<title>Simulating speckle polarization backscatter signatures from actively illuminated satellites</title>

Abstract

Speckle backscatter form actively illuminated satellites contains information on object size, shape and orientation and on the optical properties of the object materials. This speckle information is useful for object discrimination and classification tasks. In particular, the polarization properties of these materials may provide unique signatures that enhance the discrimination between certain objects. This simulation investigation explores the use of polarization information in speckle backscatter to form useful polarization signatures. The simulation uses detailed polarization renderings of objects and coherent field propagation techniques to form pupil plane speckle polarization components fields characteristic of the object materials' polarization properties. The fields are then converted to four speckle intensity fields using a four- channel polarimeter model and are spatially resolved with an array of detectors. Polarization signature information is obtained from Stokes parameters and Mueller matrices formed from these speckle measurements. The object polarization signatures considered include depolarization, diattenuation and retardance. These parameters are calculated for several object models to illustrate their possible use in object discrimination and to investigate their sensitivity to perturbing effects. The perturbing effects of detection noise, passive unpolarized background and atmospheric turbulence are considered. Spatial and temporal averaging of the Stokes parameter fields is shown to reduce the effect of these errors on the polarization parameters yielding improved signature measurements. Estimation and removal of unpolarized components from the speckle measurements is also shown to improve signature accuracy.