One-dimensional (1D) inversion problem for magnetotelluric (MT) data are addressed using two different schemes: The Levenberg-Marquardt (LM) inversion and Occam-type inversion in order to obtain information on the resolvability of each scheme in recovering the Earth’s layered model. We tested both schemes to synthetic MT data generated from a blocky and a smooth model. Standard error (SE) is used as statistical criterion for comparing the fitness between the synthetic and the calculated data as well as the test model and the recovered model. It is revealed that based on the inherent properties posed by the LM and Occam inversion, the data SE is least for Occam method for both the blocky and the smooth models (log ρa SE are 0.1167 and 0.0939 for LM and Occam, respectively; phase impedance SE are 0.0649 and 0.0432). The model fitness is achieved best by the Occam inversion (log ρ SE are 2.6504 and 1.2512).

1.
M. N.
Nabighian
(ed.),
Electromagnetic Methods in Applied Geophysics
,
Society of Exploration Geophysicist
,
1988
, pp.
131
202
.
2.
W.
Menke
,
Geophysical Data Analysis: Discrete Inverse Theory
.
New York
:
Academic Press, Inc
,
1984
3.
S.
Constable
,
R.
Parker
and
C.
Constable
,
Occam’s inversion: A Practical Algorithm for generating Smooth Models from Electromagnetic Sounding data
.
Geophysics
52
,
1987
, pp.
289
300
4.
R.C.
Aster
,
B.
Borchers
, and
C.H
Thurber
,
Parameter Estimation and Inverse Problem
,
Elsevier
,
2005
5.
W.
Srigutomo
,
E.
Agustine
,
Z. M.
Habri
,
Quantitative Analysis of Self-Potential Anomaly: Derivative Analysis, Least-Squares Method and Non-Linear Inversion
,
Indonesian Journal of Physics
, Vol.
17
no
2
,
2006
, pp.
49
55
This content is only available via PDF.
You do not currently have access to this content.