Determining the orientation of marine CSEM receivers using orthogonal Procrustes rotation analysis Kerry Key 1 and Andrew Lockwood 2 ABSTRACT Electromagnetic receivers deployed to the seafloor for CSEM surveys can have unknown orientations because of the unavailability of compass and tilt recordings. In such situ- ations, only the orientation-independent parameters derived from the measured CSEM field vector can be interpreted, and this may result in less structural resolution than possible when the sensor orientations are known. An orthogonal Pro- crustes rotation analysis OPRA technique can be used to es- timate the full 3D receiver orientation for inline and off-line CSEM receivers. The generality of this method allows it to be easily embedded into nonlinear CSEM inversion routines so that they iteratively search for both the receiver orientation and a seafloor electrical-conductivity model compatible with the data. Synthetic tests using the OPRA method jointly with a 1D inversion demonstrate that it can recover the rotation and tilt angles to about one degree accuracy for 1D data and to within a few degrees for 2D data. Application of this method to real survey data shows good agreement with a previous orientation method that is suitable only for determining the horizontal rotation of inline receivers. CSEM data collected over the Pluto gas field offshore the northwest coast of Aus- tralia were used to demonstrate how the OPRA method can be used to orient CSEM receivers prior to inversion of only the inline electric- and crossline magnetic-field components. INTRODUCTION The marine controlled-source electromagnetic CSEM method is a tool for remotely mapping the seafloor electrical-conductivity structure. Academics invented the method nearly three decades ago to study the oceanic lithosphere. More recently, CSEM has been heavily applied for hydrocarbon exploration on the continental shelves e.g., Ellingsrud et al., 2002; Edwards, 2005; Constable and Srnka, 2007. CSEM survey data are commonly recorded by sea- floor receivers outfitted with horizontal electric- and magnetic-field sensors e.g., Constable et al., 1998, whereas more recent instru- mentation also uses vertical electric- and magnetic-field sensors. Typically, the receivers are deployed from a survey vessel and then free-fall to the seafloor. On arrival at the seabed, the electric and magnetic sensor axes will be rotated to some arbitrary horizontal ori- entation and may also be tilted vertically due to regional seafloor to- pography or local rugosity. Although in principle the sensor orientations can be recorded by electronic compasses and tiltmeters, these data are not always avail- able. Furthermore, compass data are sometimes unreliable because of distortions from nearby magnetized receiver components such as the magnetic cores used in magnetometers or the inadvertent magne- tization of the battery packs used to power the compass and data log- ger. Thus it is common to encounter CSEM survey data where at least some of the receiver orientations are either unknown or unreli- able. When the receiver orientation is unknown, one way to proceed is to decompose the data into its polarization-ellipse parameters e.g., Smith and Ward, 1974. The maximum and minimum components of the polarization ellipse can be estimated independently of the el- lipse orientation angle and have been used as robust interpretational quantities for many past CSEM surveys e.g., MacGregor et al., 2001; Weitemeyer et al., 2006. However, the polarization-ellipse orientation can be indicative of seafloor conductivity variations, par- ticularly for 2D and 3D structures that modify current flow from a basic dipole geometry. Therefore, some interpretational information is lost by ignoring this parameter, which is equivalent to ignoring the absolute orientation of the sensor axes. Mittet et al. 2007 present a technique that determines the hori- zontal receiver orientation by rotating the electric-field vector until the crossline component is minimized. Although this method has of- ten been effective for the standard inline acquisition geometry, it is not suitable for off-line receivers and can fail when the transmitter dipole does not point along the towpath e.g., when lateral sea cur- rents push the head or tail of the antenna off-line. In addition, this Manuscript received by the Editor 7 July 2009; revised manuscript received 17 October 2009; published online 21 April 2010. 1 University of California San Diego, Scripps Institution of Oceanography, La Jolla, California, U.S.A. E-mail: kkey@ucsd.edu. 2 Woodside Energy Limited, Perth,Australia. E-mail:Andrew.Lockwood@woodside.com.au. © 2010 Society of Exploration Geophysicists. All rights reserved. GEOPHYSICS, VOL. 75, NO. 3 MAY-JUNE 2010; P. F63–F70, 7 FIGS., 3 TABLES. 10.1190/1.3378765 F63