Can scatter of in-situ geochemical data be diagnostic for morphological and petrological features on lava surfaces?

Abstract:

The traditional ideal geochemical whole-rock analysis captures the freshest, most homogenous material through precise collection and extensive sample preparation techniques. Additionally, analyses which have low (Σ <98 wt%) or high (Σ >102 wt%) totals are frequently replaced with analyses with totals within the 100±2 wt% threshold. Geochemical data collected on planetary surfaces by in-situ instrumentation will likely be constrained by 1) imperfect surfaces with minimal to no sample preparation and 2) the time to redo analyses with low totals. Thus, understanding the significance of geochemical analyses outside the desired totals would increase the scientific return of exploration of volcanic terrains on Mars and on the Moon. Lower totals acquired by a portable x-ray fluorescence (XRF) spectrometer may be attributed to higher water content or greater abundance of low-atomic weight elements (Z < 13). Whereas spot-analyses by portable laser-induced breakdown spectroscopy (LIBS) may vary greatly depending on what phase (glass vs. mineral) the ~50 µm laser hits. Here, we explore the conditions which vary between field and laboratory collection of this data to better understand how imperfect in-situ geochemical data can be interpreted. We present a data set of XRF and LIBS analyses of Martian analog basaltic lava surfaces, exploring the characteristics of the flow (e.g. surface roughness, vesiculation, alteration, and crystal content) and how they correlate to overall totals in XRF data and scatter in the LIBS data. Using these data sets in tandem may allow us to infer petrological changes within a suite of volcanic deposits not just despite the scatter, but because of it.