Open AccessMantle viscosity inference: a comparison between simulated annealing and neighbourhood algorithm inversion methods
Author(s) -
Piana Agostinetti N.,
Spada G.,
Cianetti S.
Publication year - 2004
Publication title -
geophysical journal international
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.302
H-Index - 168
eISSN - 1365-246X
pISSN - 0956-540X
DOI - 10.1111/j.1365-246x.2004.02237.x
Subject(s) - inversion (geology) , simulated annealing , algorithm , inference , mantle (geology) , neighbourhood (mathematics) , statistical physics , computer science , geology , geophysics , mathematics , physics , mathematical analysis , artificial intelligence , seismology , tectonics
SUMMARY Two direct search methods, simulated annealing and neighbourhood algorithm, are applied to the inversion of the viscosity profile of the mantle using relative sea level time‐histories for the Hudson Bay region. In problems characterized by a low‐dimensional model space ( N d = 2 in this study), the two inversion methods show comparable performances. When a larger number of dimensions is involved (specifically N d = 6 ), we directly show that simulated annealing is less effective than neighbourhood algorithm in overcoming the obstacles that are found in the model space when our specific data set is employed. This study confirms that modifications of the conventional Monte Carlo inversion method, such as simulated annealing and neighbourhood algorithm, are viable tools to determine the viscosity profile of the mantle, which, until recently, has been mainly tackled by means of linearized techniques.