Signal to Noise Ratio and Spectral Sampling Constraints on Olivine
Detection and Compositional Determination in the Intermediate Infrared
Region: Applications in Planetary Sciences
Abstract
The intermediate infrared region (IMIR, 4 – 8 µm) provides significant
advantages over the visible-shortwave infrared and mid-infrared for
quantitative determination of mafic mineral composition. In particular,
olivine’s sharp spectral features in IMIR spectra exhibit systematic
shifts in wavelength position with iron-magnesium content. Previous IMIR
studies have used laboratory data, with signal-to-noise ratios (SNRs)
and spectral resolutions greater than those expected of imaging
spectrometers. Here we employ a feature fitting algorithm to
quantitatively assess the influence of SNR and sampling rate on olivine
detection and compositional interpretation from IMIR data. We
demonstrate that olivine is easily distinguished from pyroxene and other
lunar-relevant minerals across IMIR wavelengths, with the
feature-fitting algorithm effectively determining olivine composition
for various synthetic, terrestrial, Martian, and lunar samples with an
average error of only 6.4 mol%. We then apply the feature-fitting
routine to degraded spectra with reduced SNRs and sampling rates,
establishing data-quality thresholds for accurate determination of
olivine composition. Spectra for the sample most relevant to lunar
exploration, an Apollo 74002 drive tube consisting of microcrystalline
olivine and glass-rich pyroclastics, required SNRs ≥ 200 for sampling
rates ≤ 25 nm to predict composition within ±11 Mg# (molar
Mg/[Mg+Fe] * 100) of the sample’s true composition. Derived limits
on SNRs and sampling rates will serve as valuable inputs for the
development of IMIR imaging spectrometers, enabling comprehensive
knowledge of olivine composition across the lunar surface and providing
valuable insight into the Moon’s crustal history and thermal evolution.