Abstract

This paper discusses the problem of uniqueness and the compromise between resolution and variance when determining plasma densities by inverting inaccurate interferometric data from cylindrically symmetric refractionless plasmas or other similar phase objects. Methods are given to show what unique information about plasma structure can be provided by a finite set of interferometric data, and how to determine the best possible resolution given that the variance of the estimate is to be less than a specific value. The problem of structural appraisal is also resolved by computing plasma densities which reproduce the data to within desired errors; the flattest model, and models produced through linear programming techniques are important in this respect. Linear programming techniques are discussed in detail because they can be used to generate models which are constrained to be non-negative, and also to provide numerical bounds on the smallest supremum of the plasma density or its gradient.

© 1979 Optical Society of America

Symmetrical and asymmetrical ideal cylindrical radiation transformers and concentrators

D. R. Mills and J. E. Giutronich
J. Opt. Soc. Am. 69(2) 325-328 (1979)

Statistical inference of particle populations from scattered light

W. L. Anderson and S. Y. Shen
J. Opt. Soc. Am. 69(12) 1684-1690 (1979)

2s² 2pk-2s 2pk⁺¹ transitions in the F i to Be i isoelectronic sequences of chromium, iron, and nickel from tokamak-produced plasmas

C. Breton, C. De Michelis, M. Finkenthal, and M. Mattioli
J. Opt. Soc. Am. 69(12) 1652-1658 (1979)

Reconstruction of Three-Dimensional Refractive Index Fields from Multidirectional Interferometric Data

D. W. Sweeney and C. M. Vest
Appl. Opt. 12(11) 2649-2664 (1973)

Matrix formulation of the reconstruction of phase values from phase differences

B. R. Hunt
J. Opt. Soc. Am. 69(3) 393-399 (1979)