Table 1. Summary of the total number of images of the Calibration Targets taken by Mastcam-Z in each filter within the first 350 sols on Mars. For each filter, the corresponding spectral range is reported.
These images were captured by Mastcam-Z at several illumination geometries relative to the cal-targets. The illumination geometry of an observation is described by three angles. We define these three angles as incidence , emission and azimuth . The incidence angle (\(0\leq i\leq 90\)) is measured between the zenith of the target and the Sun-target direction, the emission angle (\(0\leq e\leq 90\)) is measured between the zenith of the target and the detector-target direction, while the azimuth angle (\(0\leq\text{Az}\leq 180\) for hemispheric symmetry) is measured on the plane of the target and is the angle between the planes of incidence and emission. The angles \(i\) and \(e\) are measured from the zenith. The emission angle is fixed by the geometry of the rover with the primary and secondary targets seen by Mastcam-Z under emission angles of approximately 58 and 54, respectively. The overall distributions of the incidence and azimuth angles of the cal-target images are reported in Figure 3. A total of 197 cal-target sequences (which include images in only L0/R0 or all filters) were acquired by Mastcam-Z over the first 350 sols, each of which can be identified by its incidence and azimuth angles. The left plot in Figure 3 shows the combinations of incidence and azimuth angles of those 197 sequences. Summing the data points along the two axes of the plot yields the two histograms of azimuth and incidence in Figure 3 on the right. Given the changing orientations of Perseverance, the azimuth angles are evenly distributed, though with a larger frequency in the 5-10 interval, which corresponds to the Sun being approximately behind Mastcam-Z during the observation (Sun in front of the rover), and a much lower frequency of observations at opposite azimuth. In addition, most observations were made with the Sun relatively high above the horizon, with 88% of images taken at incidence angles lower than 30 and a peak around 10-15.
As studied by Buz et al. (2019) and by Kinch et al. (2020) in the pre-flight tests, at the typical emission angle of 58 all the color and grayscale materials of the cal-targets (though to a lesser extent for the white patches) show a strong forward scattering peak with incidence angles \(>45\). The smaller number of observations on Mars with the Sun opposite to the camera (high azimuth) reduced the impact of the forward scattering, which in some cases (\(i\geq 50\)) was quite significant. In addition, at higher incidence angles the diffuse light acquired importance within the dusty atmosphere, reducing the contrast between the sunlit and shadowed regions. Early martian morning or late afternoon observations proved challenging for estimating the solar irradiance as part of the reflectance calibration, and this will serve as a starting point for improvements in our data processing in the future. The best estimates of the irradiance were obtained when the Sun was high in the sky, with a smaller atmospheric path length and negligible forward or backward scattering. Mastcam-Z acquired most cal-target images at incidence angles lower than 30. However, since the diffuse light had a major impact at low Sun, this geometry was particularly useful to monitor the dust suspended in the atmosphere. This was done by comparing the radiance from the regions in the gnomon shadow on the grayscale rings to those in the direct sunlight (see sect. 4.3.2).