A.

Overall Instrument Design

As shown in Fig. 2 the SENTINEL 4 flight H/W consists of an Optical Instrument Module (OIM) and a sunshield, placed on the Nadir panel of the MTG-S platform, as well as two pieces of electronic H/W, the Instrument Control Unit (ICU) and the Scanner Drive Electronics (SDE), which are placed inside the satellite and are connected to the instrument by a dedicated harness.

Fig. 2.

Left: SENTINEL 4 configuration together with the IRS instrument on the MTG-S satellite. Right: SENTINEL 4 OIM from a radiator- and IRS-side perspective.

The OIM, with a cross-section as shown in Fig. 3 (Left), has two main view ports which can be selected with a scan mirror: an Earth observation view port and the Calibration Assembly (CAA) view port, the latter being used for both sun viewing and White-Light-Source (WLS) viewing. A single two-axis scan mirror is used for two purposes: 1) to switch between CAA port and Earth port, and 2) to scan the Earth in E/W direction in the Earth observation mode. The Earth view consists of two sub-ports, because due to the bi-annual yaw flip manoeuvre of the MTG-S satellite, pointing towards the target Earth coverage area around Europe requires opposite orientation of the scanner N/S axis in summer and winter. For this reason the aperture cover mechanism has two flaps, a summer and a winter opening flap, which are adjacent in N/S direction.

Fig. 3.

Left: SENTINEL 4 OIM in a cross-section parallel to the main optical plane. The blue arrow indicates the Earth radiance path, the yellow arrows the sun irradiance path, the red allow the star viewing path. The green arrow indicates the observation path from scanner towards the telescope, which is common to all viewings (Earth, sun, stars, WLS) and to all spectral bands (UV, VIS and NIR). Right: Core optics layout from the same perspective. The UV-VIS path is in blue, the NIR path in red.

The CAA port also has two principal sub-viewing options, which can be selected by the setting of the diffuser wheel mechanism of the CAA. In the first setting the sun is observed through one of two selectable diffusers (“nominal” and “reference” diffuser). In the other setting the flat-field WLS is observed for instrument transmission & pixel response diagnostic purposes. Deep space, and in particular star viewing is enabled by extensions of the Nadir baffle clear FOV towards the East. This allows pointing the slit sufficiently far to the east such that the entire slit points beyond Earth to deep space. In summary the instrument has 3 external views: Earth (radiance) observation, star viewing (both through the same port), and Sun (irradiance) observation.

Next in the light path after the scan mirror comes the Telescope-Spectrograph-Assembly (TSA), which features a common telescope for all spectral bands, followed by a dichroic beam splitter combined with the slit assembly, which separates the NIR from the UV-VIS band(s). The UV and VIS bands are covered by a single spectrometer all the way down to the detector, respectively Focal Plane Assembly (FPA). The relatively narrow NIR band (750-775 nm) is also covered by a single spectrometer and detector/FPA.

The light of the UVVIS band is dispersed by a transmissive grism (combination of a transmissive grating and a prism) in the UV-VIS spectrometer, and by a reflective grating in the NIR spectrometer. The NIR grating assembly includes a thermally activated co-registration compensation mechanism, which allows fine tuning of the NIR detector image position in spatial (N/S) direction and thereby optimum N/S co-alignment between UVVIS and NIR (cf. Section III).

The two FPAs feature CCD detectors acquiring the spectral channels in one direction and the spatial sampling in the orthogonal direction. In addition, each FPA includes a dedicated LED light source, which is optimized for homogeneous in-band illumination of the respective detector. These LED light sources integrated in the FPAs are used for calibration purposes (cf. Section III). The detector is the core element of the FPA where the detected light is converted into an analogue electrical signal and subsequently processed by the Front End Electronics (FEE) where the analogue signal is converted into digital data and sent via a ChannelLink interface to the instrument control unit (ICU). The ICU adds dedicated ancillary telemetry to the image data such that they are self-contained, avoiding the need to re-assemble the observation and telemetry data on ground. In addition, the ICU is responsible for operating and controlling the instrument in its various modes.

The instrument thermal design utilizes a concept based upon a combination of passive cooling and active (closed-loop) heater control for temperature stabilization. While the core optics will be maintained and controlled at 293 K, the CCDs will be operated at about 215 K, via passive cooling. In-orbit temperature stabilities (short- and long-term) on the order of 0.1-0.2 K of the key optical elements and the detectors are necessary and achieved, in order to fulfill the requisite radiometric, geometric and spectral accuracy specifications. This presents a major challenge for the instrument thermal architecture and control, taking into account the geo-stationary orbit with continuously varying sun illumination.

B.

Optical Design

The SENTINEL 4 core optical design is shown in Fig. 3 (Right). It consists of the following optical modules, designed to be independently manufactured and aligned: Scanner, Telescope Module (including beamsplitter & slits), UVVIS and NIR Spectrograph Modules. The main end-to-end performance parameters driving the optical architecture are polarization (pol. sensitivity, pol. spectral & spatial features), straylight and co-registration.

Since the system level spatial co-registration requirements are defined on an absolute and not on a knowledge accuracy basis, very good co-registration has to be achieved by design. For the optics design this means ultra-low optical distortion and also extremely good matching of the effective focal lengths of the UV-VIS and the NIR optical path.

The planar symmetry of the core optics, the on-axis lenses, and a general optimization for low angles of incidence (e.g. on scanner) and low angles of dispersion, allow achieving almost neutral polarization behavior by design. These optical architecture features are in part also enabling factors for the low optical distortion. Certain optical elements still inevitably have increased polarization sensitivity and spectral features (e.g. the grism), and therefore a depolarizing element, the polarization scrambler, is introduced before these elements in the optical path. The pre-optimization of the optical architecture towards low polarization effects has two advantages regarding this polarization scrambler: 1) The front optics, including scan mirror and telescope, features sufficiently low polarization effects that the scrambler can be introduced after these elements. This leads to a significantly smaller scrambler, which has great advantages in terms of manufacturability; 2) A rather weakly depolarizing scrambler, which is directly associated with a very small degradation of the optical point-spread-function (i.e. image quality), is sufficient for achieving the system level polarization requirements.

Another main optimization criterion of the optics design is the minimization of straylight: The main sources of straylight are scattering from surface roughness and particulate contamination, as well as ghosts. The term ghosts encompasses a variety of false light effects, such as multi-reflections from anti-reflection (AR)-coated surfaces, unwanted reflections from mechanical surfaces outside the nominal optical path (lens mounts, optical stops, baffles, etc.), and unwanted or multiple diffractions from the dispersers. All these straylight sources are mitigated by minimization of the number of optical elements. Furthermore, ghosts are suppressed by dedicated fine optimization of the optics design in all areas, e.g. spectrograph and disperser architectures, as well as by a sophisticated straylight baffling architecture, for example in the beamsplitter-slits-assembly and in the FPAs.

C.

Detection Chain

The UV-VIS and NIR detectors, which are developed by e2v (UK), are both frame-transfer CCDs featuring frame shift along the spectral direction (cf. Fig. 4). The NIR detector architecture is simpler, with a single frame, a single shift register and a single read-out port, while the UVVIS detector is divided into two spectral frames (effectively two individual CCDs), UVVIS1 and UVVIS2, with a frame split at about 340 nm. In addition, the UVVIS1 shift register has two read-out ports with different gain, the high gain being used for the low-signal spectral ranges below about 316 nm, and the low gain for wavelengths above. Furthermore, the UVVIS2 is divided into 4 individual shift registers and read-out ports. This architecture not only allows that the three main frames UVVIS1, UVVIS2 and NIR have individual gains, but also that their signal integration times can be individually adjusted. Thus optimum system SNR performance can be reached taking into account the particular spectral dynamics of the Earth radiance scenes. Furthermore, the frame periods of UVVIS1, UVVIS2 and NIR are set in multiples of the same time increment. This allows for a synchronized operation scheme (integration, frame transfer image-to-memory-zone, read-out) of the three main frames, which is used in all nominal SENTINEL 4 measurements.

Fig. 4.

Development models of the UV-VIS (left) and NIR (right) detectors

This synchronized UVVIS1-UVVIS2-NIR-sequencing avoids signal distortions by electro-magnetic-interference, which is mandatory in order to achieve the required radiometric performances, but this synchronization scheme is also a major challenge for the overall detection chain architecture and design.

A.

Overview

This subsection presents a non-exhaustive list of particular challenges faced in the SENTINEL 4 subsystem development, out of which one example, namely the UV-VIS grism (cf. bullet 2 of the following list), has been selected as an example for more detailed discussion in subsection B:

B.

Example: Grism Structure

While the groove density of the grism is fixed (to about 500 lines/mm) by the optical design of the UV-VIS spectrograph, the actual fine design of the periodic grating structure of the grism is a (multi-dimensional) design degree of freedom with significant influence on the optical performances of the grism itself. The main performances are throughput (i.e. efficiency of the employed grism diffraction order) and polarization.

Fig. 5.

Left: Etching test samples by ZEISS for the selected grism structure design. Right: Simulated grism efficiency (throughput) spectra for several example grism structure implementations, including the manufactured test samples. Photograph and simulation data are courtesy of ZEISS.

A particular constraint for the throughput is that it has to be on one hand as high as possible in the lower UV spectral range (305-320 nm) in order to achieve the system level SNR requirements, and on the other hand below certain limits at higher wavelengths in order to avoid detector saturation, especially in the VIS spectral range where the Earth radiance scenes feature very high signal levels. Regarding polarization there are several relevant performance parameters, chiefly polarization sensitivity and polarization spectral features.

The actual grism is manufactured by Carl Zeiss Microscopy GmbH (ZEISS) as a sub-contractor of the grism assembly supplier Jena-Optronik GmbH (JOP). In an iterative process, the optical performances of many possible grating structures have been simulated by ZEISS (cf. Fig. 5), subsequently integrated by Airbus Defence and Space in system level performance analyses, and evaluated in order to provide feedback to JOP/ZEISS for further structure optimization. The main challenge in this context is to fulfil simultaneously the needs for high throughput in the UV and for low polarization spectral features, because these two parameters tend to be anti-correlated, meaning that structures with high throughput tend to also have increased polarization spectral features. Out of these iterations, an optimum grism structure design has been selected.

For the selected design, etching tests have been performed by ZEISS (cf. Fig. 5) and the obtained grism structures have been mechanically characterized. These tests confirmed the manufacturability of the selected design, and also show that sufficiently low manufacturing tolerances are achieved. The selected grism design achieves impressively high efficiency on the order of 77% in the low UV, and allows to achieve on system level very good polarization spectral features performance of about 0.040% (at worst-case wavelength). At the time of writing of this paper, preliminary optical measurements of efficiency and polarization performances are available for the manufactured test samples, confirming the predictions of the simulations, and the manufacturing and characterization of full-size grism structure breadboards is on-going.