1 Introduction

MIRS is an imaging spectrometer (Barucci et al. 2021, 2024; Bernardi et al. 2024) onboard of the JAXA-led Martian Moon eXploration (MMX) mission (Kuramoto et al. 2022), in collaboration with NASA, the French space agency CNES, the German Aerospace Center (DLR), and the European Space Agency. The mission is composed of a spacecraft and a lander. The spacecraft will observe the Martian system for 3 years, making important observations that will enhance our knowledge and will land the Exploration Module on the surface of Phobos, collecting samples to bring back to Earth (more than 10 g of Phobos’ material). MMX will be the first mission to make detailed measurements of both moons and return a sample from the system to Earth for analysis. The two goals of the mission are to: i) clarify the origins of the Martian moons and constrain processes for planetary formation and material transport in the region connecting the inner and outer Solar System and ii) clarify the driving mechanism of the transition of the Mars-moon system and add new knowledge to the evolution history of Mars.

The major objective of the mission is to investigate the Mars-moons system and particularly to clarify the origin end evolution of Phobos and Deimos and constrain processes for planetary formation and material transport in the region connecting the inner and outer solar system. The second goal is to investigate the driving mechanism of the transition of the Mars-moons system and add new knowledge to the evolution history of Mars.

For the origins of the two moons, two major hypotheses are formulated and still debated: (a) Because of their spectral similarity with D-type asteroids, the moons are asteroids captured after the formation of Mars: This hypothesis shows difficulty to explain their present orbits; (b) they are fragments scattered during a giant impact with Mars after which they coalesced: This theory explains the current moon orbits (Rosenblatt and Charnoz 2012; Canup and Salmon 2018; Hyodo et al. 2018; Kuramoto et al. 2022), but not their present known composition.

MIRS has been designed to contribute to both goals of the mission. These goals will be achieved by mapping the composition of Phobos during the different MMX orbits, the composition of Deimos during multiple fly-bys, and through the investigation of Mars’ atmosphere throughout the mission. MIRS will observe between 0.9 and 3.6 µm, a spectral range where the most diagnostic signatures of minerals and molecules can be detected at the surface of Phobos and Deimos, as, for example, mafic minerals, altered phases, ices, aliphatic/aromatic CH and NH-rich compounds (Bishop et al. 2020). This spectral range will also allow MIRS to monitor both the water vapor and the dust content in the Martian atmosphere to investigate their relation.

MMX will orbit around Mars in quasi-satellite orbits (QSOs) with the relative movement of the spacecraft around Phobos in the equatorial plane at different altitudes (Nakamura et al. 2021). MIRS will acquire spectra of Phobos from high, medium and low altitudes during different phases of the mission at various resolutions and will thus help to select the two landing sites foreseen for the Exploration Module. Deimos will be observed during fly-bys at the beginning and at the end of the mission to investigate the surface composition at spatial resolutions comparable to the ones used for Phobos. Mars will be observed during orbital insertion and later phases of the mission following sampling of the surface of Phobos. In between, additional long-term monitoring of Mars’ atmosphere could also take place during Phobos eclipses.

In this paper, the MIRS instrument is presented in Sect. 2, the performed tests in Sect. 3, the measured performances in Sect. 4, as measured and known at the time of the delivery of the flight model to MELCO on April 5, 2024. In Sect. 5, a description of the in-flight calibration concept is presented, with MIRS observations in Sect. 6, ground segment for data processing and operations in Sect. 7 and the expected results in Sect. 8.

2 The MIRS instrument

2.1 Instrument overview

The MIRS instrument is based on a push-broom imaging spectrometer using a dispersion grating. A single detector acquisition (2D matrix) provides the image of a strip in one direction (spatial), and the spectrum of each point of the strip in the second direction (spectral). The second spatial dimension results from the motion of the MIRS line of sight in the along-track direction.

The final reconstruction of the image cube (Fig. 1) is performed by the Data Processing Center (DPC) on-ground (Sect. 7).

Fig. 1
Fig. 1
Full size image

A reconstructed example of a MIRS image cube, 200 × 256 spatial pixels and 256 bands in the spectral dimension spanning the range 0.9–3.6 um

In order to optimize MIRS observations efficiency and performance, a 1D scanner is implemented in the instrument to enable adjusting the line-of-sight orientation and speed in the along-track direction, i.e., the direction orthogonal to MIRS slit and following MMX ground track. This scanner will also permit to map in 2D during Phobos descent phases (Fig. 2). Cross-track observations will be achieved thanks to spacecraft maneuvers.

Fig. 2
Fig. 2
Full size image

Example of the MIRS along-track scanner projection on Phobos surface (top view), noting that MIRS is designed to operate with the slit oriented perpendicular to the track

MIRS is composed of three main components (Fig. 3): the electronics box (EBOX), the optical box (OBOX) and the harness between EBOX and OBOX (aka interconnexion) (Fig. 4).

Fig. 3
Fig. 3
Full size image

MIRS block diagram

Fig. 4
Fig. 4
Full size image

Picture of MIRS Flight Model: OBOX (on the left), EBOX (on the right) and interconnexion (white cables)

The MIRS instrument is installed on the Exploration Module of the MMX spacecraft. The EBOX is an internal component, mounted on the Exploration Module deck. The OBOX is an external component, accommodated on a dedicated bracket on the side of the Exploration Module, with a nadir pointing.

The EBOX contains two electronics boards (Fig. 5): i) the low-voltage power supply (LVPS) board that generates all power supply voltages from the unregulated primary bus voltage (33–50 V) provided by the MMX platform; ii) the Instrument Control Unit (ICU) board is the central control monitoring and processing system in MIRS, which controls all the subsystems through the FPGA (Field Programmable Gate Array). This board communicates with MMX MDP (Mission Data Processor) through two SpaceWire links (nominal and redundant). All MIRS activities are generated by commands sent by the MIRS Ground Segment, through the MDP, to MIRS where they are received and processed by the ICU.

Fig. 5
Fig. 5
Full size image

Picture of EBOX Flight Model (open configuration during integration). We can distinguish components such as the FPGA and the motor driver on the ICU board (top) and the secondary voltages circuits on the LVPS (bottom)

The ICU board is built around a FPGA device where an 8051 microcontroller is embedded. The FPGA implemented on all MIRS models is a NG-Medium from NanoXplore. To support all MIRS functions, additional electronic components are implemented on the ICU board or on the proximity electronic (PE) board.

The harness, composed of 3 harnesses wrapped up together, allows the interconnection between the EBOX and OBOX components to drive the OBOX subsystems.

The OBOX contains all the following subsystems (Fig. 6): i) a telescope to image the scene on a field stop, which is the spectrometer slit, ii) a spectrometer to disperse the light and form the image on the focal plane. After the slit, the beam is collimated to illuminate the grating that disperses the light, iii) detector assembly (cooled down to 110 K using a cryocooler to limit its dark current) and its PE to acquire and digitalize the images, iv) the along-track scanner, v) a shutter to close the spectrometer cavity and record background images, vi) a dust cover to protect the OBOX entrance window, vii) a calibration lamp to perform spectral registration and relative radiometric calibration in-flight and viii) thermal hardware to ensure the instrument remains in its flight temperature range (system based on heaters and mechanical thermostats which allow to keep the OBOX above − 40 °C, and a set of two radiators to passively cool down to limit the thermal background emission of the spectrometer).

Fig. 6
Fig. 6
Full size image

Picture of OBOX Flight Model (open configuration during integration). The light entrance is located at the dust cover location (see also Fig. 11 for the optical path)

2.2 Optical design

The MIRS optical design consists in a catoptric afocal system, followed by a diffraction grating and two lens assemblies (Fig. 7). To allow for the highest possible transmission of the wavelengths from 900 to 3600 nm, the sapphire entrance window is coated on both sides in order to prevent the transmission of the wavelengths from 400 to 800 nm (not useful for MIRS) that could generate some straylight. The scanning mirror is a flat aluminum mirror coated with a protected silver coating. The afocal system consists of an off-axis telescope with two free-form mirrors, diffraction-limited in the focal plane where the spectrometer’s entrance slit is placed, and an off-axis collimator with two free-form mirrors, diffraction-limited in the plane of the system’s exit pupil, where the grating is located. The telescope’s image focal length is 93.5 mm. The collimator’s object focal length is 51.1 mm. Together, they form an afocal system with an angular magnification equal to 1.83. The entrance pupil diameter is 23.4 mm.

Fig. 7
Fig. 7
Full size image

MIRS optical layout

The diffraction grating is an off-the-shelf component, manufactured by Richardson. It is a plane ruled grating, with 53 grooves/mm, used in the first diffraction order. The incidence angle is 17°.

The spectrometer contains two dioptric objectives. The first objective collects the collimated and diffracted beam after the grating and forms an image of the slit in an intermediate focal plane where the order sorting filter is placed. The second lens acts as an imaging relay between the intermediate focal plane and the final focal plane, and as a pupil imaging lens, imaging the pupil at the detector’s cold stop location.

The order sorting filter is a dielectric multilayer filter, whose cutoff wavelength is variable as a function of position on the filter.

2.3 The OBOX subsystems

2.3.1 The detector

The detector is based on a MCT (mercury cadmium telluride) detection chip and is optimized in the range 0.9–3.6 µm (operating at 110 K). It is hybridized on a CMOS readout circuit. This detector, called “NEPTUNE,” is supplied by Lynred. The detector is encapsulated in a cryostat equipped with a K508 cryocooler from RICOR, which permits to maintain the temperature below 110 K. This cryostat also manages the electrical interfaces with the front-end electronics.

The MIRS FPGA/AFSW implements an autoexposure algorithm to determine the optimal integration time maximizing the dynamic range of the acquired signal while minimizing the number of saturated pixels. The FSW algorithm starts with an initial integration time seed and uses successive raw images to approximate the integration time based on individual image statistics computed by the FW (firmware). A stacking function allows us to sum several raw images onboard MIRS to create one MIRS image thus improving the SNR. The detector characteristics are listed in Table 1.

Table 1 Detector assembly characteristics
Full size table

2.3.2 The dust cover mechanism

The dust cover mechanism has the role of protecting the OBOX entrance window during the four descent phases on Phobos (two rehearsals and two landings for sampling). The dust cover shall also remain in the closed position during the launch phase and when there is a risk of solar pointing. The motorization is made by a Phytron VSS-19 gear motor. It is linked to a cover part with help of Myonic bearings, hinges and a system of two rods. This system provides a large angular opening (103°) in 24 s and allows also for a stable closed position without power during mechanical environments.

The dust cover is equipped with a limit switch that triggers when the dust cover is open (Fig. 8). In case of an anomaly detected on the dust cover actuations, it is possible to activate an emergency function based on a Hold Down Release Mechanism to unstack part of the mechanism and allow MIRS to continue science observations.

Fig. 8
Fig. 8
Full size image

OBOX front view with the dust cover mechanism in the open position revealing the light entrance through the Sapphire Window

2.3.3 Scanner mechanism

A single-axis scanner allows adjusting the line-of-sight orientation and speed in the along-track direction (the direction orthogonal to MIRS slit). The scanner mechanism (Fig. 9) is a sensor-less single-axis scanning device moved by a geared stepper motor. The geared motor moves a circular cam to push a lever that imposes the flat mirror rotation.

Fig. 9
Fig. 9
Full size image

Picture of FM scanner mechanism. The square part is the flat mirror of the mechanism. The vertical cylinder behind the mirror is the stepper gear motor

Precision of the mirror motion is given by the stepper controller. The knowledge of the mirror position is better than ± 0.35 mrad (± 0.02°). Repeatability for mirror positioning is about 0.002° (0.035 mrad). The entire stroke of the mirror is about 28°. This stroke is ended by two mechanical end stops. Between these two end stops, the mirror can move by ± 10° for science scanning, and it can reach a calibration position on one side of the stroke near one end stop (+ 17°).

2.3.4 Shutter mechanism

The shutter mechanism function is to mask the spectrometer entrance slit (0.065 × 5.5 mm2) for thermal background subtraction with a stabilized position reached in less than 15 ms (close position). The designed and qualified concept can be considered as a fully integrated mono-stable, single phase, permanent magnet actuator. The shutter mechanism (Fig. 10) provides a mono-stable position (open position), when not energized or in case of any failure mode.

Fig. 10
Fig. 10
Full size image

Picture of FM shutter mechanism. The moving part is the small black pellet

2.3.5 Internal calibration lamp

The internal calibration system is used to check the spectral registration of the spectrometer (wavelength versus pixels), thanks to the spectral features of the filters. The internal calibration system (Fig. 11) is made of: i) a source assembly (graphene source, collimating lens, window, 2 filters, 1 folding mirror), ii) a folding mirror bonded beside the OBOX entrance window.

Fig. 11
Fig. 11
Full size image

Schematic layout of the internal calibration source assembly (left) and the calibration optical path (right)

The light emitted by the calibration source assembly is reflected by two flat mirrors (folding mirror fixed on the OBOX entrance wall and scanner mirror) before entering the spectrometer cavity. A typical internal calibration spectrum is reported in Fig. 12.

Fig. 12
Fig. 12
Full size image

Typical internal calibration spectrum recorded at FM thermal vacuum test of − 32 °C. Sixteen absorption spectral features of the rare earth and polystyrene (dashed lines) will be used for spectral registration

2.3.6 Technical resources

The technical resources of MIRS instrument, which are all compliant to MMX allocations, are reported in Table 2.

Table 2 MIRS technical resources
Full size table

3 Instrument tests

The MIRS qualification program is based on a proto-flight strategy due to the very short development duration. Before the delivery to JAXA for integration on MMX spacecraft, the MIRS team conducted several qualification tests and calibration activities on the FM instrument using the facilities available at LESIA and other facilities in France. The tests were divided into four main categories: i) mechanical test, ii) thermal test, iii) system tests and iv) performance and radiometric calibration. The electromagnetic tests and qualification were performed on an engineering model of MIRS, representative of the FM in terms of electrical design and interfaces.

3.1 Mechanical tests

The mechanical environment requirements (Table 3) have been one of the potential issues during the MIRS development as the OBOX embeds some off-the-shelf equipment such as the detector assembly. Some mechanical tests (shocks and random vibrations) were performed on a structural model of OBOX with dummy masses and then on the engineering models of the subsystems to assess the random vibrations levels at the subsystems’ interfaces.

Table 3 Mechanical environments requirements of MIRS
Full size table

Finally, the MIRS FM mechanical qualification campaign took place during June 2023 for both OBOX and EBOX. The two components were tested separately and tested on each axis. The MIRS FM successfully passed the mechanical qualification tests.

3.2 TVAC campaign: thermal test and calibration

During the three-week thermal vacuum campaign (TVAC) carried out in October 2023, MIRS underwent several tests at different temperatures to check the thermal response and performance in various possible scenarios during the mission. The goal of the TVAC campaign was to demonstrate that the MIRS FM was ready to operate in the flight environment and achieve the specified performance providing at the same time the inputs for the operational constraints. The TVAC was composed of 2 thermal cycles under vacuum alternating hot and cold non-operational qualification in [− 55 °C; + 60 °C] at both the EBOX and OBOX interfaces. During the cold cycle, two main sessions of testing were carried out:

  • The first cold cycle was dedicated to functional tests and verification of the requirements, in particular spectral range, spectral resolution and sampling as well as information on the field of view, signal-to-noise ratio (SNR) and the modulation transfer function.

  • The second cold cycle was focused on radiometric calibration and performance evaluation with the main objective of acquiring knowledge of absolute radiometric calibration better than 10% and relative radiometric calibration better than 1%.

All the tests were carried out in clean rooms of ISO 8 level of the MESPAL (Moyens d'essais, Salles Propres, AIT/AIV du LESIA) at LESIA-Paris Observatory with controlled environmental parameters: temperature 20 ± 2 °C, humidity 55 ± 5% and over pressure > 25 ± 2 Pa. The MIRS FM was placed inside the SimEnOn vacuum chamber at LESIA-Paris Observatory. In the first cycle of the TVAC campaign, SimEnOn chamber was coupled with an optical bench called “Banc 2 M” to measure alignment of internal optics and performances using an additional optical cell filled with pure methane. Results on the performance test are reported in Sect. 4. During the second cycle of TVAC, the Banc 2M was replaced with the CARA optical bench visible in Fig. 13 for radiometric calibration. The CARA bench consisted of a calibrated black body source illuminating a parabolic gold mirror reflecting the collimated light from source to the MIRS entrance window passing through the SimEnOn sapphire window. The entire setup was flushed with pure nitrogen during the calibration test in order to limit the H2O and CO2 atmospheric contribution to the measurements. Six temperatures of the black body were investigated with the MIRS FM OBOX at three different temperatures acquiring more than 8000 images.

Fig. 13
Fig. 13
Full size image

View of the CARA optical bench interfaced with SimEnOn chamber during MIRS radiometric calibration campaign

At the conclusion of TVAC campaign, all Thermal Qualification wells passed successful with Survival Duty cycles consistent and correct. MIRS was consequently declared ready for space thermal environment.

3.3 System tests

In addition to the MIRS instrument validation performed by the MIRS LESIA team, the MIRS CNES team defined a series of tests aiming at the MIRS system validation. The system validation encompasses the following points:

  • Components tests: qualification of the different functional components of the system, individually (MIRS instrument, MIRS CGS and its sub-components for MIRS programming and flight dynamics, MIRS DPC, etc.) [see Sect. 7]. This qualification is done by the entity in charge of the component development who checks the component features, robustness, performance, etc.

  • Interface tests: validation of all the interfaces between the system components

  • Qualification tests:

    • Operability tests involving MIRS flight software and MIRS CGS for validation of the monitoring and control of the instrument in operational conditions (typical nominal cases, constraining nominal cases and contingency cases)

    • Operational routine tests involving all system components

  • Final operational tests for flight procedure validations (taking into account all operational constraints) and operation team training.

System tests are mainly performed on simulation means at CNES (instrument simulator but also image simulator). Some system tests are performed on the engineering model at LESIA (including a test at cold temperatures during TVAC campaign) to check test procedures against MIRS HW. System tests are led jointly with the development of the flight software at LAB as well as the development of the ground segment at CNES, LESIA and JAXA. A coordinated validation workflow allows us to test and validate step by step the interfaces and functions developed in the different components of the system.

The system qualification tests are still ongoing, aiming at MIRS operational qualification to be started next year.

In addition to the MIRS instrument validation and the MIRS system validation, the MMX system validation ensures the operability of the MIRS instrument integrated on MMX through the JAXA ground segment. The MMX system validation is performed by JAXA.

4 Performances

The MIRS verification and validation process was based on early performance measurements during the OBOX integration at cold temperature (− 27 °C) and atmospheric pressure. Final performances (Table 4) of the instrument were characterized during the 3 weeks of thermal vacuum testing. A dedicated optical test bench was developed and installed in front of the vacuum chamber, allowing to scan the spectral range and the full field of view.

Table 4 MIRS instrument performance results
Full size table

Few points have a status partially compliant but with no or minor impact on the scientific analysis. The instrument covers the full spectral range from 0.9 to 3.6 µm. with spectral sampling better than 10 (± 10%) nm over 2.8 µm. The spectral resolution of MIRS has been estimated using both an external CH4 gas cell and the internal calibration lamp of the instrument (Fig. 14). The requirement, that is a spectral resolution of 21 ± 2 nm, is met for wavelengths < 3.2 µm. Beyond 3.2 µm, the acquired spectra have very low signal and do not permit to have an accurate estimation. The pessimistic estimation, derived by fitting the observed methane features with the HITRAN-HAPI methane model (Gordon et al. 2022) convolved with Gaussian kernels of different widths, gives a spectral resolution < 32 nm in the 3.2–3.6 µm range. This estimation is very likely affected by the limitation in the instrumental setup placed in front of MIRS. Moreover, using the internal graphene lamp did not help in estimating the spectral resolution at longer wavelengths, because of the lack of features and of the low SNR.

Fig. 14
Fig. 14
Full size image

Observed methane cells line (blue line) with superposed the best fit models produced using the HAPI methane lines convolved with Gaussian kernels having FWHM of 21 nm up to about 3 µm and 31 nm beyond 3 µm. The vertical lines show the bands used to compute the spectral resolution

Requirements are satisfied up to 3.2 µm using only CH4 cells measurements, beyond data are too noisy or have low SNR. Even in case of the worse estimation in spectral resolution, MIRS will keep the ability to observe and identify the main compounds that may be present at the surface of Phobos and Deimos, as reported in Sect. 8. Spectral resolution analysis gave similar results at the three different temperatures for the OBOX showing good stability of the instrument to changes in the environmental conditions. Better calibration of the range 3.2–3.6 µm will be achieved with the planned in-flight calibration during the cruise phase of MMX.

The modulation transfer function (MTF) was assessed by imaging a monochromatic source point in different conditions and fitting an instrument model with optical aberrations to be computed. However, the aberrations are not fully understood, and we suspect that part of them come from the instrument as well as from the test bench. In order to conclude on the acceptability of the PFM model, we performed two estimations of the MTF, including movement of the satellite:

  • The optimistic case considers that all observed aberrations come from the test bench,

  • The pessimistic case considers that all observed aberrations come from the instrument.

The resulting images are shown in Fig. 15, and the degradation considering all the aberrations come from the instrument is barely noticeable, due to the smooth spatial features of the observed scene.

Fig. 15
Fig. 15
Full size image

Typical image simulation with optimistic (left) and pessimistic (right) modular transfer function (MTF). The MMX-Phobos distance is about 50 km, and the projected pixel scale is about 14 m

SNR predictions can be performed thanks to a model developed at the beginning of the development for the instrument sizing. This model was improved and completed taking into account the instrument measured characteristics (optical transmission, detector dark current, quantum efficiency and readout noise, cold stop diameter). Validation was made by comparing the predictions to measured data during the thermal vacuum tests. It helped in particular to check the correct evaluation of the thermal background. Finally, this model uses as inputs the scene parameters (sun illumination, albedo), and the observations constraints (maximum allowed integration time). In Fig. 16, SNR predictions are shown for typical spectrometer temperatures (246–253 K). The requirement of reaching a SNR higher than 100 up to 3.2 µm in less than 2 s for 253 K max is met for phase angles lower than 30°. Better SNR is achieved at colder instrument temperatures and lower phase angles.

Fig. 16
Fig. 16
Full size image

Temperature sensitivity in achieving SNR levels in Phobos spectral data

The relative and absolute radiometric calibration accuracy was measured using the CARA bench and varying both the instrument OBOX temperatures (33 °C, − 20 °C and − 10 °C), the blackbody temperatures (300 °C, 400 °C, 500 °C, 600 °C, 800 °C and 1000 °C), and the integration time (ranging from 0.2 to a maximum of 25 ms, with longer integration time depending on the selected blackbody temperature to avoid strong saturation of the detector). We checked first the blackbody flux stability at different blackbody temperatures and exposure times, discarding data where the blackbody flux was not stable enough, or saturated. The pixel-by-pixel relative response was derived combining the data acquired at the 3 different OBOX temperatures, and at the 3 blackbody temperatures (400, 500 and 600 °C), guarantying the wider spectral coverage without saturation and the higher flux stability. The derived relative radiometric accuracy is < 1% beyond 1.5 µm, therefore compliant with the requirements. At shorter wavelengths, the blackbody flux at all tested temperatures was too low to permit any reliable radiometric measurement. The absolute radiometric calibration was estimated using the same setup and the different OBOX and blackbody temperatures. The theoretical blackbody flux entering in the instrument was compared to the measured one to derive the instrumental response of MIRS. In estimating the absolute radiometric calibration, several sources of errors should be taken into account: a) the blackbody flux stability and uniformity, estimated at 2% beyond 1.5 µm; b) the transmission accuracy of the SimEnOn chamber window and of the parabolic mirror directing the light from the blackbody to MIRS in the CARA optical bench; the accuracy in these optical elements is estimated to be about 2%; c) the MIRS detector gain map accuracy, which is the largest error contribution (7–8%); d) the contribution of the atmospheric transmission; this is very localized in the spectra in particular window and may be corrected by model; the associated error is 0.5%. Adding quadratically the individual errors sources, the absolute radiometric calibration is better than 10%, thus compliant with the specification, beyond 1.5 µm. We stress again that the ground-based MIRS calibrations are not reliable in the 0.9–1.5 µm range because of the extremely low SNR of the collected signal. For both relative and absolute radiometric calibration, in-flight observations of standard spectrophotometric stars are foreseen during the cruise phase of MMX toward Mars and will permit the radiometric characterization of MIRS in the whole spectral range, and repeated regularly during the mission, as detailed in the following section.

5 In-flight calibration concept

In addition to ground-based calibration, in-flight calibration will be conducted during the whole mission in order to complete the previous calibration done on ground, as well as to check and correct from instrument temporal variations, that could be initially due to excessive mechanical stress during the launch or to temporal hazards affecting the instrument during the cruise and main observation phases. For this purpose, several observations will be performed periodically, and each type of observation will be devoted to checking and improving the accuracy of the spectral calibration of MIRS.

For this purpose, there will be four different calibration procedures that will be used at different frequencies. These procedures will consist in:

  1. 1.

    Checking and improving the absolute radiometric curve from observations of bright stars

From ground-based calibration, we checked that the accuracy of the absolute radiometric solution was mainly better than 10%, except for a small part of the detector, which was not enough illuminated to reach high enough SNR level (typically before 1.5 µm, as initially expected). In addition to that, the global absolute performance was mainly limited to the summation of various errors, dependent to the setup used for this purpose, and could be partly improved in-flight. For these reasons, we will observe several bright infrared sources (the two brightest stars will be Betelgeuse, aka alpha Orionis, and Arcturus, aka alpha Bootes) during the cruise and main phases. The acquisition of these stars will provide us sufficient flux, especially to fill the caveat at short wavelength and improve the current absolute radiometric solution at an expected few percent level. Each star will be mainly pointed on the central part of the slit, along the spatial direction, and the slit will then be shifting along the cross-track direction with enough margins to be sure to point each star (see Fig. 17, where 11 consecutive frames have been simulated in the case of Betelgeuse). Repetition of these acquisitions will be done to reach sufficient SNR at any wavelength and be compared to the absolute tabulated flux of these calibration sources. For Betelgeuse, as flux variations are expected during the mission, additional ground-based observations of this target will be done, closely in time.

Fig. 17
Fig. 17
Full size image

Left: Reconstructed data cube of Betelgeuse from 11 consecutive MIRS frames as seen at 1.5 and 3.3microns. Right: Extracted spectra of Betelgeuse from each frame (dashed lines) and from the whole frames (continuous line)

  1. 2.

    Checking the pixel-to-pixel response from bright stars and resolved homogeneous areas of several targets:

Observations of the same bright stars will also be performed using different positions along the slit in order to compare the solution at different locations on the detector and check for a few columns the validity of the pixel-to-pixel response map edited from the ground. For this purpose, we also plan to get the spectral response of homogeneous and similar areas of resolved objects (i.e., parts of Phobos and Mars surface are currently compliant for this task). The idea of this procedure is to point and swipe the same chord of the target using the motion of the spacecraft, along the slit direction, allowing each part of the slit, and then each column of the detector, to encompass exactly the same area of the target and therefore the same amount of light.

  1. 3.

    Checking the stability of the registration

As performed during the TVAC, we will acquire spectra of the graphene lamp to monitor the registration provided by the internal lamp solution. Since the internal solution was stable during the ground-based calibration (within the 5 nm specification), whatever the OBOX temperature (ranging from − 15 to − 36 °C) and scanner position (from edge to edge), this procedure seems to be the best tool for this purpose.

  1. 4.

    Constraining the spectral effects caused by the observational geometry using particularly smooth and regular areas on resolved targets

The goal of the three previous procedures is to calibrate the detector from a radiometric and spectral registration point of view in order to be able to fully link each ADU measured by a pixel of MIRS detector to the associated physical value (radiance) coming from the observed scene. This last calibration attempts to take into account, or at least constrain, the effects of the observational geometry on the amount of light received by MIRS within the spectral wavelength range and be able to disentangle between geometric effects from real physical variations of the surface. This procedure will mainly concern the treatment of Phobos data, from which compliant analyses are possible. To fulfill this task, smooth and homogeneous areas will be observed using several different geometries. The areas to be observed need to be smooth enough to avoid shadows formation and hiding effects, both modifying the scene characteristics (i.e., the physical information).

6 MIRS observations

To achieve its scientific goals, MMX will collect samples from Phobos to return to the Earth, will observe Phobos and Deimos and monitor Mars’ atmosphere. MMX trajectory during the five years of its mission is designed accordingly and divided into five main phases (Nakamura et al. 2021).

6.1 Cruise

A cruise of around one year will be necessary to reach the Martian’ system, during which the instruments will be commissioned and calibrated radiometrically and geometrically thanks to stars observations. As MIRS is a push-broom spectrometer, during the calibration its internal scanner will move its line of sight in the cross-slit direction. Joint observations with the wide field-of-view optical camera OROCHI and the narrow high-resolution optical cameras TENGOO and CAM-T will allow a fine determination of their relative alignments (Fig. 18).

Fig. 18
Fig. 18
Full size image

Geometric calibration of MIRS with TENGOO on Betelgeuse (left). Comparative Fields of View of MIRS, TENGOO, CAM-T and OROCHI. MIRS has a field of view of 0.02° × 3.3°, TENGOO and CAM-T have a field of view of 1.1° × 0.82°, and OROCHI has a field of view of 73° × 58° (right)

Such calibrations will also be performed regularly during the whole mission to adjust the instruments settings when necessary.

6.2 Insertion into Phobos co-orbit

The time necessary to insert MMX into its first Phobos co-orbit will be very productive with the first Phobos distant observations and several Deimos fly-bys. This strategy allows to compare the composition of both Mars’ moons while conserving MMX’s fuel resources.

6.3 Phobos detailed observation and landing sites selection.

To prepare MMX touchdowns on Phobos, a very good understanding of its composition and topography will be achieved during a one-year co-orbit phase, with several stages of decreasing altitude, from 100 to 7 km. In particular, MIRS hyperspectral capabilities will allow for characterizing the surface composition, and the potential presence of key features regarding the origin of Phobos such as phyllosilicates or organic compounds, which will help orient the landing and sampling site selection.

During the first high-altitude co-orbit MIRS will perform a global map with about 100 m resolution (Fig. 19) with a limited number of observations (18) and a total observation time lower than 4 h thanks to the use of its internal scanner. This efficiency is particularly important in that phase for which most of the available Phobos observation time will be dedicated to its shape characterization with OROCHI and TENGOO. The optimization of the number of telecommands to upload as well as the coverage margins between observations is particularly important as MMX, like all interplanetary missions, has to manage a limited Earth.

Fig. 19
Fig. 19
Full size image

MIRS Phobos Global Coverage resolution in meters in high-altitude co-orbit (left). On the right MIRS performing a scanner observation in high-altitude co-orbit

communication bandwidth.

This map will contribute to the selection of landing candidate regions that will be more closely observed thanks to a medium altitude co-orbit during which another global composition map of medium latitudes at a better resolution will be performed. This improvement of resolution entails an increase in the number of observations necessary to cover the target area which leads to a complex scheduling problem that we solve by combining a Tetris inspired heuristic and a genetic algorithm. Despite the difficulty to observe the toward-Mars region induced by the Sun eclipse by Mars season (Fig. 20a), a global coverage of the medium latitudes with a low phase angle can be reached (Fig. 20b).

Fig. 20
Fig. 20
Full size image

Illustration of the difficulty to observe the Phobos in the Mars-facing direction, during eclipse in Mars’ shadow (left). MIRS Phobos Global Coverage Phase Angle in medium altitude orbit (right)

The best landing candidate regions will be selected for finer observations during lower altitude co-orbits. In addition, MIRS will perform Phobos thermal inertia observations at specific sets of local time and phase angle.

6.4 Phobos sample collection (phase 3)

The collection of samples from Phobos on its toward and anti-Mars face will be preceded by rehearsals, during the first of which the CNES-DLR Rover Idefix will be released. During each descent and ascent, the landing sites will be finely observed down to about 400 m above Phobos’s surface to provide detailed composition maps. Below 400 m, the MIRS shutter will be closed to avoid damage and contamination to the instrument by dust lifted off Phobos’ surface.

6.5 Phobos complementary observation and Mars atmosphere monitoring

An additional co-orbit phase of several months will give the opportunity to complete Phobos coverage, potentially on the high latitudes. During its whole mission MMX trajectory with respect to Mars will be globally equatorial. This position will be beneficial to monitor the low-to-mid-latitude Martian atmosphere. In particular, this will allow observing the same area of Mars with different local times. During orbital insertion, Mars observation could be carried during each of its available observation slot of ~ 3h50’. This observation period corresponds to late northern summer on Mars and the beginning of the dust season (Wang et al. 2023). This should be the occasion for first time activities with at least one observation with each possible Mars atmosphere observation mode: 1) global mapping, allowing for an extended coverage of the Martian low-to-mid-latitudes within a few observation slot; 2) limb scan; 3) region of interest (ROI) tracking mode with both high spatial (~ 2.5 km) and temporal resolution (< 45 min) of a designated area of Mars together with high-resolution imagining by TENGOO; and 4) continuous nominal monitoring mode to maximize the observation of the same area on Mars with different local times (Fig. 21) The main challenge for observations of Mars’ atmosphere will be the handling of data budget given the larger surface to cover than Phobos and the transient nature of the observations.

Fig. 21
Fig. 21
Full size image

Illustration of local Mars time across the surface when measured by MIRS in the Mars Nominal Mode

6.6 Deimos observation and complementary Mars atmosphere monitoring (Phase 5)

The final MMX observation phase will last several months and will be dedicated to Mars and Deimos. The 4 Mars observations modes will be used but adapted in order to cope with the challenge of the MMX elliptic orbit with respect to Mars. The challenge for Deimos is to get the best possible coverage through observations during each of the high velocity fly-bys. The use of the MIRS internal scanner will enlarge the coverage of each observation (Fig. 22).

Fig. 22
Fig. 22
Full size image

Example MIRS coverage and resolution in meters on Deimos obtained from a fly-by observation (left); MIRS, and the slit during a Deimos fly-by observation (right)

6.7 The challenge of MIRS observational strategies

From the scientific objectives that define what to observe and why, the observational strategies determine when to observe and how. This iterative process follows all the phases of the project, from the initial design to the operations and requires a close cooperation with the scientific teams and all the components of the space system: Spacecraft, Trajectories, Instruments, Operations and Ground Segment. Along with the scientific objectives, expected performances leading to the final scientific products are progressively refined, allowing for an incremental definition of how and when to observe. Also, the assessment of the necessary budgets for data volume, total observation duration and number of observations is initialized very early in the project and continuously refined in order to ensure their compatibility with the System and the instrument design and constraints (Table 5).

Table 5 MIRS data volume, observations duration, number of observations and scanner stroke
Full size table

7 Ground segment for data processing and operations

The MIRS Ground Segment is a distributed segment using Japanese and French resources. The outline of the MIRS system is shown in Fig. 23.

Fig. 23
Fig. 23
Full size image

Flowchart outline of the MIRS System and CNES-LESIA/LIRA Ground Segment

MIRS Ground Segment communicates with the Instrument (for commands and telemetry) through the JAXA control Center on the ground and through SMU (Spacecraft Management Unit)/MDP (Mission Data Processing) subsystems on MMX Spacecraft.

MIRS Ground Segment is then composed of two main entities:

  • The MIRS Control Ground Segment (CGS) at CNES (Toulouse), which is the only interface with JAXA Ground Segment and is in charge of the following functions: a) instrument monitoring and control, b) definition of instrument programming and commands generation, c) flight dynamics (FD) assessment for MIRS, d) interfacing with the science team (on raw science data products, processed products, and science requests) and e) operational functions (ground segment process scheduling, data management, MIRS memory management, etc.). MIRS in-flight Operations will be performed at CGS.

  • The MIRS Data Processing Center (DPC) at LESIA (Paris Observatory), which is responsible for the following functions: a) elaboration of MIRS Science Requests; b) decoding of the Raw Science Data Telemetry (TM); c) generation of the MIRS Data Products until level 2; d) validation of MIRS Products performance; e) delivery of MIRS Data Products to Co-Is and MIRS instrument team; f) delivery of MIRS Data Products to MIRS CGS for Archiving and Distribution to JAXA and MMX Science Team; g) MIRS Instrument Database and Products Database delivery; and h) development of tools to support ground testing.

Regarding CGS, tools are developed to perform the following detailed activities (Fig. 24):

  • Downlink: all the processes dedicated to data processing (spacecraft data, mission data and health status), instrument monitoring, data archiving and bandwidth resources management necessary for downloading the data.

    • MIRS data (HKTM and science TM) are fetched on a JAXA server

    • CGS checks the instrument health as well as the correct execution of the executed command plans

    • Science Raw data, MMX Attitude and Orbit files as well as other additional data are transferred to DPC for processing

    • MMX Attitude and Orbit files are sent to and processed by FDS

    • Processed products (L1, L2) will be sent to CGS after production and assessment by DPC. Those products will be eventually sent to JAXA for PDS archiving.

  • Uplink: all the processes dedicated to uplink activities from science planning up to command plan upload to the spacecraft.

    • Based on MMX activity plan, MMX attitude and orbit files, FDS inputs, MIRS team’s science requests and all known mission constraints, CGS defines the best MIRS instrument programming for the next programming horizon.

    • MIRS and MMX related Command plans are generated accordingly and checked against flight rules.

    • Once cleared, the command plans are formatted and sent to JAXA for uploading

Fig. 24
Fig. 24
Full size image

Architecture overview of the MIRS Control Ground Segment (CGS)

Regarding DPC, a pipeline is developed to decode raw TM and to generate MIRS data products.

This pipeline will be hosted on LESIA servers. It is based on 4 software components:

  • Decoding of science TM.

  • Geometry computation. This will be done using NAIF SPICE software and MMX SPICE kernels information: reconstructed attitude and orbit (provided by JAXA). TENGOO/OROCHI images and in-flight stars observations will also be used to readjust pointing information.

  • Radiometric calibration. Several corrections such as bad pixels, odd/even effect and distortion are done using on-ground calibration measurements. The spectra are also divided by the instrument transfer function obtained with on-ground calibration. This function is dependent on multiple parameters.

  • Spectral calibration. On-ground calibration measurements are also used to obtain a correlation between wavelengths and pixel positions. In flight, internal calibration with spectral lamp will be used to correct spectral calibration if needed.

The pipeline also performs cubing of the data and PDS4 formatting (at first, FITS format will be delivered for MIRS scientific team).

In Fig. 25, the different levels of the data are illustrated.

Fig. 25
Fig. 25
Full size image

Overview of MIRS Level 0, Level 1 and Level 2 data products

L1A products are uncalibrated 2D spectral images (Fig. 26: (left) example of L1A image, (right) schematic of L1A product). They are decoded, reconstructed and uncalibrated in instrument native format (Digital Number—readable format). One elementary image = 256 pixels × 200 pixels.

Fig. 26
Fig. 26
Full size image

MIRS L1A data example (uncalibrated 2D spectral images)

L1ACAL products are spectral calibration obtained onboard with the internal lamp.

L1ADK products are dark data for radiometric calibration purpose. Dark subtraction is done onboard. But one dark is downloaded before and after each observation to check the detector’s health.

L1C products are uncalibrated 3D cube data (3D images reassembled using the scanner or push-broom track). Typical size is 200 × N × 256 pixels with N = number of lines during the scanning. The schematic describing the 3 dimensions of the cube is reported in Fig. 1.

L1BGEO are data cube associated to each 2D spectral cube images that provides information related to geometry (needed for 3D projections at Phobos/Deimos/Mars surfaces). Those data will be predicted geometry just after acquisition and later reconstructed geometry based on refined attitude Spacecraft.

L2B are calibrated 3D cube data in radiance (W/m2/µm/sr). They are composed of 3D images reassembled using the scanner or push-broom track. They are obtained from L1C data that are calibrated using housekeeping information and internal calibration sources.

L2C are reflectance calibrated cube data obtained by division of L2B data by an appropriate solar spectrum.

The MIRS data products will be available for MIRS scientific team within 1 day of the science telemetry availability on JAXA server. This delay is reduced to 6 h for critical phases such as LSS (Landing Site Selection).

8 Expected results

The near infrared imaging spectrometer MIRS has been built to optimize the observations of Martian Moons from the MMX orbits, as well the Martian atmosphere to achieve the MMX objectives.

MIRS will observe the global mapping of Phobos at different altitudes and resolutions, with a spatial resolution of about 20 m during the equatorial QSO-M orbit, for observations in the range of about ± 45° latitude, and the polar regions of Phobos for above/below 30° latitude, during the QSO-3DM orbit.

MIRS will be able to characterize the composition of Phobos in both the red and blue units and investigate the local compositional heterogeneity at the different geomorphological features across Phobos’ surface. MIRS will characterize the surface composition of Deimos and allow us to compare it with that of Phobos at similar surface resolution.

MIRS will measure the surface temperatures of selected areas and from these measurements compute the surface thermal inertia of Phobos. At wavelengths beyond 2.5 μm, the thermal tail due to Phobos’ thermal radiation becomes more important than the reflected sunlight. The thermal tail is in function of the temperature and is also a function of the infrared emissivity. The thermal inertia will give important constraints on the grain size of the regolith and the material porosity.

MIRS will spectroscopically observe the surface of the Martian moons from 0.9 to 3.6 µm obtaining a continuous spectrum with no gaps as was the case of previous observations by instruments like OMEGA (Observatoire pour la Minéralogie, l’Eau, les Glaces et l’Activité) on board of Mars Express and CRISM (Compact Reconnaissance Imaging Spectrometer for Mars) onboard the MRO (Mars Reconnaissance Orbiter) (Fraeman et al. 2012, 2014). MIRS will be able to detect the presence of all the molecular bands with a spectral resolution of 20–30 nm and to constrain the mineralogy, species’ abundances and composition of Phobos and Deimos. In particular, MIRS will be able to detect the presence of anhydrous silicates (features in the 0.9–1.0 and 2.0 μm regions), minor overtones (at 1.4 and 1.8 μm), hydrous silicate minerals (at 2.7–2.8 μm), water (ice) (absorption bands at 1.5, 2.0 and 3.0–3.2 μm) and organic matter (3.3–3.5 μm). Figure 27 shows the spectra of Bennu and two meteorites convolved at the spectral resolution of MIRS (of 20 nm up to 3.2 μm and up to the worst estimation of 30–35 nm for the spectrum beyond 3.2 μm, see section on performance) to demonstrate the capability of MIRS to detect organics. The unique performances of MIRS will allow us to detect the presence of all bands even with band depths as weak as 3%. A detailed investigation of the band center, depth and area of the detected absorption bands will allow for a detailed analysis of the surface composition of Phobos and Deimos. In particular, if MIRS detects patches with large quantities of organic material, this should provide sufficient SNR in the 3.2–3.6 µm region to investigate their composition, in particular regarding their aliphatic or aromatic nature, whether they are nitrogen-bearing or pure carbonaceous. If especially organic rich terrains are observed, an evaluation of the organic’s degree of unsaturation may even be achievable through a deconvolution of their –CH2 and –CH3 symmetric and asymmetric stretching modes. In addition, MIRS, with an unprecedented spatial resolution, could be able to detect exogenous material, if present, on the surface of Phobos, either from meteorite and micrometeorites or from ejecta. This will provide access to transport mechanisms between the planet and the moons.

Fig. 27
Fig. 27
Full size image

Comparison of selected spectra of Bennu asteroid (obtained by OSIRIS-REx mission), Tagish Lake and Orgueil meteorites (obtained in laboratory) convolved at different spectral resolutions on the range 20–35 nm as expected from MIRS. The worst case in terms of spectral resolution (35 nm) shown on top in the three panels displays the expected noise evaluated using the prediction of Fig. 16. Even with stronger noise in the 3 micron region, faint organic bands can still be detected. Spectra in each panel are shifted for clarity

MIRS will be a critical support for the selection of landing and sampling sites. MIRS will observe the surface composition of Deimos and compare it with the composition of Phobos at a similar spatial resolution. The obtained spectral data will allow to search for heterogeneities that could be linked to the Martian surface composition. Potential dichotomy between the leading and trailing hemisphere would help investigate the effect of micro impact induced space weathering, while if present, a dichotomy between the sub-Martian and anti-Martian hemispheres would likely be related to the effect of oxygen ions escaping from the Martian upper atmosphere (Nenon et al. 2019).

Regarding Mars, the exceptionally unique orbit of MMX in the equatorial plane of the planet, will be optimal for low-to-mid-latitude observations. Compared to previous near-IR spectrometers such as OMEGA (Bibring et al. 2004) or CRISM (Murchie et al. 2007) aboard missions operating in polar orbits, MIRS will be able to monitor continuously an area of Mars for ~ 3.5 h during the daytime and return to it ~ 3.5 h later after the night-time, allowing for an unprecedented investigation of local-time effects. This capacity of MMX to observe at different local times will be a key to understanding the diurnal variations of water vapor, water ice clouds formation (in particular in the equatorial cloud belt), dust cycle and dust storms, surface pressure and CO maps and the formation of water/CO2 ice on surface areas. Compared to the EMIRS Emirates Mars Infrared Spectrometer) instrument aboard Emirates Mars Mission (EMM), covering the spectral range from 6.0 to 100 µm (Edwards et al. 2021), the only other mission in the equatorial plane, MMX offers a much closer to Mars MMX orbit (~ 6000 km from the Martian surface) and hence affords and increased spatial resolution by up to a factor 50. In particular, MIRS will be able to monitor atmospheric density, pressure and temperature through the analysis of the CO2 2.0 and 2.7 µm bands; water vapor through the H2O 2.6 µm band; and water ice clouds and dust with their broad spectral features between 0.9 and 3.6 µm or their effect on the 2.0 and 2.7 µm CO2 bands. Depending on the SNR and the actual number of observations of Mars atmosphere performed, a detection of CO 2.35 µm band and O2 airglow emission at 1.27 µm could also be achieved. In addition to atmospheric monitoring, MIRS spectra will also allow us to investigate Mars’s surface regolith and its phase angle and potential local-time variation.

In few days, MIRS will be able to complete a global map of Mars and to cover temporal variation of specific area at timescale of hours. Observations performed in ROI mode by MIRS with close observations by the onboard cameras (OROCHI and TENGOO) will be able to observe at even higher temporal resolution and investigate temporally short-lifespan or local-time dependent events in the Martian atmosphere such as the development of orographic clouds and dust storms.

Overall, MIRS will observe the distributions of dust content and storms, water ice clouds and water vapor and will be able to constrain dust and water transport processes in the Martian atmosphere. Observations of water vapor and water ice clouds will provide insights on the formation mechanisms of water ice clouds and their interactions with the dust cycle and topography, on water transport mechanism between polar caps and the equatorial region, and on the interactions between surface/subsurface water reservoirs and the atmosphere. Regarding dust, MIRS will be able to observe the development and evolution of local, regional and global dust storm, which remain poorly understood. The observations should provide key insights into convection processes within these storms and explain how they affect the vertical dust distribution in the atmosphere, the redistribution of dust at the surface and the global dust cycle in general.

Observations during the northern spring and summer season (solar longitude, Ls ~ 0 to ~ 180°), if occurring toward the aphelion cloud belt period, will allow for obtaining an instantaneous global spectral imaging of Mars atmosphere, a daily monitoring of low and mid latitudes water vapor and a local time investigation of clouds evolution. When intense dust activity is predicted around Hellas and the Valles Marineris region, these regions will be privileged targets for dust observations. With observations covering the southern spring and summer seasons (Ls ~ 180 to ~ 360°), during which local and regional storms are ubiquitous on Mars, we will be able to investigate the mechanisms of formation of local/regional/global storms and the diurnal variation of the dust cycle. In particular, the well-known channels for flushing storms and the regions around Hellas will be especially relevant to investigate large-scale dust storms. Observations with the ROI mode will provide high temporal information on both dust and water during the onset of dust storms, helping to further decipher the relation between dust and water in the Martian atmosphere. In addition, MIRS can perform limb observations to further understand how storms transport dust vertically, how detached layers form and subsist in the atmosphere and how these layers evolve over a diurnal cycle.

9 Conclusions

MIRS is an important instrument of the MMX mission, which will operate together with the rest of the payload to ensure a successful mission. MIRS will observe Phobos obtaining global mapping from high, medium and low altitudes. MIRS will support the selection of the two landing sites and will be able to observe during its descent phase down to an altitude of 400 m. The derived composition of Phobos and Deimos will allow us to decipher the origins of the two moons: The detection of phyllosilicates as well as of organics on their surface would imply that the origin of Phobos and Deimos from a giant impact is less likely. In fact, in the case of Phobos and Deimos composition being closer to primitive dark asteroids, similar to carbonaceous chondrites with possible presence of organics, phyllosilicates and/or ices, it would imply an asteroid-type capture origin. While if the composition of the two satellites is containing, even in small quantities, high-temperature phase minerals, representing a mixture of Martian silicates, or if a depletion in moderately volatile elements compared to Mars mantle is measured, this would imply an origin during a giant impact.

MIRS will observe the Martian atmosphere spectroscopically and, together with the onboard cameras, will provide from the equatorial orbit the first opportunity to investigate Martian diurnal variations. MIRS will enhance our understanding of the presence and of the temporal variations of water, CO2 and dust cycles, thanks to long-term monitoring.

MIRS, along with the complete payload of the mission and sampling return, will be able to contribute to the main objectives of the mission on the origin of the Martian moons and on the process of the evolution history of the Martian environment.