Seeing Red: Shocked Plagioclase as Pre-Noachian Stratigraphic Tracers
Bibliographic record
Abstract
IntroductionThe Colour and Stereo Surface Imaging System (CaSSIS; Thomas et al., 2017) has been acquiring 4 m/pixel images of the martian surface since 2018, building a database of over 38,000 images in up to four spectral bands across the planet. CaSSIS NIR-PAN-BLU (NPB) and HiRISE IR-RED-BLUE (IRB) images reveal conspicuous, red-toned outcrops (RTO) at a range of scales that are distributed across the southern highlands. While visually striking, the composition, distribution, and origins of these outcrops remain unclear. Here, we present a global analysis of RTOs aimed at resolving this.MethodsTo investigate the RTO, we manually identified a selection of ~30 examples of RTO covering a range of sizes, morphologies, textures, colours, and lighting geometries in CaSSIS NPB images, and used these examples as a training dataset for a convolutional neural network (CNN), which is able to detect occurrences of RTO similar to those in the training dataset (see techniques in Bickel et al., 2024 and references therein).The CNN assessed all >38,000 images in the CaSSIS database (as at 28/02/2024), and produced an output of 2232 detections at >60% CNN confidence level. These outputs were manually sorted to remove false positives and to assess RTO morphology/texture. True positive RTOs, as well as those from the training dataset, were compared with CRISM footprints, and those CRISM MTRDR cubes that overlapped were analysed following the standard ratioing method (e.g. McGuire et al., 2009; Murchie et al., 2007).ResultsWe observe a total of 923 “true positive” RTO CaSSIS detections (e.g. Figure 1). RTOs are most common in the areas north of the Hellas and Argyre basins, in lower elevation regions of eastern Valles Marineris, and in the Nili Fossae region. These usually crop out in the ejecta of large (>10 km) craters in these areas but can also be observed in the walls of craters and on relatively flat-lying terrain. Most RTOs are associated with Noachian-aged terrains with 76% occurring within Noachian Units (Tanaka et al., 2014), 12% in Hesperian Units, 4% in Amazonian Units, and 8% in Hesperian/Amazonian Impact Units.RTOs typically possess a blocky, massive appearance (Figure 1a), and occur as both discrete outcrops with well-defined margins (Figure 1b), interpreted as ejecta blocks, or as diffuse deposits (Figure 1c), interpreted as intermixing with other components within ejecta blankets.CRISM spectra from RTOs reveal a consistent and broad, yet extremely weak and slightly long-shifted 1.25 μm absorption feature, consistent with partial transformation of Fe-plagioclase to maskelynite, a diaplectic amorphous glass associated with shock metamorphism from hypervelocity impact events (e.g. Jaret et al., 2015; Spudis et al., 1984).Figure 1: False-color examples of Red-Toned Outcrops (RTOs) from across the Hellas region. a) Large blocky RTO in Terby Crater (CaSSIS NPB); b) Discrete fractured RTO north of Hellas (HiRISE IRB); c) Large diffuse RTO with some sharp boundaries NW of Hellas (CaSSIS NPB); d) RTO under younger dark material NE of Hellas (CaSSIS NPB).DiscussionWe infer RTOs to be the products of large basin-forming impacts, with their current distribution shaped by subsequent impact gardening. Their occurrence surrounding Hellas and Argyre suggests that these units are part of the original basin ejecta fields and therefore, at least in the case of Hellas, predate the Noachian. In these regions, RTOs likely derive from Fe-plagioclase-bearing massifs uplifted from depth during the impact events, now observed as degraded massifs along the basin rims (Phillips et al., 2022; Phillips and Viviano, 2025).RTO exposures are commonly found within ejecta in impact-modified terrains superposed on the Hellas and Argyre ejecta blankets. The absence of similar features in Noachian-aged terrains outside of these ejecta fields suggests that large basin-forming impacts are a pre-requisite to exhume deeply buried plagioclase material, and that subsequent impact gardening of basin ejecta is a primary mechanism for their exposure.Collectively, the observed distribution, morphology, and mineralogy of RTOs point to the presence of a plagioclase-rich lower crustal component that was widespread in the Pre-Noachian martian crust, likely explaining the lower-than-expected observed density of the martian crust (Knapmeyer-Endrun et al., 2021; Bouley et al., 2020; Baratoux et al., 2014). The evidence indicates that this plagioclase-rich material underwent shock metamorphism, excavation, and emplacement within the ejecta of early basin-forming events, providing a unique window into the composition and modification history of Mars’ ancient crust.ReferencesBaratoux, D., et al. (2014) Journal of Geophysical Research: Planets, vol. 119, no. 7, pp. 1707–1727 Bickel, V. T., et al. (2024) Scientific Data, vol. 11, no. 1, p. 845 Bouley, S., et al. (2020) Nature Geoscience, vol. 13, no. 2, pp. 105–109 Jaret, S. J., et al (2015) Journal of Geophysical Research: Planets, vol. 120, no. 3, pp. 570–587 Knapmeyer-Endrun, B., et al. (2021) Science, vol. 373, no. 6553, pp. 438–443 McGuire, P. C., et al. (2009) Planetary and Space Science, vol. 57, no. 7, pp. 809–815 Murchie, S., et al. (2007) Journal of Geophysical Research E: Planets, vol. 112, no. 5, pp. 1–57 Phillips, M. S. and Viviano, C. E. (2025) LPSC 56, Houston, Texas, abs#1653 Phillips, M. S., et al. (2022) Geology, vol. 50, no. 10, pp. 1182–1186 Spudis, P. D., et al. (1984) Journal of Geophysical Research, vol. 89 Thomas, N., et al. (2017) Space Science Reviews, vol. 212, no. 3–4, pp. 1897–1944
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.006 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".