The Study On Polarization Correction Of Imaging Spectrometer For Space-Borne Ultraviolet Atmospheric Remote Sensing

The scattered sunlight contains essential information about the globaldistribution of atmospheric constituents, so the measurements of the spectralirradiance of the light from the sun and spectral radiance of the scattered light fromthe earth can provide information about a number of trace gases and aerosols tomonitor the global greenhouse effect, the change of ozone layer, gas pollution et al.Light reflected from the Earth's atmosphere is linearly or partially linearly polarizedbecause of scattering of unpolarized sunlight by air molecules and aerosols. Thespace-borne ultraviolet spectrometers always contain optical components such asdiffraction gratings, mirrors whose reflectance or transmission are often polarizationdependent, so the instrument are normally sensitive to the state of polarization of light.In order to measure the radiance accurately, the polarization response of theinstrument must be eliminated from the polarization-sensitive measurement. Thespace-borne spectrometers have2options to treat the atmospheric polarization of theincoming light. Either the polarization information is destroyed by scrambling, as inthe OMI and SBUV/2instruments or the polarization of the incoming light has to bemeasured to correct for the polarization dependence of the instrument, as in theGOME and SCIAMACHY. The polarization characteristics of the space-borneultraviolet spectrometers, polarization measurements in the ultraviolet, the scramblingand measuring the polarization of incoming light methods are explored.In this dissertation the polarization characteristics of the scattered sunlight whichis the targeted object of ultraviolet hyperspectral ozone profile spectrometers(UHOPS)is analyzed, accounting for rayleigh scattering, mie scattering and multiple scattering.The radiance and polarization model calculations are performed using the LibRadtransoftware package. The radiometric measurement equation considering for polarizationis presented using mueller matrix calculus. The uncertainty in the on-orbit radiance measurement is presented, and the simulation results show that the measurementsuncertainty due to polarization is small for instruments with polarization responsesless than0.01.The linear polarization sensitivity of space-borne ultraviolet spectrometers ispresented using mueller matrix calculus. The diattenuation of the main portion of theinstruments (mirror and grating) is analyzed. A set of ultraviolet polarization testsystem is designed with250-500nm wavelength range to test optical components andinstruments. The uncertainty of optical components measurements is about1.5%, andof the optical instruments test is about1.6%.The double plate horizontal-vertical depolarizer, rotation depolarizer, improvedLyot depolarizer and the dual Babinet compensator depolarizer which are commonused in remote sensing are evaluated, and the spatial average of their Mueller matrixare gained. The depolarization properties of these depolarizers are described. Theaberrations of these depolarizers are analyzed using geometrical optics calculating,mtf and software simulation to evaluate the image degradation of imagingspectrometers.The polarization characteristics of the scan system and small groove spacinggrating that are the main sources of linear polarization sensitivity of UHOPS areanalyzed and the mueller matrix for the two main observations are gained. Accordingto polarization characteristics of UHOPS, the polarization correction algorithm ofinstruments is set. The uncertainty of polarization factor is presented and themathematical approach to obtain the polarization values from the instrumentobservations shows that the error on the measured radiance is within2%. The resultsof solar backscatter ultraviolet spectrometer experiment are within the specification ofmeasurement accuracy and the polarization correction algorithm satisfies therequirement.