A new method for anthropogenic noise remova in the ELF band based on ICA technique

A new method to separate anthropogenic noise from natural electromagnetic field measurements is presented in this communication. This new method is based on independent component analysis technique, which allows us to face the problem of separating a measured signal into independent components (ICs) that verify statistical independence. If the signal is the result of the mixture of independent sources that verify statistical independence, then the ICs are scaled versions of the original sources, which means that we can solve the problem of blind source separation. This technique has been successfully applied to extremely-low frequency band magnetic field measurements taken in Sierra Nevada, Spain, to remove anthropogenic noise from measured natural electromagnetic field data.