Quantitative estimation of intrinsic induced polarization and superparamagnetic parameters from airborne electromagnetic data

The primary aim of my research is to improve the characterization of induced polarization (IP) responses in airborne electromagnetic (AEM) survey data. The principal objectives are to test alternative methodologies for quantitative modeling and inversion to extract the spatial variation of IP parameters using the inductively thin-sheet model. The methods tested first fit, by nonnegative least squares, an AEM decay to the early delay time data, using thin-sheet basis functions. This modeled AEM decay is assumed to represent the IP source. It is then convolved with a few Cole-Cole models spanning the range of parameter sensitivity to get IP basis functions appropriate for the AEM excitation. Method 1 fits a linear sum of several AEM basis functions plus one IP basis function at a time and chooses the model with least-fitting error at late delay times. Method 2 fits a linear sum of several IP and several AEM basis functions. Both methods fit IP affected airborne data well, with normalized fitting errors being reduced by a significant factor when IP affects the data and is taken into account. Using penalty weights, superparamagnetic (SPM) effects can be simultaneously estimated in the fitting process. Without such weighting, SPM and IP parameter estimations are unstable. Cole-Cole models predict that the sensitivity of inductive airborne IP collected at 25 or 30 Hz base frequency indicates little overlap with galvanic ground IP collected with a 0.125 Hz waveform. Many easy IP sulfide targets with IP physical properties determined by ground surveys are predicted not to have a detectable airborne IP response. Clays, however, are predicted to have a small detectable background response that for airborne data would not be well-fitted by a single Cole-Cole response.