The future ESA Earth Observation Sentinel-4/UVN is a high resolution spectrometer intended to fly on board a Meteosat Third Generation Sounder (MTG-S) platform, placed in a geostationary orbit. The main objective of this optical mission is to continuously monitor the air quality over Europe in near-real time. The Sentinel-4/UVN instrument operates in three wavelength bands: Ultraviolet (UV: 305-400 nm), Visible (VIS: 400- 500 nm) and Near-infrared (NIR: 750-775 nm). Two dedicated CCD detector have been developed to be used in the Focal Plane Subsystems (FPS), one for the combined UV and VIS band, the other covering the NIR band. Being a high resolution spectrometer with challenging radiometric accuracy requirements, both on spectral and spatial dimensions, an effect such the Random Telegraph Signal (RTS) can represent a relevant contribution for the complete system accuracy. In this work we analyze the RTS effect on data acquired during the FPS testing campaign with qualification models for the Sentinel-4/UVN detectors. This test campaign has been performed in late 2016. The strategy for the impact assessment of RTS is to measure the effect at room temperature and then to extrapolate the results to the at instrument operational temperature. This way, very-long lasting data acquisitions could be avoided since the RTS frequency is much lower at cryogenic temperatures. A reliable technique for RTS effect detection has been developed in order to characterize the signal levels amplitude and occurrence frequencies (flipping rate). We demonstrate the residual impact of the RTS on the global In-Orbit Sentinel-4/UVN instrument performance and products accuracy.

RTS effect detection in Sentinel-4 data

Driven by the need of always more accurate models, space optics instrument-based observations push constantly towards high accuracy measurements that require an excellent knowledge of the instrument. To achieve this, current classical technologies are limited by the complexity of current instruments, calling for disruptive technologies to take over. Therefore, Airbus is currently integrating Artificial Intelligence (AI), responding to the call for new concepts. Here Airbus takes benefit of deep learning to detect complex patterns that would otherwise be impossible to properly characterize classically, opening the door for completely novel characterization paradigms and enabling manifold accuracy improvements. This work first focuses on obtained results on the detection of random telegraph signals (RTS) of CCD detectors under tests. By training a convolutional neural network (CNN) with RTS data, it has been possible to setup an algorithm achieving 20x faster data processing while increasing accuracy, providing unprecedented fast and performant RTS characterization. In another domain, multi-reflection-induced ghost stray light have been also characterized using CNN. Here, Airbus uses simulated data from optical software to generate 2D ghost maps used to train an algorithm capable of segmenting individual patterns. We show in this work that the appropriate architecture with optimized hyper-parameters achieves 97% accuracy. These ground-breaking results pave the way for a complete characterization of optical instrument ghosts that were so far neglected because of their complexity. It hence enables in the future more performant straylight correction algorithms as well as providing extended freedom in the design of space optical instruments.

/pdf/space-optics-instrument-optimization-and-characterization-2eu7nb2zac.pdf

Space optics instrument optimization and characterization with artificial intelligence

The Sentinel-4 instrument is an imaging spectrometer, developed by Airbus under ESA contract in the frame of the joint European Union (EU)/ESA COPERNICUS program. SENTINEL-4 will provide accurate measurements of trace gases from geostationary orbit, including key atmospheric constituents such as ozone, nitrogen dioxide, sulfur dioxide, formaldehyde, as well as aerosol and cloud properties. Key to achieving these atmospheric measurements are the two CCD detectors, covering the wavelengths in the ranges 305 nm to 500 nm (UVVIS) and 750 to 775 nm (NIR) respectively. The paper describes the architecture, and operation of these two CCD detectors, which have an unusually high full-well capacity and a very specific architecture and read-out sequence to match the requirements of the Sentinel- 4 instrument. The key performance aspects and their verification through measurement are presented, with a focus on an unusual, bi-modal dark signal generation rate observed during test.

The Sentinel-4 detectors: architecture and performance

The Sentinel 4 instrument is an imaging spectrometer, developed by Airbus under ESA contract in the frame of the joint European Union (EU)/ESA COPERNICUS program with the objective of monitoring trace gas concentrations. Sentinel 4 will provide accurate measurements of key atmospheric constituents such as ozone, nitrogen dioxide, sulfur dioxide, formaldehyde, as well as aerosol and cloud properties. Sentinel 4 is unique in being the first geostationary UVN mission. The SENTINEL 4 space segment will be integrated on EUMETSAT's Meteosat Third Generation Sounder satellite (MTG-S). Sentinel 4 will provide coverage of Europe and adjacent regions. The Sentinel 4 instrument comprises as a major element two Focal Plane Subsystems (FPS) covering the wavelength ranges 305 nm to 500 nm (UVVIS) and 750 nm to 775 nm (NIR) respectively. The paper describes the Focal Plane Subsystems, comprising the detectors, the optical bench and the control electronics. Further the design and development approach will be presented as well as first measurement results of FPS Qualification Model.

The Sentinel 4 focal plane subsystem

The Sentinel-4 payload is a multi-spectral camera system which is designed to monitor atmospheric conditions over Europe. The German Aerospace Center (DLR) in Berlin, Germany conducted the verification campaign of the Focal Plane Subsystem (FPS) on behalf of Airbus Defense and Space GmbH, Ottobrunn, Germany. In this publication, we will present in detail the temperature dependence of dark signal for the CCD 376 (NIR) from e2v. Dark current is strongly temperature-dependent and is based on the thermal excitation of electrons in the conduction band. During the testing the temperature of the detectors was varied between 215K and 290K. Different models were examined, and corresponding deviations determined. Presenting the dark current by means of an Arrhenius plot, it can be shown, that the activation energy is about half of the band gap. As an important result it could be shown that the temperature dependence can be described by two activation energies.

Temperature Dependence of Dark Signal for Sentinel-4 Detector

The METimage instrument is part of the EPS-SG (EUMETSAT Polar System Second Generation) program. It
will be situated on the MetOp-SG platform which in operation has an objective of collecting data for
meteorology and climate monitoring as well as their forecasting. Teledyne e2v has developed and
characterised the CMOS VISNIR detector flight module part of the METimage instrument.
This paper will focus on the silicon results obtained from the CMOS VISNIR detector flight model. The
detector is a large multi-linear device composed of 7 spectral bands covering a wavelength range from 428
nm to 923 nm (some bands are placed twice and added together to enhance the signal-to-noise
performance). This detector uses a 4T pixel, with a size of 250μm square, presenting challenges to achieve
good charge transfer efficiency with high conversion factor and good linearity for signal levels up to 2M
electrons and with high line rates. Low noise has been achieved using correlated double sampling to
suppress the read-out noise and give a maximum dynamic range that is significantly larger than in standard
commercial devices.
The photodiode occupies a significant fraction of the large pixel area. This makes it possible to meet the
detection efficiency when front illuminated. A thicker than standard epitaxial silicon is used to improve NIR
response. However, the dielectric stack on top of the sensor produces Fabry-Perot etalon effects, which are
problematic for narrow band illumination as this causes the detection efficiency to vary significantly over a
small wavelength range. In order to reduce this effect and to meet the specification, the silicon manufacturing
process has been modified. The flight model will have black coating deposited between each spectral
channel, onto the active silicon regions.

Characterisation results of the CMOS VISNIR spectral band detector for the METimage instrument

The detectors of the Sentinel 4 multi spectral imager are operated in flight at 215K while the analog electronics is operated at ambient temperature. The detector is cooled by means of a radiator. For thermal reasons no active component has been allowed in the cooled area closest to the detector as the passive radiator is restricted in its size. For thermal decoupling of detector and electronics a long distance between detector and electronics is considered ideal as thermal conductivity decreases with the length of the connection. In contradiction a short connection between detector and electronics is ideal for the electronic signals. Only a short connection ensures the signal integrity of both the weak detector output signal but similarly also the clock signals for driving the detector. From a mechanical and thermal point of view the connection requires a certain minimum length. The selected solution serves all these needs but had to approach the limits of what is electrically, mechanically and thermally feasible. In addition, shielding from internal (self distortion) and external distorting signals has to be realized for the connection between FEE(Front End Electronics) and detectors. At the time of the design of the flex it was not defined whether the mechanical structure between FEE and FPA (Focal Plane Assembly) would act as a shielding structure. The physical separation between CCD detector and the Front-end Electronics, the adverse EMI environment in which the instrument will be operated in (the location of the instrument on the satellite is in vicinity to a down-link K-band communication antenna of the S/C) require at least the video output signals to be shielded. Both detectors (a NIR and a UVVIS detector) are sensitive to contamination and difficult to be cleaned in case of any contamination. This brings up extreme cleanliness requirements for the detector in manufacturing and assembly. Effectively the detector has to be kept in an ISO 5 environment and additionally humidity has to be avoided - which does not comply with the usual clean-room atmosphere. This paper describes how in Sentinel 4 the given challenges have been overcome, how the limited load drive capability of the detector component has been considered on a flex length of about 20 cm (7.87 in) and how EMC shielding of the highly sensitive analog signals of the detector has been realized. Also covered are design/manufacturing aspects and a glance on testing results is provided

Michael P. Skegg

Papers

RTS effect detection in Sentinel-4 data

The Sentinel-4 detectors: architecture and performance

Characterisation results of the CMOS VISNIR spectral band detector for the METimage instrument

Realization of the electrical Sentinel 4 detector integration

The Sentinel 4 focal plane subsystem