Spatial Spectroscopic Models for Remote Exploration

Abstract

Ancient hydrothermal systems are a high-priority target for a future Mars sample return mission because they contain energy sources for microbes and can preserve organic materials (Farmer, 2000 ; MEPAG Next Decade Science Analysis Group, 2008 ; McLennan et al., 2012 ; Michalski et al., 2017 ). Characterizing these large, heterogeneous systems with a remote explorer is difficult due to communications bandwidth and latency; such a mission will require significant advances in spacecraft autonomy. Science autonomy uses intelligent sensor platforms that analyze data in real-time, setting measurement and downlink priorities to provide the best information toward investigation goals. Such automation must relate abstract science hypotheses to the measurable quantities available to the robot. This study captures these relationships by formalizing traditional "science traceability matrices" into probabilistic models. This permits experimental design techniques to optimize future measurements and maximize information value toward the investigation objectives, directing remote explorers that respond appropriately to new data. Such models are a rich new language for commanding informed robotic decision making in physically grounded terms. We apply these models to quantify the information content of different rover traverses providing profiling spectroscopy of Cuprite Hills, Nevada. We also develop two methods of representing spatial correlations using human-defined maps and remote sensing data. Model unit classifications are broadly consistent with prior maps of the site's alteration mineralogy, indicating that the model has successfully represented critical spatial and mineralogical relationships at Cuprite. Key Words: Autonomous science-Imaging spectroscopy-Alteration mineralogy-Field geology-Cuprite-AVIRIS-NG-Robotic exploration. Astrobiology 18, 934-954.

Exploring, Mapping, and Data Management Integration of Habitable Environments in Astrobiology

“…Mars sensors have observed vibrational overtone absorption features associated with phyllosilicate minerals and isolated occurrences of carbonate overtone features in Noachian-aged terrains (Bibring et al, 2005; Mustard et al, 2008; Niles et al, 2013) and water and sulfate overtone features in early Hesperian terrains (Bibring et al, 2005; Gendrin et al, 2005). Terrestrial airborne imaging spectroscopy using platforms such as NASA’s “Classic” and “Next Generation” Airborne Visible Infrared Imaging Spectrometers, AVIRIS-C (Green et al, 1998) and -NG (Thompson et al, 2018a) respectively, have been used to map sites of hydrothermal alteration and unique sedimentary environments (e.g., Kruse, 1988; Crowley, 1993; Thompson et al, 2018b). Such datasets have also been used to detect minerals generated through microbiologic activity (e.g., Anderson and Robbins, 1998).…”

“…Thus, new algorithms and artificial intelligence can find and recognize image patterns at multiple scales: from the micro-scale, where biogenic structures may be visible; to the mesoscale, where geologic fabrics and boundaries indicate compositional units with common properties; to the macro-scale, where wide-area physical processes indicate regions where preservation potential is greatest. Spectroscopic methods for automated mapping, characterization, and anomaly detection allow automatic pattern identification in similar fashion (Thompson et al, 2015, 2018b) to interpret mineralogy. A third important component is a geostatistical model capable of inferring spatial and contextual relationships from measurements made at multiple scales (Thompson et al, 2018b).…”

“…Spectroscopic methods for automated mapping, characterization, and anomaly detection allow automatic pattern identification in similar fashion (Thompson et al, 2015, 2018b) to interpret mineralogy. A third important component is a geostatistical model capable of inferring spatial and contextual relationships from measurements made at multiple scales (Thompson et al, 2018b). This can also integrate: standard probabilistic models drawn from geostatistics (Cressie and Cassie, 1993); factor analysis, which decomposes a dataset into component trends that may vary independently, and may imply different physical effects (Bishop, 2006); and potentially latent semantic analysis, which infers the pure “topics” represented by a dataset, even if no single instance expresses just one (Blei et al, 2003).…”

“…Biosignatures (e.g., elemental, mineral, textural, or other scientific evidence of life) have mappable distributions at microscopic (<mm) to macroscopic (>km) levels. These scales of investigation can be linked via probabilistic models , statistical representations that represent dependencies permitting inference and prediction between the scales (Thompson et al, 2018b; Chan et al, 2018, 2019). Based on observations of Earth analogs, it is expected that extraterrestrial life also requires habitable environments – places where environmental conditions are conducive to sustain life (Preston and Dartnell, 2014; Cockell et al, 2017).…”

Expert Guided Rule Based Prioritization of Scientifically Relevant Images for Downlinking over Limited Bandwidth from Planetary Orbiters

“…We then determine the relative importance of the instances based on these attributes, where it is crucial for the module to understand the content in the image like an expert. Prioritization has been studied with planetary rover data to select rocks with a specific target signature (Estlin et al 2012), for autonomous exploration by integrating past orbital datasets with a science hypothesis (Thompson et al 2018). We focus on formulating a generalized representation of the already observed data and allow the module to determine if a new instance is unusual.…”