CHAPTER IX - Linearity Analysis

Abstract

This chapter discusses linearity analysis, the major objective of which is to characterize the sensor transfer function. A remote target can be characterized in four nearly independent domains—spatial, spectral, temporal, and polarization. It therefore follows that the calibration must characterize the sensor responsivity in these four domains. The responsivity of a spectroradiometer is a multivariable functional relationship between the incident flux and the instrument output. All the relative terms are defined on the basis of a linear response. Thus, where each of these terms is evaluated theoretically or experimentally, it is assumed that the sensor output signal is linearly related to input flux. It is necessary to linearize the data before evaluating the terms. Many detector–preamplifier systems tend to be nonlinear over part of the useful dynamic range. Also, various nonlinear schemes are occasionally used to extend the dynamic range of systems. It is the shape of the transfer function that must be obtained, rather than its absolute value. It is very convenient to model the relative transfer function as a mathematical equation that describes the entire dynamic range of the sensor output. The absolute transfer function differs from the relative transfer function by a constant. The chapter also discusses the problems that arise in the interpretation of transfer function data.