from the known integrals of f over the geodesics of metric (0.1) which join the points of the circle ρ = ρ1. His method, exposed in detail in § 3.4 of the book [4], also bases on the above-mentioned geometric fact. Due to this fact, some Volterra integral equation of second kind appears for each Fourier coefficient of the function f(ρ cos θ, ρ sin θ). This equation implies uniqueness of a solution to the problem under assumption (0.2). It was shown in § 3.7 of [4] that the linearization of the inverse kinematic problem for anisotropic media leads to the integral geometry problem for a symmetric tensor field of second degree. As was mentioned there, a solution to the latter is not unique. Theorem 3.6 of [4] states only that some part of components of a sought tensor field can be recovered under the assumption that the other components are known. The integral geometry problem for tensor fields of higher degree is of interest in its own right. For instance, it was shown in Chapter 7 of [5] that, in the case of quasi-isotropic elastic media, the problem of determining the anisotropic part of the elasticity tensor from the results of measuring the phase of a compression wave is equivalent to the integral geometry problem for a symmetric tensor field of fourth degree. The present article treats the integral geometry problem of symmetric tensor fields along the geodesics of a spherically symmetric metric. We completely describe the extent of nonuniqueness for a solution to the problem. The main result states
Read full abstract