In the Optical Assembly the incident radiance and irradiance are collected and split according to wavelength and focussed on the detector modules of the two optical channels. The Optical Bench consists of a single aluminium structure in which all elements are mounted. Attached to this aluminium structure are the two Detector Modules. The Optical Assembly Thermal Hardware consists of a thermal radiator and heaters. The thermal radiator is thermally connected to the detector modules via cold-fingers and flex-links. The heaters are attached to the Optical Bench structure and the CCD detectors.
The OMI CCD's are thinned, backside illuminated CCD's from EEV (currently E2V) , model EEV 55 with 576 rows and 780 columns. The detectors have a HfO2 anti-reflection coating to obtain a high quantum efficiency, especially in the UV. This coating is optimised to have minimum reflectivity and best quantum efficiency at 270 nm.
OMI CCD description
The OMI detectors are frame transfer CCD's. This means that the CCD's have three sections: an image section, a storage section, and a read-out register. The pixels are 22.5 × 22.5 micron in size and there are no regions in between the pixels that are not sensitive to light, which is important for accurately detecting spectral features of trace gases. Each image part of the CCD consists of 580 (spatial) × 780 (spectral) CCD pixels, of which 480 × 750 are used for Earth and sun measurements. The storage region of the CCD is an exact copy of the image part of the CCD. The read-out register has 17 extra pixels on each end.
The CCD uses 480 rows of 750 pixels for the Earth and sun measurements. The wavelength information is distributed along a CCD row, the spatial information (viewing angle) is distributed along a CCD column. The image section is in the focal plane of the spectrograph and after exposure the frame is transferred rapidly to the storage region. This is the snapshot mode of the CCD detectors. During transfer illumination of the detector continues, causing smearing of the detected signals, that will be corrected for in the Level 0 to 1B software. After transfer to the storage region, the rows are transfered to the read-out register and the pixels are read out.
Binning of images
As described above the rows of the storage area are clocked to the read-out register. In this process, 4 or 8 rows are added in the read-out register, before clocking them out over the read-out amplifier. This on-chip binning is needed to reduce the instrument data rate. Binning of 4 rows results in the spatial zoom-in mode pixel size of 13 × 12 km², binning of 8 rows in the global mode pixel size of 13 × 24 km². For the UV-1 these ground-pixels sizes are 13 × 24 km² and 13 × 48 km², respectively.
Exposure times and co-adding of images
The nominal exposure time of the CCD is 0.4 seconds. In order to improve the signal-to-noise and to reduce the downlink data rate, 5 sequentially taken exposures are co-added on-board. Co-adding is done by the electronics in the Electronics Unit (ELU). Co-addition of 5 exposures results in a total 2 seconds interval in which measurements are taken. This results in the 13 km spatial resolution in the flight direction. This 2 seconds interval is called the Master Clock Period (MCP), which is the basic time interval in the ELU.
The extreme rows and columns of the image area are masked and not sensitive to light. These masked rows are used to determine the dark signal (columns) and exposure smear (rows).
Straylight, dark-current, offset and gain-switching
Between the image areas on the CCD detectors and the masked areas above and below the image areas there are a number of rows that are not masked, but that are not illuminated directly by the optical system. These stray light rows are used to estimate and correct for spatial stray light. The stray light rows above and below the image are binned to achieve better signal-to-noise.
Information on the electronics offset and dark current can also be obtained from the read-out register. The read-out register within the CCD has 17 excess pixels compared to the CCD itself. The offset can be measured via initial register readout preceding all image data. As the register is cleared just before this, it also contains the offset data.
The detector electronics is equipped with a 12 bit Analog to Digital Converter (ADC). The Electronics Unit (ELU) allows for switching the electronic gain as a function of the CCD column number (wavelength) to optimise signal over the available ADC range. Switching as a function of column (wavelength) can be used to obtain more electronic gain for low radiance levels. The highest gain is used for the dark current, smear rows and stray light rows above and below the directly illuminated image areas on the CCD detectors. The register readout uses the same gain switching as the image area. In this way electronic offset values for all gains are obtained. OMI allows to select four gain values: x1, x4, x10 and x40.
It has been observed in flight that high energetic protons (>10 MeV) trapped
in the magnetic field of the Earth can cause damage to the individual OMI CCD detector
As a result, the dark current of the individual CCD pixels and the average dark current of the complete CCD detector will
increase over the lifetime of OMI. To maintain sufficiently low dark current
towards the end of the mission, the nominal detector operating
temperature is 265 K. In order to account for the pixel-to-pixel variable increase in dark current of pixels
that have been affected by a proton hit, a dynamical background subtraction scheme has been devised. In this
scheme the background signals are updated each day.
In order to reduce the impact of proton radiation damage in flight a considerable amount of aluminium shielding
has been added to minimise detector degradation.
Pixels that have been affected by a high-energetic proton can show two effects: a considerable increase in dark current and so-called Random Telegraph Signal (RTS) behavious. RTS behaviour manifests itself as jumping of the signal level between two or more quasi-stable levels. Pixels that have been hit by a proton can feature modest RTS behavious, but in exceptional cases also severe RTS behaviour. The latter behaviour is detected by the in-flight calibration and monitoring activities and pixels that can no longer be used for science purposes are flagged as bad pixels. With the above precautions the in-flight proton radiation damage is within acceptable limits and does not seriously affect the quality of the OMI level-1b and level-2 data products and the objectives of the OMI mission.
In order to reduce the risk of ice-formation on the surfaces of the detectors or the optical components, the instrument was outgassed at an elevated optical bench and detector temperature of 303 K for a time period of more than 6 weeks immediately after lauch.
The instrument optical components introduces optical distortion of the image that is focused on the CCD. The optical distortion has two components. Firstly, optical distortion in the swath direction (i.e. curvature of lines of equal viewing direction). Secondly, optical distortion in the spectral direction (i.e. curvature of lines of equal wavelength). The optical distortion in the swath direction introduces spatial aliasing: pixels which are part of the same row (i.e. pixels which are part of the same spectrum) having different viewing angles. As a result of binning of the rows in the read-out register the optical distortion in the spectral direction potentially introduces a decrease in the spectral resolution. However, this has a negligible effect on the spectral resolution. Therefore the optical design has been optimised to minimise to the distortion in the spatial direction. The optical distortion in the spectral direction is more pronounced than the distortion in the spatial direction. It is more pronounced for the UV channel than for the VIS channel. The instrument calibration corrects for these distortion effects.