Methods for Improving the Stability of Electromagnetic Geophysical Inversions

In this paper, two methods are presented for improving the stability of geophysical inversions. The first method uses the weighted least squares technique to reduce the effects of additive noise in the measurement data. This is accomplished by giving a lesser weight to those measurements that are thought to be more noisy. Some simulations are presented that demonstrate the effectiveness of this method. In addition to the above method, we explore the possibility of using post-reconstruction image processing to reduce the effects of noise in the measurements. A new technique, which is referred to here as mini-mux partitioning, is developed. This method, in conjunction with image filtering, is used to locate high contrast anomalies in a region. Examples are presented that show the superiority of this partitioning scheme over other techniques.