How targets are defined

Science targets are included in the MER Opportunity and Spirit and MSL Curiosity Notebooks. For each rover, the science team defined a target by noting its position on an image (called a "locator image" in the Notebook). Knowing the rover's location and camera pointing information when the locator image was acquired, we can determine the target's position on other images acquired nearby.

In the Notebook, a target is associated with the sol on which it was defined. When using the Sol Summaries or Target Search, a target will only be associated with the sol on which it was defined, even if observations of the target were made on later sols.

Target names, and why can't I find a target I know exists?

Target names are tricky things. The names are not approved by the IAU or designated as having "the name NASA uses". Targets are defined and named during surface operations, but other names may be applied informally to the same target by various instrument teams and working groups. In addition, a target may be referenced in publications by different names.


The Opportunity and Spirit science team used a hierarchy of a feature (more general) that comprised one or more individual targets (more specific). In the Notebook, we use the target names. Features are listed in Target Search, but are not integrated in the name search function at this time. In addition, target aliases used by the APXS, Moessbauer, Microscopic Imager, and Rock Abrasion Tool (RAT) teams are not referenced by the Notebook search. The full list of MER target feature names and instrument team aliases for contact science targets is given in the MER target archive bundle at the PDS Geosciences Node.

Note that only MER contact science targets are included in the Notebook. Remote sensing targets were not systematically captured during the mission and a formal list does not exist at present.


The Curiosity science targets are taken from the science team planning tool in which targets were defined by the team. Other names may be used informally by team members and in the literature. The Mars Target Encyclopedia literature references in the Notebook formal science team target names.

Finding other images of a target

Because we know a target's position relative to where the rover was located, we can find other images of a target. For images acquired at the same rover site and drive position as the locator image, we know the position well because the rover has not moved since the locator image was acquired. For an image acquired nearby the locator image rover position, we can approximate the target location, with uncertainty due to slip and skid associated with rover movement. As the distance increases between the locator image rover position and any other rover position, the uncertainty increases.

In the Notebook, we present a target with its locator image and a list of other images of the target. The list is one of the views on the Target detail page. The list is restricted to images acquired at the same site as the locator image and within 50m of the locator image rover position.

Targets in the Image viewer tool

The Image viewer automatically displays targets on an image when available. Targets are categorized by how well their position on the image is known (see discussion above). For images acquired at the same rover site and drive position as the locator image, a target's position is denoted as "known". For images acquired at other locations, the target's position is listed as "approximate".

Auto-location and position uncertainty

As described above, target locations on an image might be well know or approximate depending on the rover's location when an image was acquired. A systematic study of target placement uncertainty on images has not been carried out. Review of select images by science team members shows that auto-located target offsets in the Notebook are well under the expected uncertainty of up to 50 cm.

The Notebook auto-location function is not rover-aware or terrain-aware. For a given image, rover hardware may be positioned between the camera and target, resulting in the target being drawn on top of the rover hardware. Likewise, a small hill might lay between the camera and a distant target, resulting in the plotted target appearing on the near side of the hill.

Further information of algorithms used is available upon request.