Solar Light Radiometry Calibration Unit for a ScanPol Polarimeter of the Aerosol-UA Space Mission

Te Aerosol-UA space mission will study aerosol microphysical characteristics in the Earth’s atmosphere based on the multispectral scanning polarimeter (ScanPol) and imaging polarimeter (MSIP). Both polarimeters must be precisely calibrated on the ground and in orbit to provide correct measurements. Tis paper considers the results of developing an experimental device for the radiometric calibration of the ScanPol. We consider the calibration unit design and operation principle to form a luminous fux with unchanged or well-predicted characteristics in a specifed direction. Te construction of the radiometric calibration unit is based on a sun-illuminated refective difuser made from the white opal glass MS20. We evaluated the scattering and polarization characteristics of the difuser in laboratory experiments at a wide range of wavelengths. Te results suggest that the polarization properties of the difuser are negligible. Te difuser scattering parameters are close to Lambertian for illuminance conditions, which is necessary for radiometric calibration. Te calibration unit was manufactured and tested during feld observations of solar radiation. Te results will be used for its improvement, mainly to reduce the observed stray scattered radiation entering the telescopes of the ScanPol polarization state analyzer.


Introduction
Te primary purpose of the Aerosol-UA space mission is to determine the microphysical characteristics of aerosols in the Earth's atmosphere from the orbit using spectropolarimeters described in previous papers [1][2][3][4][5][6]. Te mission is planned for launch on a subpolar sunsynchronous orbit of approximately 700 km altitude and an inclination of 98°using the YuzhSat satellite platform with a tentative launch in 2025 by the Yuzhnoye State Design Ofce. Te MODIS and MISR instruments on the Terra satellite still operate on this orbit providing aerosol property measurements [7,8]. Two spectropolarimeters will be installed on the following YuzhSat platform: the multispectral scanning polarimeter (ScanPol) and the multispectral imaging polarimeter (MSIP).
Te ScanPol and MSIP polarimeters must be precisely calibrated on the ground and in orbit to ensure accurate measurements of solar radiation's intensity and polarization characteristics scattered by the Earth's atmosphere and surface with signifcant spatial coverage. To build an efective calibration procedure, polarimetric models of ScanPol and MSIP were elaborated [4]. Tese models take into account the main sources of measurement errors, such as shifts in the levels of dark currents, shifts in the photodetector's spectral sensitivity, and others. Te possibilities and methods of polarimeter calibration are outlined in papers [4,5]. Te proposed calibration methods aim to reduce the errors in determining the degree and angle of linear polarization (DoLP and AoLP) for the ScanPol to 0.15% and 0.2°, respectively.
Various factors can afect satellite instruments' characteristics during launch and fight. Launching into orbit is characterized by strong mechanical overload and vibrations in a wide range of frequencies [9,10]. During a space fight, when transitioning from the subsolar part of the orbit to the Earth's shadow and back, signifcant mechanical stresses arise in the spacecraft's structure due to sharp temperature gradients. Tat can cause the deformation of its units and equipment. Te vacuum condition contributes to the pollution of surfaces, especially optical elements, due to the outgassing of the space platform element. Irradiation by short wavelength emission and high-energy rays and bombardment by micrometeorites also afect the characteristics of satellite instruments [11,12]. Terefore, calibration of satellite instruments in fight is demanded to identify ofset for adapt their characteristics after the spacecraft launch into orbit and throughout its functional life.
Special systems are designed for on-orbit calibration of earth-based remote sensing devices. Tey often use the light from the Sun and the Moon for calibration [13][14][15][16][17][18]. For example, the MODIS, VIIRS, and MISR multiangle radiometers [15,19,20] use the Sun's radiation for radiometric calibration. Te APS polarimeter of the Glory mission, the prototype of the ScanPol, also included a unit for radiometric calibration of the instrument in orbit by the Sun,the Solar Reference Assembly [21,22].
In the process of design a calibration device, it is necessary to consider changing conditions of illumination of the structural elements of spacecraft by the Sun. Tese demands are dictated by the parameters of the space platform orbit due to the apparent movement of the Sun relative to the apparatus and the confguration of the satellite depend on them, as well as by the design and parameters of the instrument itself. Finally, the Sun illumination conditions for the ScanPol polarimeter as part of the YuzhSat satellite platform are analyzed in [23].
Tis work aims to design and test the radiometric calibration unit based on a refective difuser for the multispectral scanning polarimeter ScanPol. In Section 2, the satellite platform and Aerosol-UA instruments are briefy described. Te optical layout and construction of the calibration unit are discussed in Section 3, followed by the assurance scheme and test measurement results in Section 4 and 5. Finally, a short discussion and conclusions are presented in Section 6.

YuzhSat Satellite Platform and Complex of Scientific Equipment Aerosol-UA
Te YuzhSat microsatellite platform can acquire a payload of up to 50 kg. Te Aerosol-UA project will be implemented using a modifcation of the 30 kg platform. Te payload includes the following two devices: the high precision scanning polarimeter, the ScanPol, and the wide-feld multispectral imager polarimeter MSIP [2,5], mounted in a single frame [1,24]. Te ScanPol operates in six spectral  bands from the near ultraviolet to the shortwave infrared  spectral with wavelengths of 370, 410, 555, 865, 1378, and  1610 nm. Te ScanPol measures the three Stokes parameters I, Q, and U of the sunlight, scattered by the atmospheric aerosols and the Earth's surface, from about 200 viewing angles along a track in +50°and −60°directions from nadir [4]. Te MSIP provides images of the state of the atmosphere and the Earth's surface. Tese images are used for the cloudiness correction of the ScanPol measurements.
Using the data of both polarimeters allows for performing in-orbit calibration of the MSIP image signal by the ScanPol data. Te critical element of the MSIP is the four prisms image-divided optical unit that separates a primary input image into four images on the same image sensor. Te MSIP includes three polarizing paths and two radiometric paths to quantify Stokes parameters I, Q, and U in one shot in the broad feld of view 60°× 60°at the wavelengths 410, 555, and 865 nm and retrieve the aerosol optical depth in eight spectral wavebands. Te spatial resolution of the MSIP polarimeter is 6 km on the Earth's surface at the nadir, with 15 viewing angles in a single scene [4,5].
In Figure 1, a general view of the satellite platform with the installed payload is shown. Successful implementation of the scientifc tasks of the Aerosol-UA mission is possible if the ScanPol polarimeter measures the polarimetric parameters of solar radiation scattered by the underlying aerosols and surface with an error of 0.15%.
To maintain high accuracy in measurements of polarization parameters in orbit, ScanPol is equipped with four units for in-fight calibration [4]. Tese units are as follows: (1) the dark unit for calibration of the polarimeter signal zero value, (2) the polarizer unit that forms a linearly polarized light with a high DoLP value and a known azimuth, (3) the depolarizer unit that forms unpolarized light, and (4) the solar radiometric calibration unit that forms light with a given intensity [23]. Te frst three units maintain the high polarimetric accuracy of the ScanPol. Tese units use as input light the solar scattered radiation by the atmosphere and Earth's surface that enters the device from the nadir.
Te fourth is the ScanPol solar radiometric calibration unit, which is considered in this paper in detail. Since the Sun's radiation is well-measured, stable, and unpolarized, it is an excellent candidate to be a source of reference radiation for radiometric calibration [25]. Te unit aimed to form a reference intensity of light at the entrance pupils of the ScanPol telescopes VIS1, VIS2, IR1, and IR2 using direct solar radiation. Te radiometric calibration should be carried out in orbit periodically when the optimal sunlight conditions at the calibration unit's input are achieved [26].
Te optical properties of the unit should not be changed noticeably during the entire operation in orbit. Te spectral sunlight distortions and polarizations should not be introduced by the unit or have to be stable and well defned in the laboratory. Te design of the unit has to prevent additional stray light from entering the interior of the ScanPol both during calibration and during operational measurements.
Due to the energy and weight limitations, it is impossible to direct solar radiation from the optical path to the calibration unit by maneuvering the device or mechanically moving any of its structural elements. Te Aerosol-UA satellite will operate on a sun-synchronous orbit, which enables the positioning of the input hole of the calibration unit in an optimal way for directing solar radiation to the measurement channels once a day during the device's movement in orbit. Te calibration time is reached when the satellite crosses the terminator, and the operational measurements are not performed.
Te chosen optical scheme of the radiometric calibration unit allows for providing calibration procedures at the intersection with the northern part of the terminator when the calibration unit's optical axis is directed along the orbit. Te deviation of the axis from this direction is determined by the parameters of the orbit and the height of the Sun above the upper edge of the atmosphere during calibration. Te vertical angular size of the input diaphragm depends on the segment of orbital time (number of frames) allocated for calibration and does not exceed 2°. Te horizontal angular size of the input diaphragm is determined by the distance between the positions of the Sun on the horizon during the summer and winter solstices and does not exceed 7°. In the current layout of the ScanPol-MSIP optical assembly, the space inside the assembly to install the radiometric calibration unit is limited (see Figure 1).

Optical Layout and Components Selection
3.1. Te Optical Layout. Te radiometric calibration unit operates within a 360-1700 nm spectral range and is integrated into the ScanPol-MSIP optical assembly. Te optical confguration of the unit consists of a convex mirror and a difuser ( Figure 2). Solar radiation enters the mirror that illuminates the difuser, then the radiation scattered by the difuser passes through the window of the scanning mirror assembly and enters the telescopes of the multispectral polarization state analyzer. Te geometrical parameters of the optical system elements are determined by the calibration unit dimensions of 90 × 70 × 130 mm and by the lighting conditions of all elements' light diameters.
Another important energetical parameter of the optical system is the attenuation coefcient τ of radiation in the optical system, which is determined by the following equation: where D mir and D dif − the diameters of the mirror and difuser, correspondingly, r mir , r dif − the integral refection coefcients of the mirror and difuser. In the case of values of D mir � 32 mm, D dif � 60 mm, r mir � 0.7, and r dif � 0.9, the attenuation coefcient τ is of ∼0.18. Te calibration unit creates an optical signal at the ScanPol photodetectors with values equal to or close to the signal values received from the nadir during observations of the Earth's surface. We have presented the results of developing the ScanPol multispectral polarization state analyzer in [26], where the signal-to-noise ratio was estimated for diferent spectral ranges based on the spectral density of the energy brightness of the Earth's surface B(λ). It is known that the refectivity of the Earth's surface-the Earth's albedo   International Journal of Optics can vary from ∼0.1 to 0.7. When observing the Earth's surface with an average albedo of 0.3, the radiation attenuation coefcient at the calibration unit should be of the same order. Terefore, the value of the attenuation coefcient τ = 0.18 illustrates the agreement between signals during calibration and measurements.

Te Mirror.
Te convex mirror forms the illumination of the difuser by solar radiation (see Figure 2). Te choice of a mirror coating is mainly determined by the operating conditions in a wide spectral range and the coating's resistance to the space environment conditions. Chromium and silver are preferable for the various mirror coatings used in the 360-1700 nm spectral range. Chromium is resistant to radiation exposure and has a refection coefcient of 0.6-0.7 in the visible-infrared spectrum range. Silver-based mirror coating is widely used in space astronomical instruments and has a refection coefcient of 0.85-0.95 in the spectral range of 370-900 nm. Studies show that the degradation of surface quality is insignifcant [27][28][29].

Difuser. Geometric parameters and material options
for the radiometric calibration unit are restricted by the required angular distribution of scattered light and the difuser spectral band. For the manufacture of difusers (scatterers) with a wide operating optical spectrum range, which work for difuse refection of beams, specifcally standards of whiteness, screens from magnesium oxide (MgO), barium sulfate (BaSO 4 ), and fuoropolymer are used in laboratories in diferent countries [30,31]. Te shortwavelength limit of refectivity of these patterns is much less than 0.3 μm, although the slope of its decline depends on the compound of the pattern and on the way how it is applied to the surface. Te long-wavelength limit is larger than 2 μm, with a decrease in refectivity of a few percent in the spectrum area of approximately 1.5-1.6 μm.
Te specifed working spectral range of the ScanPol device also corresponds to some options in ceramics with difuse refection, for example, ceramics of the SOO-U6 type [32] or Accufect ® B6 [33]. Optical ceramics, or Irtran, are glass-type materials made from a polycrystalline array of various substances (for example, magnesium fuoride (MgF 2 ), magnesium oxide (MgO), and some others) by the method of hot pressing at a temperature approximately 2/3 of the melting point of the substance in a vacuum [34]. Various optical devices, including light-scattering screens, are made of these materials, mainly for the infrared part of the spectrum, which is used under high mechanical and thermal loads.
Spectralon patterns have been used for over 20 years to calibrate devices for Earth's atmosphere remote sensing and the surface from orbit [35]. Tese scattering screens are made of a fuoropolymer (polytetrafuoroethylene). In [36,37] the spectral and scattering characteristics are well described. Te refectivity of Spectralon is uniform in the spectral range of 0.25-2.5 μm. Te Spectralon characteristics of angular scattering are similar to those of the other scatterers. In the mentioned spectral range, when the angle of incidence varies between 50°and 60°, depending on albedo, the refectivity of Spectralon is close to Lambertian. As with other difusers, the Spectralon scattering indicator shows a specular component, accompanied by some degree of polarization of the scattered beams, especially for large incidence angles. Consequently, careful measurements of the angular characteristics of working samples are necessary when using such a scatterer. In space conditions, the refectivity of spectral fuxes deteriorates due to impurities, especially in the frst months, diferently in diferent spectral channels, for example, [15,20,38]. However, scattering difusers made of Spectralon are used as a convenient working tool for orbital calibration for remote sensing satellite instruments, such as MODIS and VIIRS [15], MISR [39], POSP [40], OMI [41], and others. Spectralon was used in the APS instrument of the Glory satellite [13,22]. Various auxiliary devices are used to assess and consider the degradation of such scatterers, as mentioned in [18,20,41], and special spacecraft maneuvers are aimed at the Moon [18,20,22].
Refectors made of white opal glass are used mainly as whiteness standards in the visible part of the spectrum (0.4-0.75 μm), where their spectral properties are well studied (see, for example, [42][43][44]. However, as it is shown in the cited publications, white opal glass samples of the types NS20 and POS1-POS4 which are, in particular, fabricated in Ukraine (https://ipz.com.ua/), as well as white opal glass samples of the types 2015 and 2016 manufactured in the USA, have high refectance at both limits of the indicated visual part of the spectrum. In [32], the spectral properties of the difuse refection of a sample of white opal glass MS20 in the range of 0.3-2.4 μm is given. At wavelengths of 350 nm and 2000 nm, the refection coefcient of this MS20 is approximately 0.80. Te angular properties of the MS20 difuse refection were also studied in detail by various authors and are given in many publications, particularly in [32,33,[42][43][44].
Despite the relatively low refectance of this sample compared to the Spectralon, the MS20's other scattering characteristics are comparable. Terefore, given its availability in Ukraine, MS20 is considered promising for use as a difuser screen in the ScanPol polarimeter radiometric calibration unit of the Aerosol-UA space mission.

Estimation of the Diffuser's Scattering Properties
In the radiometric calibration unit, we use the refective white opal glass MS20 (AlF 3 -SiO 2 ) based difuser with a diameter of 60 mm. We estimate its scattering properties in a series of experiments. First, the Mueller polarimeter with a wavelength of 632.8 nm was used to estimate the polarization efect and complete Mueller polarimetry. Note that the operation features of ScanPol allow polarization measurements and polarization calibration regardless of radiometric calibration [4]. Te partial polarization in the radiation scattered by the difuser will redistribute the intensities in the ScanPol channels with orthogonal polarizations but will not change the total intensity. Tis circumstance makes it possible not to consider the partial polarization of the radiation scattered by the difuser. Te experiment's geometry is shown in Figure 3, where PSG-polarization state generator and PSA-polarization state analyzer, Θ i -the angle of light incidence, and Θ s -scattering angle. Te PSG includes a beam expander that broadens the input laser beam to a 10 mm in diameter. Te measurements were carried out on the incident plain. Te design features of the measuring setup determine the minimum possible angle between the direction of the incident beam and the direction of observation as Θ i + Θ s � 10°. Te difuser is mounted at the rotating stage to change the incident angle. Te range of observable scattering angles is expanded with increasing incidence angle.
Te angular dependences of Mueller matrix elements were measured for the cases when the light incident the difuser at angles of Θ i = 0°, 15°, and 30°. A measurement error of normalized Mueller matrix elements δm ij was 1.5%. Te results demonstrated that the values of only the normalized Mueller matrix diagonal elements were far from zero. Te absolute values of these elements slightly decrease when the scattering angle grows. Other matrix elements are close to zero within the error range, except for m 12 . Te dependencies of the mentioned nonzero Mueller matrix elements are shown in Figure 4. Terefore, the difuser does not produce any instrumental polarization under the given conditions since the DoLP for totally depolarized input is determined by the following equation: Depolarizing efect of the difuser for linear polarized input light at wavelength 632.8 nm can be expressed by the following equation: DoLP ≤ ���������� m 22 2 + m 33 2 ≈ 0.14, since the Mueller matrix of the difuser is diagonal and the maximum number of diagonal elements' modules is about 0.14. In (3), an equal sign should be used for totally polarized linear input light. Te difuser's depolarization efect grows when the scattering angle increases, as shown in Figure 4. Second, the bidirectional refection distribution function (BRDF) values were obtained for the MS20 sample when the laser sources with wavelengths 450 nm, 532 nm, 632.8 nm, and 1064 nm were used for the geometry shown in Figure 3 with modifcation. Te reference intensity channel was introduced to compensate for the sources' signal fux, and the PSG and PSA were removed. Te results are presented for normalized BRDF in Figure 5. Te BRDF is normalized on a value corresponding to the measured one at Θ s = 10°since it is a minimal observable scattering angle at normal light incidence (Θ i = 0°).
Te irregularities on dependencies in Figure 5 could be explained by an error in the angular positioning of the receiving part Figure 3 which was of δΘ s ≈ 1°. Tere is also some bias in the axis of the rotating stage from the axis of measurement arm rotation at which the receiving part is mounted. Te BRDF of Lambertian scattering is a strain line parallel to X-axis. In Figure 5, the scattering of the difuser is the closest to the Lambertian scattering the closer the incidence angle to 45°. In general, the scattering of the studied difuser deviates from the Lambertian scattering by less than 1% within the 20°range of scattering angle for all incidence angles at all used spectral wavelengths.

Test Measurements with the Radiometric Calibration Unit
Te experimental design of the calibration unit is presented in Figure 6. A unit of diaphragms is attached to the Aerosol-UA assembly where the convex mirror and the difuser are fxed, forming the working FOV of the calibration unit, which is of 7°in the horizontal plane and 2°in the vertical plane [23]. Additional intermediate diaphragms are also placed in the diaphragm unit to minimize stray light.
In the standard the ScanPol-MSIP optical assembly for the test experiment of the calibration unit, a scanning mirror unit, and a camera with a matrix detector for recording the light fux was installed (Figure 7(a), upper panel). Te upper part of Figure 7 schematically shows the operation of the input lenses of the ScanPol multispectral polarization state analyzer. Sunlight illuminates the difuser of the calibration unit. Te scattered radiation passes through the mirror scanning unit and enters the input lenses of telescopes VIS1, VIS2, IR1, and IR2, which focus the light beams in their focal planes, where the feld diaphragms are installed. Next, the light passes through collimators, Wollaston prisms, dichroic mirrors, spectral flters, and camera lenses and is focused on sensors. Tis part is described in detail in our paper [26].
However, we performed the calibration unit performance test at the frst step without using the multispectral polarization state analyzer to evaluate the light beams formed after the feld apertures by the scheme in Figure 7(a) in the lower panel. In this case, the Hartmann diaphragm replaces the input lenses (telescopes) with apertures equal to their light diameters. Finally, a focusing lens with the matrix detector is installed behind the diaphragm.
Te Hartmann diaphragm can be installed in position 1 or 2. Position 2 is preferable because it is more convenient and the diaphragm will not glare in the sunlight. Te aperture of T in the focal plane of the lens is essential. It ensures the telecentric path of the beams, and the Hartmann diaphragm separates the working areas of the difuser. If the detector is defocused so that the camera lens builds an image of the scatterer, we will get a matching of all scheme elements. Terefore, we model the operation of the input lenses of the ScanPol multispectral polarization state analyzer and evaluate the light fuxes after the feld diaphragms. Te focusing lens with the matrix detector was also placed in the standard assembly of the ScanPol-MSIP unit and installed on the roof of the laboratory building. Finally, the entrance hole of the external diaphragm was directed to the Sun, with the subsequent recording of the solar radiation fuxes (Figure 7(b)).
To evaluate the work efciency of the calibration unit in dependence on the change in the angle of the position of the igure 4: Dependencies of normalized by m 11 nonzero Mueller matrix elements for the white opal glass MS20 (in relative units) at scattering angles for incidence angles Θ i � 0°, 15°, and 30°of light wavelength 632.8 nm. 6 International Journal of Optics Sun, the FOV was scanned along the central axis of the external aperture ( Figure 8). Quantitative characteristics of the radiometric calibration unit recorded by the detector (Figure 7(a)) depending on the Sun angle position are presented in digital numbers W proportional to averaged light intensity shown in Table 1.
In Table 1, the W values are presented, which are proportional to light intensity at an enter in four ScanPol   Figure 7(a) (lower panel). During calibration, the sunlight enters the FOV calibration unit, illuminates the difuser (β min ), and is passes through the center of the FOV (β 0 ), next to the edge of the FOV (β max ). A matrix detector measures the signal value W in each of the described positions.

. Discussion and Conclusions
Te design features of the YuzhSat platform and the limited volume allocated to the satellite's payload made it possible to design a calibration unit based only on a refective difuser that determines the illumination of the difuser in a range of incidence angles of 14.5°÷ 21.5°. Te angle values depend on the change of season. At the same time, the scattered light from the difuser comes to the input window of the scan unit from a direction near the scattering angle in the range of −1°÷ 1°. Te input window looks directly at the difuser. Tese angles come up so that we scan the difuser and get multiple signal readings. As mentioned above, the scattering properties of the considered difuser are close to the Lambertian scattering within mentioned incidence and scattering angles, at least for the spectral range of 450-1064 nm. Terefore, providing the size of light spots on the difuser are larger than the FOV of telescopes, it will be possible to signifcantly reduce intensity fuctuations in the measuring channels when the Sun is displaced relative to the input window of the calibration node during radiometric calibration.
Te core part of the radiometric ScanPol calibration is to obtain a calibration coefcient A r in the calibration model of the polarimeter that was described in detail in [4]. We demonstrate (see Figure 4) that the difuser based on the white opal glass MS20 at wavelength 632.8 nm does not produce stray instrumental polarization under considered conditions when the diameter of the illuminated difuser's area is at least 10 mm. At the same time the stray instrumental polarization of the convex mirror can be signifcant for of-normal light incidence cases. Te theoretical analysis, involving Fresnel equations, carried out for silver and chromium mirrors, shows the DoLP of the refected sunlight that incidents the mirrors at the angle 17°doesn't exceed 0.6% for silver mirrors and does not exceed 5% for chromium mirrors for the ScanPol spectral range of 360-1700 nm. However, the difuser will signifcantly reduce the DoLP. For instance, for wavelength 632.8 nm, the diffuser reduces DoLP for approximately seven times. Terefore, the efect of the instrumental polarization of the convex mirror is also negligible for the considered calibration unit design. In addition, as mentioned in Section 4, even a signifcant partial polarization of radiation scattered by a diffuser (for example, at other wavelengths) will only redistribute the intensities in ScanPol channels with orthogonal polarizations. But it will not change the total intensity I, which should be used to calculate the radiometric coefcient A r . In this case, the following expression should be used: where RD 0 and RD 90 is the zero-signal bias compensated digital values from the output of the ScanPol channels with analyzers azimuths α � 0°and 90°correspondingly; q and usecond and third Stokes parameters of solar radiation scattered by difuser; q inst and u inst − second, and third Stokes parameters of stray instrumental polarization introduced by the scan system of ScanPol; K1 − isotropic gain coefcient for 90°ScanPol channel [4]. Te values q inst , u inst and K1 are calibrated regardless of A r . In equation (4), |q inst q+ u inst u| ≤ 10 −4 for DoLP of radiation scattered by difusor is up to 2%. Te problem of the possible difuser's characteristics degradation due to contamination can be reduced by the preliminary cleaning procedure. Te parasitic scattering inside the radiometric calibration unit will introduce errors into the calibration, which must be minimized, for example, by including diaphragms. Computer modeling of additional apertures and blends, their sizes, shapes, and positions, which can signifcantly reduce scattered parasitic radiation in the calibration unit, was also previously explored.
Finally, our estimation of the difuser's characteristics is qualitative. Nevertheless, it gives us reason to assume that the calibration unit will precisely provide on-orbit radiometric calibration for ScanPol. During the unit's operation test, the measurements of light beams from the Sun were carried out according to the ScanPol scheme of operation. We obtained the relative intensity of the light beams entering the input of the multispectral polarization state analyzer for four channels at diferent positions of the Sun. We also scanned the entire FOV of the calibration unit. Te maximum irregularity of the light beam intensity from one edge of the FOV to the other is less than 1% for VIS channels and less than 5% for IR channels. Such a sufciently large range of irregularity is primarily associated with the scattered parasitic radiation entering the input of the optical channels of the Stokes parameter converter and requires further, more detailed research.
Tese results are necessary for specifying ways to improve the parameters of the radiometric calibration unit and for a more detailed analysis of the real parameters of the scanning unit. At the next stage, we plan to analyze the light beams in the working spectral channels of ScanPol in the range of 360-1700 nm using a multispectral polarization state analyzer and measure the light beams with real photodetectors.

Data Availability
Te data used to support the fndings of this study are available from the corresponding author upon request.

Conflicts of Interest
Te authors declare that they have no conficts of interests regarding the publication of this paper.
International Journal of Optics 9