Viking Satellite UV Imager
As mentioned before, Canada’s, and more specifically the University of Calgary’s, contribution to the Viking satellite program was the Ultraviolet Auroral Imager. It was a fundamental part of the mission and to the progression of auroral research as it made dayside aurora accessible for analysis. It also proved the importance of auroral observations on other data. There were various considerations that influenced the design of the instrument:
Experience gained from previous imagers such as the one flown on ISIS-II.
Observations were made in the vacuum ultraviolet part of the spectrum where there was minimal solar interference. Auroral observations in the visible part of the spectrum were ruled out for both dayside and nightside areas of the oval because of the intense background coming from the earth and sunlit atmosphere.
A fast reflecting optical system with 25° and minimum number of optical elements was required. To achieve this, an f/1 inverse Cassegrain Burch camera involving two spherical mirrors was used which resulted in a spherical focal surface with its radius of curvature being 22.4 mm.
Separate imaging at the two most prominent UV auroral emissions, LBH and OI, was performed in order to derive information on the precipitating electrons responsible for their emissions from their relative intensities and spatial distributions.
Curved microchannel plate (MCP) intensifiers were required to conform to the spherical focal surface. Open intensifiers were used with the photocathode material deposited directly onto the front surface of the MCP. Because of the composition of the cathodes for the chosen wavelength ranges, the intensifier assemblies needed to be in a vacuum or dry nitrogen environment at all times.
Multistage MCPs presented many resolution lifetime problems so a single-stage MCP was utilized as its gain was more than adequate and avoided the problems that the multistage version faced. The intensifier was directly coupled to the detector with fibre optic blocks for optimal transfer efficiency and the fibre optic blocks also provided the desired mapping of the intensifier’s spherical surface to the plane of the CCD.
To achieve the desired sensitivity, one-second exposures were necessary from the spinning satellite. This was done by synchronously stepping the charge accumulating in the CCD with the image motion due to the satellite spin. This meant that the CCD had to be aligned with the spacecraft equatorial plane within ±0.2°.
Interactive control of the instrument pointing and image size was required and done to ensure the effective use of the available telemetry bandwidth and meet the scientific objectives. CCD exposure was triggered at any point during the spacecraft’s spin using a reference pulse from either the spacecraft limb or the sun sensors. Flexible CCD readout electronics allowed for the image dimensions to be shrunken or enlarged and to shift the field of view within the optical system’s bounds. These adjustments could all be made in nearly real-time by the instrument operator however, it required extensive viewing and planning at the ground telemetry site in Kiruna, Sweden.
The camera was kept sealed and secured tightly until orbit was achieved. The components of the system were anchored inside a graphite fiber epoxy composite tube (GFEC) which was sealed with a dust cover. The camera was kept under a dry nitrogen purge until a few hours prior to launch, which meant that when it was removed, the camera was sealed with a slight positive pressure. However, the dust cap was not released until several hours after a stable orbit was achieved to minimize the risk of damage to the instrument. The system was able to yield a good optical throughput and sensitivity in the UV range with its FOV of 25°, compact f/1 lens, and reflecting surfaces throughout the system. It did well in minimizing spherical and off-axis aberrations. The image intensifier limits the overall resolution to approximately 0.076°.
The optics being designed concentrically ensured that the image formed is free of angular distortion because the external field is accurately mapped onto the spherical image surface. A tapered fiber optic assembly was inserted between the MCP and the CCD to remove degradation caused by projection of the surface onto the CCD. The entrance and exit surfaces of the assembly are made to match the corresponding surfaces of the MCP and the CCD, while the interior surfaces are made to compensate for distortions caused by the projections.
The camera had to operate in temperatures as low as -30°C, but final focusing adjustments could only be made at room temperature. Therefore, it was desirable for the camera to stay in focus over this temperature range. The sensitivity of the Burch design to the separation of the primary and secondary mirrors was minimized due to the GFEC housing. This was because it had a near-zero thermal expansion along the camera’s optic axis. The mirrors were made from Zerodour which also has a very low expansion coefficient. Because of the choice in materials, for best results, the camera stayed in focus through the operating temperature range.
The imager involved two separate cameras with one of the passbands, frequency bands in which a signal is transmitted by a filter with no reduction in force, ranging from 1235-1600 Å and the other from 1340-1800 Å. Camera 1 was the shorter wavelength camera that transmitted the strong 1304 Å and 1356 Å OI multiplets as well as some N2 LBH bands. Camera 0, on the other hand, cut off at the OI lines but passed the LBH as well as some contribution from 1493 Å NI line. These passbands were achieved in the following ways:
Camera 0 had a BaF2 filter and a CsI photocathode.
Camera 1 had a CaF2 filter and KBr photocathodes as well as reflective coatings on the secondary mirror.
These two types of filters provided the short-wavelength cutoff to each of the passbands whereas the photocathodes provided the long-wavelength cutoffs. Additionally, the photocathodes had very low sensitivity to scattered visible light. The filters were placed in the central hole of the secondary mirrors and due to this location and the fact that they were polished for low UV scattering, they had no impact on the image quality. The photocathodes conferring such low sensitivity is significant as it means that data collection from the dayside aurora was possible. To maximize the quantum efficiency while minimizing factors that could inhibit it in the visible range, Camera 1’s secondary mirror was overcoated with a selectively reflecting coating supplied by Acton Research Corporation. Camera 0, however, was not overcoated because it was more important for this camera to have the ability to maximize its sensitivity in the far UV rather than minimize the signal received from the visible range. The remaining mirror surfaces were overcoated with aluminum by Acton Research Corporation to maximize reflectivity in the 1250-1500 Å range.
Preliminary estimates of the imager’s sensitivity indicate that the sub kilorayleigh aurora would be detectable in a 1-second exposure which is consistent with the expected sensitivities of a few photoelectrons per pixel-spin and electro-optic gains in the range of 0.1-0.4 dn per photoelectron. A dn is a unit of 8-bit digitized video signal from the cameras. The rejection of visible light in the dayglow region was handled by the photocathodes, the attenuation through multiple scattering in the channels of the MVP, and by the aluminum overcoating on the fiber optic tapers. Preliminary observations of dayglow at 1340 Å were consistent with the expected fluxes.
A baffle system was required to prevent the scattering of ultraviolet light into the detector from illuminated surfaces in the optical path. More specifically, it prevented the sun from directly illuminating the filter, the mirrors, or the stop at the edge of the filter. This greatly reduces the amount of stray light. Despite these efforts, it is still possible for light to scatter from the interior of the baffle tube onto the filter so, to reduce the stray light levels caused by this, a series of vanes provided with knife edges was included. The remaining scatter then arises from the tips of the knife edges and multiple scatter from the fronts of the vanes to the back surfaces of preceding vanes. The diffuse and specular reflectance of the vane and baffle tube surfaces are important in controlling the amount of scatter.