Science targets

Throughout the Spirit and Opportunity rover surface operations, MER science team members identified and named locations of interest for carrying out contact science measurements. These locations were defined in a loose hierarchy of feature and target in which a relatively broad feature contained one or more specific targets. Features and targets also were defined for remote sensing measurements by the mast-based instruments Mini-Thermal Emission Spectrometer and Panorama Cameras, and a similar archiving effort is planned to capture those targets. Names given by the science team are not formally recognized by the International Astronomical Union.

A list of contact science targets was not available from the project at the end of the MER mission. At the beginning of the mission, targets were defined by team members using the Science Activity Planning (SAP) tool from which the rover planners could retrieve specific location information for commanding the spacecraft. When SAP was replaced by the Maestro planning tool, target locations were communicated to the rover planners separately, with the result that targets were not always added to Maestro.

Initial work identified resources for enumerating the contact science targets: partial lists from SAP and Maestro along with thousands of sol reports from rover planners, science operations working group documentarians, and the mission managers. From these, we compiled a list of observation targets, using the Basilisk information extraction system to scan these reports and highlight possible target names for human review. Basilisk was trained using a seed target list generated from the planning tools and from instrument lists and plan summaries in the Analyst's Notebook. It learned frequently occurring English patterns in which targets are mentioned, such as “images of X”. Applying these patterns yielded many additional candidate targets.

A team of human reviewers used the brat web-based text annotation system to browse Basilisk's candidate targets in the context in which they were found and then approve or remove each candidate as appropriate. After reviewing 1502 Mission Manager reports, we obtained a list of 326 MER-A and 638 MER-B targets. In addition, several target aliases were discovered, including typos (e.g., “Boabab” for “Baobab”), punctuation variants (e.g., “Lemon-Rind” for “Lemon_Rind”), and phonetic variants (e.g., “Stewart_Island” for “Stuart_Island”). These target references would have been missed using a strict matching search. The resulting list of targets from the Basilisk system was integrated with target lists from instrument teams and planning reports to create a master list of targets.

Having the list of targets, two questions needed to be answered: what science data were acquired, and where was the target? Activity lists from the APXS, MB, MI, and RAT instrument teams were used to match targets to science activities and to find the sols of data acquisition. Target location relative to the rover was gleaned from rover planner reports and cross-checked against planning tool lists when available. The rover location (site and drive) for each target was sourced from the AN. For each target, a representative image showing the target in context is selected as a finder frame. Ideally, a Navigation Camera or Panorama Camera image is used. Candidate images must be taken from the same rover site location where the target was defined. Each target’s ground coordinates and image metadata, the CAHVOR image model is used to auto-locate target positions within an image. Preference is given to full-size (1024 by 1024 pixels) images with the target appearing near the center of the frame. In addition, the distance from the camera to the target position in an image is calculated to avoid near-field shadows. The Notebook image measurement tool is used to validate finder frame selection.

Validation of the contact science target database was provided by MER science team members who participated in each step of its development.