| The frequency-domain marine controlled source electromagnetic (CSEM) method has recently become a tool in determining subsurface resistivity related to hydrocarbon formations in the deep water environment. In shallow water, this frequency-domain method is subject to airwave saturation that severely limits sensitivity to targets at depth. It has been suggested that time-domain CSEM may offer an improved resolution to these deep targets, as well as increased sensitivity to resistors in the presence of the airwave. In order to examine and test these claims, a modeling code has been developed for computing time-domain responses for layered 1D models with arbitrarily located and oriented transmitters and receivers. The code extends the open-source frequency domain code Dipole1D by efficiently computing the time-domain, step-on, and impulse responses by Fourier transformation of the frequency-domain kernels. Impulse responses are used along with pseudo-random binary sequences (PRBS) to generate synthetic time-domain data. A realistic noise model and waveform scaling effects are then applied to synthetic step-on, PRBS, and the frequency-domain SIO "Waveform D" data generated from this code. Wiener deconvolution is applied to recover impulse responses from the PRBS data, allowing for a systematic examination of the sensitivity and resolution of time-domain and frequency-domain CSEM to representative targets of interest for offshore hydrocarbon exploration. These studies suggest that there is no large advantage to time-domain techniques, as previously suggested, and rather that the frequency-domain Waveform D should give better results in the presence of noise for the shallow marine setting. |
