The polarization sensitivity of the Visible/NearIR (VISNIR) bands in the Joint Polar Satellite Sensor 1 (J1) Visible Infrared Imaging Radiometer Suite (VIIRS) instrument was measured using a broadband source. While polarization sensitivity for bands M5-M7, I1, and I2 was less than 2.5 %, the maximum polarization sensitivity for bands M1, M2, M3, and M4 was measured to be 6.4 %, 4.4 %, 3.1 %, and 4.3 %, respectively with a polarization characterization uncertainty of less than 0.38%. A detailed polarization model indicated that the large polarization sensitivity observed in the M1 to M4 bands is mainly due to the large polarization sensitivity introduced at the leading and trailing edges of the newly manufactured VISNIR bandpass focal plane filters installed in front of the VISNIR detectors. This was confirmed by polarization measurements of bands M1 and M4 bands using monochromatic light. Discussed are the activities leading up to and including the two polarization tests, some discussion of the polarization model and the model results, the role of the focal plane filters, the polarization testing of the Aft-Optics-Assembly, the testing of the polarizers at the National Aeronautics and Space Administration’s (NASA) Goddard center and at the National Institute of Science and Technology (NIST) facility and the use of NIST’s Traveling Spectral Irradiance and Radiance responsivity Calibrations using Uniform Sources (T-SIRCUS) for polarization testing and associated analyses and results.
A movable pixelated filter array is proposed to provide low cost, on demand polarimetry and wavefront sensing. With this concept, an optical system can turn polarimetry on and off by using a shutter to move a microgrid polarizer array in and out of the optical path of the system. This allows an optical system to operate in two modes, a non-polarimetric mode in which sensor range is maintained, and a polarimetric mode in which it is reduced. In implementing this concept, adequate knowledge of the position of the filter in the optical path and calibration procedures become critical topics. This paper discusses simulated and hardware-tested results of this invention.
This paper describes measured performance of VIIRS Flight unit 1 (F1) in response to variations in linearly polarized light
input into the instrument. Measurements and analysis show that VIIRS F1 meets specified polarization sensitivity
requirements with maximum measured degree of linear polarization (DoLP) ranging from 0.5% for the M7 band (865 nm) to
2.8% for the M1 band at 412 nm. Estimated uncertainty in the DoLP ranges from 0.l7% for the M6 band at 746 nm up to
0.37% for I2 at 865 nm. Detailed tables of measured polarization sensitivity characteristics for all spectral bands, all detectors
and both sides of the VIIRS de-rotation mirror are provided to help the VIIRS users’ community with analysis of measured
spectral radiances in the solar reflectance channels.
Polarimetry sensor development has been in work for some time to determine the best use of polarimetry to differentiate
between manmade objects and objects made by nature. Both MWIR and LWIR and 2-color staring Focal Plane Arrays
(FPAs) and LWIR scanning FPAs have been built at Raytheon Vision Systems each with exceedingly higher
performance. This paper presents polarimetric performance comparisons between staring 2562 MWIR, 2562 LWIR, 5122
LWIR/LWIR staring FPAs and scanning LWIR FPAs.
LWIR polarimetry has the largest polarimetric signal level and a larger emissive polarimetric signature than MWIR
which makes LWIR less dependent on sun angles. Polished angled glass and metal objects are easily detected using
While single band 9-11 um LWIR polarimetry has advantages adding another band between 3 and 7 um improves the
capability of the sensor for polarization and spectral phenomenology. In addition the 3-7 um band has improved NEDT
over the 9-11 um band due to the shorter detector cutoff reducing the Noise Equivalent Degree of Linear Polarization.
To gain acceptance polarimetric sensors must provide intelligence signatures that are better than existing nonpolarimetric
Infrared sensors. This paper shows analysis indicating the importance of NEDOLP and Extinction ratios.
Significant anisotropy in as-deposited CVD ZnS at several length scales has been demonstrated through investigation of
structural and optical properties. Compressive strength of cylinders of CVD ZnS oriented in the growth direction is
~50% higher than cylinders taken perpendicular to the growth direction. Lattice parameter measurements of mandrel
side (first-to-grow) material is ~0.4% smaller than growth side (last-to-grow) material in a cored sample representing
~500 hours of CVD growth, indicating significant strain along the growth direction. X-ray diffraction also shows
evidence of preferred orientations for hexagonality which differ depending on position in the growth history. In crosssection,
the cored sample shows several large bands which are correlated with different degrees of infrared absorption
and BTDF scattering. However, no universal trend is found that applies to the whole length from the mandrel to the
growth side regarding optical properties. The extinction in the visible and infrared is lower for measurements
perpendicular to the growth axis than parallel to it, possibly due to scattering from the growth bands.
A requirement for the Visible/Infrared Imager Radiometer Suite (VIIRS) is that its polarization sensitivity be 3% or less
for all VISNIR bands (412-865 nm). A test using a rotating polarizer sheet was performed on the sensor to validate this
requirement, and though the test results show that the requirement is met, they also show a large variation in this
polarization sensitivity (as much as 2%) across the field of view (FOV) in track. Though this result is unexpected, it may
be the result of natural variations in the diattenuation and retardance of the VIIRS optics as a function of field angle. To
test this theory, a raytracing model of the system was constructed using measured ellipsometric data from the VIIRS
optics, and the polarization sensitivity of the model was computed. Using the nominal ellipsometric data, good
correlation between the predicted and measured polarization sensitivity was not achieved. However, by applying small
variations to the ellipsometric data as a function of position on the optics, it was possible to achieve good correlation.
This paper gives the details of the sensor polarization sensitivity measurements, ellipsometric measurements, and
When measuring the BSDF of a surface, it is necessary to determine the instrument signature of the measurement device,
which quantifies its intrinsic scatter with no sample present. For scatter angles near the specular beam, the equivalent
BSDF of this signature is greater than the BSDF of the sample, and therefore the signature defines the minimum
measurable scatter angle. For flat samples, the signature can be measured directly, but this is not possible for curved
samples because the optical power of the sample changes the angular distribution of the signature. A method is described
in K. A. Klicker et. al.  in which the signature of the device with a curved sample is determined using a raytracing
simulation, however, few details of this simulation are given. Here we give the details of a simulation in which the
instrument signature of a commercial scatterometer was modeled. We will show that this method is effective by
comparing the modeled signature of a flat sample to the measured.
A theoretical model of optical scattering in materials consisting of densely packed spherical particles is developed that
can be used to predict its optical properties given its physical characteristics. The inputs to this model are the waveband
of interest, the complex refractive indices and particle size distribution of the materials that comprise the media
(including any contaminants), the density and sizes of any contaminants in the media, and the dimensions of the media
slab. The outputs of this model are the specular transmittance and emissivity vs. wavelength of the media, and it's
Bidirectional Scattering Distribution Function (BSDF) vs. scatter angle vs. wavelength. The results of this model are
compared to measured transmittance and BSDF data from optical ceramics comprised of densified nanopowders
(nanocomposite optical ceramics).
Particulate contamination scatter is often modeled using Bidirectional Scatter Distribution Functions (BSDFs) based
upon Mie scattering by a distribution of spherical particles. Starting with the basic model described in P. R. Spyak and
W. L. Wolfe [1,2,3,4], we improve upon it by adding multiplicative geometrical form factors. These factors prevent the
Total Integrated Scatter (TIS) from exceeding unity and ensure that reciprocity is always obeyed. Preventing the TIS
from exceeding unity is necessary for energy to be conserved in the raytrace, and obeying reciprocity is necessary to
obtain consistent results between forward and backwards raytraces. As will be shown, this improved model fits
measured data better than the previous model.
It is often necessary in optical analysis software to trace millions of rays in order to determine flux transfer, color
chromaticity, and the distributions of intensity, irradiance, or radiance onto one or more targets. The key question of
"have I traced enough rays?" cannot be answered unless the statistical error in the final simulated result is below some
measurement criteria; for instance, if you are to determine irradiance uniformity to less than 1%, having a 2% statistical
error across the target will wash out what you are trying to analyze. Some optical analysis software packages do not
provide error estimation methods, while others use error estimation algorithms having assumptions that are not valid for
all cases. This paper describes how subdividing and recombining raytraces provides a robust method for estimating
error. We will show that this error estimation technique can be used with most optical analysis packages and we will
compare it with algorithms employed currently. Example systems will be analyzed and presented.
Methods for computing bin-by-bin error estimates of 2-D illumination and chromaticity distributions generated from Monte-Carlo raytracing data will be introduced, as well as algorithms for choosing the optimal number of bins based on the desired accuracy. Methods of improving the accuracy of such distributions will also be discussed, along with methods for smoothing these results for display purposes.
Many surfaces scatter light in an anisotropic way, that is, for a normally incident beam, the distribution of scattered light varies as a function of the azimuthal angle of the scattered direction. Examples of surfaces with anisotropic scattering characteristics are brushed metal reflectors and certain types of diffusers. A model, based on an anisotropic scatter model proposed by Ward is introduced. The ability to fit this model to various sets of measured BSDF data is investigated. Raytracing simulations are performed using the fitted parameters, and the results are compared with experiments.
Interpolation routines based on polynomials, splines, linear triangulation, and distance weighting techniques are tested. Two data sets containing irregularly distributed point values with two independent variables (wavelength and angle of incidence) are used as input data. The accuracy of interpolated values at unvisited points and processing time are used as criteria to determine the merits of the various interpolation algorithms. Effectiveness of distance weighting methods was found to be largely dependent on the number of neighbors used. For both gradually and abruptly changing data, the most accurate models used squared inverse distance weighting. Linear triangulation was found to be the fastest method.
This course explains the basic principles of designing, building, and testing optical systems whose stray light performance is adequate for their intended purpose. It teaches methods to identify stray light problems in the design phase when they can be most easily and inexpensively fixed, and does not emphasize the use of any particular stray light analysis software, but rather the fundamental principles of radiometry and optical design necessary to use such software effectively. Application of the course material is demonstrated in class by measuring the stray light performance of a simple camera system and comparing the measurement to both first order estimates and detailed ray tracing results.