Earlier work · Laboratory spectroscopy
What makes a good analogue of Phobos?
During my master's internship in Japan, I studied laboratory-made Phobos analogues: how their grain size changes the light they reflect, and how those spectra compare with spacecraft observations when sampled for MMX's instruments.
The question
Phobos and Deimos raise a question about the history of Mars: were they captured, or did they form from impact debris? Spectra offer clues about surface materials, but composition, grain size and surface processing can leave overlapping signatures.
My study tested UTPS-TB, a Tagish Lake-based simulant developed as an analogue for the asteroid-capture scenario. The aim was to evaluate its spectral behaviour and its limits as a comparison material.
Phobos and Deimos would originally have been asteroids orbiting the Sun, later captured by Mars's gravity. Their dark, asteroid-like surfaces motivate this idea, but explaining their nearly circular orbits close to Mars's equatorial plane is difficult.
A large impact on early Mars would have launched debris from both Mars and the impactor into orbit. Phobos and Deimos could then have formed as that debris accumulated into moons. This scenario links their composition to a mixture of Martian and impactor material.
Phobos was the experimental focus. Deimos belongs to the wider origin question; the spectral comparisons shown here use Phobos observations.
In the laboratory
Holding the mixture constant makes it possible to examine how particle size and sample preparation affect reflectance and spectral slope.
I separated the simulant with sieves, prepared the sample holders and measured their reflectance with the VERTEX 70v. The spectra were calibrated with Spectralon and Infragold, then compared in absolute and normalised form.
The “finer than” fractions are cumulative and overlap; these are not six disjoint size bins. The smallest sieve used was 53 µm.
Through MMX's eyes
A laboratory spectrum contains more detail than a camera's filter bands. I sampled the laboratory and CRISM spectra under the same instrument assumptions, then compared brightness, spectral shape and absorption features.
The report averages the spectra across seven nominal filter bands. Normalising at 0.55 µm removes the overall brightness scale so that differences in spectral shape are easier to compare.
Absolute reflectance · %
Normalised at 0.55 µm
Absolute reflectance retains brightness differences between the analogue and Phobos. Those differences are not removed by choosing a grain-size fraction.
Normalised reflectance highlights shape. Finer analogue fractions generally show redder slopes, while the Phobos curves still differ from the tested material.
The near-infrared comparison looks at continuum slope and the hydroxyl-related absorption in the analogue near 2.7 µm. The report normalises these spectra at 2.5 µm.
Absolute reflectance · %
Normalised at 2.5 µm
Reading the legends. The six grain-size entries identify the UTPS-TB laboratory fractions listed above. Phobos “red” and “blue” refer to spectral units on the moon, rather than grain-size classes.
What I learned
The tested fractions show differences in reflectance, slope and absorption depth. Packing and coarse-grain geometry also matter; brightness is not a simple grain-size ruler.
Normalisation helps compare spectral shape, but UTPS-TB does not reproduce every aspect of the Phobos observations. Composition and space weathering remain possible contributors to the mismatch.
This study does not establish Phobos's grain size, confirm hydrated minerals on its surface, or decide between the origin scenarios. It identifies comparisons and caveats for interpreting MMX observations.
Master's internship, 2023: three months at AIST in Tsukuba with Moe Matsuoka and one month at Tohoku University in Sendai with Tomoki Nakamura. The report also includes Ryugu sample work; this page focuses on the UTPS-TB spectroscopy and Phobos comparison.
The report describes a roughly 5 mm footprint, 30° incidence, 0° emission and 2,048 scans per sample in the FTIR measurements. These laboratory conditions differ from spacecraft observing geometry.
| Band | Centre · µm | Width · µm |
|---|---|---|
| ul | 0.39 | 0.05 |
| b | 0.48 | 0.03 |
| v | 0.55 | 0.03 |
| d | 0.65 | 0.04 |
| e | 0.73 | 0.03 |
| x | 0.86 | 0.03 |
| p | 0.95 | 0.06 |
The rectangular diagram represents these nominal widths. It is not a measured spectral-response curve or an end-to-end flight-instrument simulation.
The report discusses very fine grains, porosity, organic material and space weathering as possible explanations. Spectral similarity alone does not uniquely recover those properties or validate an origin scenario.
The 2.76 µm value used for one band-depth comparison is the nearest available CRISM sample to the analogue's 2.71 µm feature. Thermal correction and incomplete coverage limit that comparison.
The graphs are approximate digitisations of the JPEG images embedded in the 2023 report. Both panels retain their original axes and comparisons. The OROCHI redraw shows only the band-average curves; MIRS retains the visible spectral traces and CRISM scatter. JPEG compression and overlapping curves prevent exact recovery of coordinates or marker counts. The downloadable points are for displaying these comparisons, not for new scientific inference or band-depth calculations. Original numerical measurements are not available here.
