1.1

Metosat Third Generation Program

The Metosat Third Generation mission involves three satellites flying in constellation in geostationary orbit above Europe (Figure 1):

Figure 1:

Metosat third generation mission involve a constellation of thee satellites.

MTG-IRS spectrometer is Fourier transform spectrometer (FTS) based on a Michelson Interferometer ([1], [2]). The theoretical background for this type of instrument is synthetized by Peter R. Griffiths, James A. de Haseth [3]. Demonstration of the interest of this kind of instrument for meteorology and atmosphere chemistry study was demonstrated with Infrared Atmospheric Sounding Interferometer (IASI) instrument, which is flying on Meteorological Operational satellite (Metop) [7].

The instrument is composed of four main sub-elements developed and manufactured in parallel (see Figure 2): The front telescope that scans the Earth and collects light from the Earth or from calibration scenes, the Michelson Interferometer Assembly, which is the core of the instrument, the back telescope that collects the flux of the interferometer and the detection and electronics assembly which contains optics, detection chain and signal processing units. The Core Spectrometer is a sub-assembly of the IRS. It is representative of the spectral behavior of the instrument. It was tested in a dedicated campaign using non-flying Engineering Models of the three involved subparts.

Figure 2:

The MTG-IRS architecture is made of four main elements built by European manufacturers. The Core spectrometer is a subpart that is representative of the spectral behavior of the instrument.

1.2

MTG-IRS spectral performances specifications overview

The spectral performances are specified in the Metosat Third Generation system requirements document [4]. The main spectral performance requirements are:

The spectral response function knowledge and stability requirements are particularly challenging. With the aim to fulfill this specification, Thales Alenia Space had to implement a new patented metrology system [8]. This system allows to know the 3D relative positioning of the two arms of the Michelson interferometer. It allows to get a subnanometric knowledge of Optical Path Difference (OPD) fluctuations for each sounding points (i.e. for each detector pixel) of the instrument field of view. The validation of this technology is the core topic of this paper.

Table 1:

Spectral band and spectral resolution specifications [4]:

Table 2:

Spectral Response Function knowledge and stability specifications:

Table 3:

Spectral lines/SRF positioning specifications [4]:

Figure 3:

The rendering of a spectral line in the measured spectrum depends on the SRF shape, that’s why it is important to have a nearly perfect knowledge of those functions.

3.1

Test set-up: Core spectrometer illuminated by laser sources

Test set-up is presented in Figure 6. The idea is to produce a spatially uniform monochromatic scene at core spectrometer entrance with the aim to measure the instrument line shape (ILS).

For this purpose, two laser beams (one for each spectral band) are injected into an integrating sphere placed at core spectrometer entrance. The test equipment needs to be carefully dimensioned, in particular:

Figure 6:

Core Spectrometer spectral test set-up: two MWIR and LWIR laser beams are injected in an integrating sphere to produce a spatially homogenous monochromatic scene to enlighten the Core spectrometer entrance. The integrating sphere and the core spectrometer being placed into a vacuum chamber. The laser is injected into the chamber through porthole.

3.2

Comparison of measured line instrument line shape with prediction

For each pixel of each spectral band, the ILS (core spectrometer response to the monochromatic light) is computed and compared to the expected shape. The obtained average ILS is, for each pixel, a temporal average of 236 ILS measurements. The result is illustrated in Figure 7.

The obtained match shall be compared with system specification (recalled in Table 2). For this purpose, the specification reference spectrum is convolved with the two ILS (measured and predicted) and the difference between the two results is expressed in radiometric unit (as defined in [5] => Noise Equivalent delta Temperature definition). The following results are obtained:

Note 1: This result supposed that the ILS shape error does not depend on the wavelength, which is not rigorously exact.

This result is slightly better than expected and allows to demonstrate the quality of the SRF estimation model that will be used to provide the SRF information to MTG-IRS spectra end users.

Figure 7:

Comparison between measured core spectrometer ILS and prediction. The match appears to be visually nearly perfect. Remote side-lobes are perfectly described by the model even for lobes 10000 smaller than the main peak. The measurement curve on this plot is corresponding to the ILS measured on a MWIR pixel. It is an average of 236 ILS measurements.

3.3

Line shape positioning stability results

During Core Spectrometer test campaign, the stability of the ILS position was also checked.

The accurate spectral positioning of MTG-IRS spectra will be ensured thanks to a dedicated spectral calibration process. The spectral scale positioning will be checked regularly with the aim to perform the scale correction.

However the raw instrument positioning mismatch needs to be stable enough between two calibrations to ensure the quality of the correction.

This point was checked on core spectrometer. In LWIR, no ILS position instability was identified. In MWIR, a 1ppm fluctuation was identified (see Figure 8), but this fluctuation is corresponding to the real laser line position fluctuation, as measured by MWIR laser lambda-meter. Hence, this fluctuation is not due to the Core Spectometer, but to the test setup. The stability of the core spectrometer itself does not bring any significant instability.

Figure 8:

Measurement of Instrument Line Shape (ILS) position evolution during one hour in MWIR (blue) compared with laser source wavenumber stability measured by a lambdameter (green). The two curves are matching. No line positioning instability due to the core spectrometer is detected.

4.1

Set-up for interferometer assembly micro-vibration sensitivity test

As for core spectrometer test, the Interferometer Assembly is illuminated with a spatially uniform monochromatic scene produced by a laser injected in an integrating sphere. The interferometer output signal is collected by a ground optical system that simulates IRS back optics (Back Telescope Assembly and Detection and Electronics Assembly) with the aim to measure ILS.

The Interferometer is placed on a heavy structure that is voluntarily vibrated using a shaker (Figure 9).

Acquired interferograms are processed using the same algorithm as the ones that are implemented in the flight data processing unit.

Residual error due to the injected perturbation is identified, measured, and compared to the result predicted using numerical model simulations.

Figure 9:

Interferometer Assembly sensitivity to vibration test set-up. OGSE = Optical Ground Support Equipment.

4.2

Micro-vibration sensitivity test results

Thanks to the high level of injected perturbation, a vibration residual error is visible on ILS measurement. The residual can be unambiguously identified as it is a ghost (repetition of the ILS peak shape) which is located at a predictable distance from the ILS main peak:

… where f_perturbation is the injected perturbation frequency and opd_speed is the nominal Optical Path Difference scanning speed of the interferometer.

Figure 10 illustrates the result for the 22 Hz perturbation frequency: The residual ghost is detected at the expected location. Its amplitude is measured (1.01e-3 with respect to main ILS peak) and compared to the expected level of residual ghost amplitude assessed using numerical simulation models (0.94-3 with respect to main ILS peak).

The good match between measurement and prediction allows to validate the accuracy of the model to predict performance error residual knowing the mechanical perturbation level.

It is now possible to rely on this numerical model to compute flight performances using estimations of flight mechanical perturbation levels.

Figure 10:

Interferometer Assembly sensitivity to vibration test result for 22Hz perturbation. OGSE = Optical Ground Support Equipment.

4.3

Overview of structural an thermal model micro-vibration testing

The MTG-IRS Structural and Thermal Model (STM) is a non-functional engineering model of the MTG-IRS instrument. It is only representative in terms of mechanical and thermal behavior. It is used to test and validate thermal and mechanical behavior.

The STM test campaign includes a dedicated test to measure perturbation that reach the interferometer and that produce perturbations on the Optical Path Difference scanning. With the aim to do this test, the Interferometer Assembly STM was removed from the IRS STM. It was replaced by a functional Engineering Model of the Interferometer Assembly. The 3D metrology of the interferometer was then used as a mechanical vibration sensor.

The test provided information about the mechanical perturbation levels induced by instruments equipment such as scanning mirrors or the cryo-cooler.

The main perturbations come from the cryo-cooler. The cryo-cooler frequency and its harmonics were identified thanks to the Interferometer 3D metrology system. Those perturbation levels are lower than predicted by mechanical simulations used up to now to assess future MTG-IRS flight performances.

This test also demonstrates the ability of the metrology system to detect sub-nanometric perturbations: the smallest detected perturbations have an estimated amplitude of 0.03nm for harmonic 13 of the cryo-cooler fundamental frequency.

4.4

Synthesis of 3D metrology performances.

The results of interferometer micro-vibration sensitivity tests (§4.2) allows to validate the numerical model used for performance prediction. The micro-vibration test held on Structural and Thermal model of the instrument (§4.3) allows to confirm that the micro-vibration perturbation hypothesis used during interferometer development to assess performances are conservative.

Hence, we can conclude that the previous global performance predictions of the spectral noise due to vibrations are valid. It confirms the validity of the new 3D metrology solution used on MTG-IRS.

This result will be used for the qualification review of the interferometer assembly that is expected in 2021.