matrix ͑GLCM͒ method to compute the texture-based energy, entropy, homogeneity, and contrast attributes. A selection of the GPR attribute volumes allows us to highlight key aspects of the fault zone and observe important features not apparent in the standard images. This selection also provides information that improves our understanding of gravel deposition and tectonic structures at the study site.Anew depositional/structural model largely based on the results of our analysis of GPR attributes includes four distinct gravel units deposited in three phases and a well-defined fault trace. This fault trace coincides with a zone of stratal disruption and shearing bound on one side by upward-tilted to synclinally folded stratified gravels and on the other side by moderately dipping stratified alluvial-fan gravels that could have been affected by lateral fault drag. When used in tandem, the coherence-and texture-based attribute volumes can significantly improve the efficiency and quality of 3D GPR interpretation, especially for complex data collected across active fault zones.

Three-dimensional modeling of marine controlled-source electromagnetic (CSEM) data is vital to improve the understanding of electromagnetic (EM) responses collected in increasingly complex geologic settings. A modeling tool for simulating 3D marine CSEM surveys, based on a finite-difference discretization of the Helmholtz equation for the electric fields, has been developed. Optimizations for CSEM simulations include the use of a frequency-domain technique, a staggering scheme that reduces inaccuracies especially for horizontal electric-dipole sources located near the seafloor, and a new interpolation technique that provides highly accurate EM field values for receivers located in the immediate vicinity of the seafloor. Source singularities are eliminated through a secondary-field approach, in which the primary fields are computed analytically for a homogeneous or a 1D layered background; the secondary fields are computed using the finite-difference technique. Exploiting recent advances in computer technology and algorithmic developments, the system of finite-difference equations is solved using the MUMPS direct-matrix solver. In combination with the other optimizations, this allows accurate EM field computations for moderately sized models on small computer clusters. The explicit availability of matrix factorizations is advantageous for multisource modeling and makes the algorithm well suited for future use within an inversion scheme. Comparisons of simulated data for (1) 1D models to data generated using a 1D reflectivity technique and (2) 3D models to published 3D data demonstrate the accuracy and benefits of the approach.