Abstract
Magnetotellurics (MT) can be applied to monitor resistivity change at depth that is for example due to fluid injection in enhanced oil recovery or CO2 storage. The observed MT data changes at the surface may be insignificant, but the effect can be enhanced using the layer stripping method, i.e. calculating MT data changes that would be observed at depth based on data from the surface. Two well-known formulas for MT 1D forward modeling were reformulated to allow for calculation of the impedance at depth based on the impedance at the surface. We applied the layer stripping technique to synthetic data associated with models that were representative of a likely CO2 storage site. We also used an equivalent model and the Monte Carlo approach to estimate the sensitivity of the method to cope with the uncertainty of the host model and the input data. The layer stripping calculation has the greatest uncertainty at short periods, where the real and imaginary parts of the complex impedance tend to be equal, i.e. an homogeneous medium response. The layer stripping technique should be used with great caution based on a relatively precise 1D host model.