| |
Inversion of this data set involved modeling the earth with a 40 x 40 x 10 grid of cells (16,000 cells), each with horizontal dimensions of 100 x 100 metres. A "minimum structure" (or "smoothest") model was specified for the model objective function, a reference model with susceptibility equal to zero was assigned, and positivity was included as a further constraint. Remanent magnetism and self-demagnetization effects were neglected.
| Data shown to the right were assigned an error of 2% plus 5 nT, and the inversion converged to the desired misfit after 25 iterations.
Move your mouse over the figure to see the predicted data generated by the model recovered by 3D inversion.
Almost all of the misfit occurs at the outcrops because the cells are simply too large to properly model the rapid variability of magnetic susceptibility in the top few 10's of meters. However, the data outside these outcrops are adequately reproduced, as can be seen by comparing observed and predicted data in these two figures. |
 |
The recovered model shown in the interpretation section provides the suseptibility of each subsurface cell in terms of a contrast because the reference model was set to zero, and data were provided as anomaly values. Recovered susceptibility contrasts had a maximum of 0.31814 and a minimum of zero.
| © UBC-GIF
January 9, 2007
|
|
|
|
|
