Statistical Information Content of Radiation Measurements used in Indirect Sensing

Abstract

The information content of radiation measurements used in inferring profiles is defined as a reduction in uncertainty in the estimation of a profile after the measurements are introduced. The information is shown to depend directly on the kernel of the equation of radiative transfer, the covariance matrix of experimental error, and the covariance matrix of the a priori statistical information. Calculations based on the minimum rms inversion method are applied to the indirect probing of the vertical temperature distribution by microwave measurements of oxygen thermal emission. Choice of optimum location of measurements is discussed and comparison of the proposed method with that of Twomey is given.