測地学会誌
Online ISSN : 2185-517X
Print ISSN : 0038-0830
ISSN-L : 0038-0830
東北日本の地殻歪(I)
―解析方法―
藤井 陽一郎杉田 要中根 勝見
著者情報
ジャーナル フリー

1986 年 32 巻 1 号 p. 32-42

詳細
抄録

A least square adjustment is developed in order to deduce earth's horizontal strain from displacement vector. Let (xi, yi) (i=1, 2, …, P) be the coordinates of p triangulation stations. These coordinates are determined with respect to the centroid of the geodetic figure consisting from p stations. Let d (xi, yi) (i=1, 2, …, P) be the displacement vectors at the triangulation stations. These vectors are given as the deviations from translation vector of the centroid that is estimated by any suitable method from original displace ment vectors. We assume homogeneous strain u inside the geodetic figure as
ut=[εx, εy, εxy, ω], (1)
where εx and εy, are normal strains, γxy, shear strain in the sense of engineering shear strain, and w rotation. Displacement vectors d are
dt=[dx1, dy1, …, dxp, dyp] . (2)
All the components are assumed to be independent.
The relation between u and d is given from elasticity theory
d=Bu, (3)
where B is design matrix and has a form
B=(4)
Least square solution of (3) is
u=(BtPdB)-1BtPdd . (5)
Pd is weight matrix of d. Variance-covariance matrix Σ(u) of u is
Σ(u)=σ02Ωuu (6)
Cofactor matrix Quu is, using notation N=(BtPdB)
Quu = N-1 (7)
and a posteriori reference variance σ02 is
σ02=VtPdV|c-u=VtPdV|γ. (8)
V is residual of displacement vectors calculated using the determined strain u. c is number of observation equations, u number of unknowns.
Let maximum and minimum principal strains be ε1 and ε2, and azimuth of maximum principal strain ε1 be α, then(9)
Vector of principal strain E is defined as
E[ε1, ε2, α] (10)
Variance-covariance matrix of E is Σ(E), and this is propagated from 3×3 elements inΣ(u), Using Jacobian Jeu, Σ(E)=JeuΣ(u)euJeut(11)(12)Dilatation Δ and maximum shear strain γmax is obtained as Δ=ε12 λmax=ε12(13)Variance of Δ and γγγmax, covariance between Δ and γmax are σΔ2, σΔ 2, σ2γmax, and σΔγmax(14)where, σε12 and σε22 are variances of ε1 and ε2, and ε1ε2 is covariance between ε1 and ε2.

著者関連情報
© 日本測地学会
前の記事 次の記事
feedback
Top