PlanetPhysics/Geiger's Method

From testwiki
Jump to navigation Jump to search

Geiger's method [1] is an iterative procedure using Gauss-Newton optimization to determine the location of an earthquake, or seismic event. Originally his method was developed to obtain the origin time and [[../Epicentre/|Epicentre]] but it is easily extended to include the [[../FocalDepth/|Focal Depth]] for [[../Hypocenter/|Hypocentre]] determination.

Given a set of M arrival times ti find the origin time t0 and the hypocentre in cartesian coordinatios (x0,y0,z0) which minimize

the objective [[../Bijective/|function]]

F(𝐗)=i=1Mri2.

Here, ri is the difference between observed and calculated arrival times

ri=tit0Ti,

and the unknown [[../Parameter/|parameter]] [[../Vectors/|vector]] is

𝐗=(t0,x0,y0,z0)T

In [[../Matrix/|matrix]] form (1) becomes

F(𝐗)=𝐫T𝐫

The Gauss--Newton procedure requires an initial guess of the sought parameters, denoted here as

𝐗*=(t0*,x0*,y0*,z0*)𝐓,

which are then used to calculate the adjustment vector

δ𝐗=(δt0,δx0,δy0,δz0)T

in

(1)𝐀T𝐀δ𝐗=𝐀T𝐫.

The Jacobian matrix 𝐀 is defined as

𝐀=(r1/t0r1/x0r1/y0r1/z0r2/t0r2/x0r2/y0r2/z0rM/t0rM/x0rM/y0rM/z0).

The partial derivatives are evaluated at the initial guess, or trial vector, 𝐗*. Equation (45) can be rewritten as

(2)𝐆δ𝐗=𝐠.

Using (46) and an initial guess 𝐗* an adjustment vector can be calculated. The initial guess can then be updated 𝐗*+δ𝐗 and used as the inital guess in the next run of the [[../RecursiveFunction/|algorithm]]. In this manner the sought parameters 𝐗 can be determined to some tolerance.

All Sources

[1] [2] [3]

References

  1. ↑ 1.0 1.1 Geiger, L., Probability method for the determination of earthquake epicenters from the arrival time only. {\it Bull. St. Louis Univ.} vol. 8, pp. 60-71.
  2. ↑ Lee, W. H. K. and Stewart, S. W. {\it Principles and Applications of Microearthquake Networks,} Academic Press, New York. 1981
  3. ↑ Gibowicz, S. J. and Kijko, A. {\it An Introduction to Mining Seismology,} Academic Press, New York. 1994.

Template:CourseCat