Efficient 2D Inversion of Long ERT Sections

In this work a new algorithm for the efficient 2D inversion of long ERT lines was introduced. The algorithm incorporates an experimental procedure to avoid calculation and storage of the entire Jacobian matrix. This approach speeds up the Jacobian matrix calculations and also reduces the required memory resources to store it. The efficient storage of the sparse Jacobian and Smoothness matrices and the efficient inversion using the LSQR method increase significantly the inversion speed. The application of the new algorithm to synthetic and real data sets resulted in reconstructed models of comparable accuracy to the standard approach.