Lateral Model Correlation for 1D Inverse Modeling of Large Data Sets

In this paper, we present novel developments of the Lateral Parameter Correlation (LPC) method for invoking lateral smoothness in model sections of one-dimensional (1D) inversion models. The LPC method is a three-step procedure consisting of (1) individual inversion of all soundings, (2) a lateral correlation procedure, and (3) a final individual inversion. The lateral correlation involves solving a simple constrained inversion problem with a model covariance matrix that ensures lateral smoothness. The method separates inversion from correlation and is much faster than methods where inversion correlation are solved simultaneously. In the new developments, a strictly horizontal correlation can be performed, thereby avoiding the model artifact sometimes seen when correlating along layers. Furthermore, a solution to the intractable computation times arising with large data sets is formulated employing a tessellation of the plane and an averaging scheme within subareas that reduces the size of the numerical LPC inversion problem while maintaining correct correlation within a very large area. A field example shows the improvements obtained with the strictly horizontal correlation. The LPC method is very flexible and is capable of correlating models from inversion of different data types including information from boreholes, and it lends itself easily to ‘embarassingly parallel’ computation.