|
| | Abstract | | Magnetotelluric (MT) method, one of the electromagnetic survey methods, is used for resource exploration, active fault survey and so on. Subsurface resistivity structure is generally estimated by inversion algorism, based on the least-square scheme using observed apparent resistivity and phase. The optimal resistivity model is a kind of approximated solutions, therefore, the reliability of estimated model should be discussed. However, reliability tests in the previous works are usually qualitative and subjective.In this research, we develop a new way of the reliability test of optimal model objectively and quantitatively based on the principle component analysis (PCA) and Monte-Carlo methods. The principle components show the common and major structure in resistivity model. By changing the principle component in each direction randomly, we can modify the optimal model slightly for checking the acceptable range of model parameters. We applied our method to the synthetic MT data (TE mode) and the inversion result, and finally confirmed that the obtained reliability map well corresponds to known sensitivity trend of MT inversion in TE mode. |
|
|
|
| |
|