Figure 6. Calibration Target spectra from sol 12 and 339. The circles are the observed calibrated R* reflectances of the clean spots, while the solid lines represent the corresponding laboratory spectra.
The Radiometric Coefficients Files
A major portion of the cal-target sequence processing included the generation and storage of information on the images and their ROI selections. We refer to these as Radiometric Coefficients files, or simply RC-files, which were created simultaneously with the linear fits described in section 3.2. The first part of a typical RC-file included metadata (e.g., name of the corresponding cal-target image, local martian time of the original image). The body of the file included a list of arrays, each of which was formed by 41 fields (one for each ROI selection). The algorithm that selected the ROIs for each region included the complete name of the region, its average radiance over the pixels and uncertainty of this radiance, number of pixels selected, illumination geometry angles and model reflectance related to the color material and to the geometry of the region. The inverse of the slope of the linear fit described in section 3.2 was also included in the RC-files, along with three Boolean arrays of flags (expressed in terms of 0 and 1) that specified which regions had been selected, which ones were “bad” and should not be used (usually due to unwanted shadows), and which should be used to make the linear fits for calibration. We used a graphic interface to mark the regions appropriately. Whenever a region had not been selected, its place in the data arrays of the RC-files did not show any numerical value. An example of an RC-file is given in appendix 1.
Results
All quantitative results that we present here have been obtained from the values stored in the entire set of RC-files covering the time range from landing to sol 350. In each color image in L0 and R0, one RC-file was extracted individually from each of the three broad-band channels (L0B, L0G, L0R, and R0B, R0G, R0R). Hence, the total number of RC-files in our set was 3366. The regions of the cal-targets that were not selected or were marked as bad were not included in the following analysis.
Imaging of the Calibration Targets on Mars
The considerable number of color images of the cal-targets acquired by Mastcam-Z allowed a regular visual check of their conditions over time, including the accumulation and displacement of airfall dust and monitoring the color and grayscale materials.
Figure 7 is a comparison of four selected Bayer RGB images of the cal-targets taken by the left eye of Mastcam-Z. The most evident details that can be visually evaluated in time are the progressing change of hue of the primary white patch (that effect is reflected in the fits shown in section 3.2) and the accumulation of dust and sand on the surfaces. The former is commonly referred to as “yellowing” effect that degraded the reflectivity at shorter wavelengths of the visible layer of the white material (AluWhite98) since landing. The term “yellowing” merely comes from the fact that the white patch was noted to fade towards a yellowish tone in the color images. A more detailed treatment of the problem is reported in section 4.5.
Figure 7 also illustrates the progressive increase in the amount of magnetic dust with time on the primary magnet rings, manifested as an overall dark reddish color. This outcome, which was also observed on the MER (Madsen et al. , 2009) and MSL (Bell et al. , 2017) rovers, was expected from the cal-target design. The underlying colors, that were already attenuated but still visible after the first sols (Figure 7a), were not as discernible at the end of the period under analysis in this work (Figure 7d). On other surfaces, such as the other parts of the cal-targets or the deck surroundings, the deposition seemed to be ruled by either original deposition of small grains during the landing event, wind activity capable of transporting larger grains on top of the deck, and/or airfall dust. In general, airfall dust and larger grains were displaced frequently, allowing the cal-targets to remain rather clean. In addition, the deposition of dust on the secondary target was affected by its position on a vertical surface, which would limit the fraction of dust falling on the horizontal tiles. During the first 350 sols of the mission Perseverance experienced several significant episodes of high wind and local dust lifting. The strongest event occurred between sols 314 and 316 (Lemmon et al. , 2022), with consequences visible for several sols (Figure 7d). Appreciable layers of brownish airfall dust and sand were conspicuous on the grayscale rings, on the deck next to the primary target (including the vertical side of the golden base), and on the secondary horizontal target following this event.