Figure 14. (a) Time evolution of the relative difference between the slopes of the two- and one-term fits for four narrow-band filters. Each color corresponds to a different filter. The relative difference is computed as\([(\text{slope}_{two\_term}-\text{slope}_{one\_term})/\text{slope}_{one\_term}]\). The circles are observations with \(i\leq 30\), the triangles are\(i>30\). The vertical dashed line is the martian aphelion. (b) Average of the relative difference of slopes for all 12 narrow-band filters, over the range from sol 100 to 180 (corresponding to the two vertical solid lines in panel (a)).

The offsets for all the two-term fits over time from landing to sol 350 are represented in Figure 15 for the primary clean spots, the grayscale rings, and the secondary horizontal tiles. The white patch was excluded from the computation of the fits for the primary clean spots and the secondary horizontal target, but not for the grayscale rings, since it was not observed there. The offsets are calculated in radiance units, which we converted to units of IOF for easier comparison by filter-to-filter and sol-to-sol. The offset values are roughly wavelength dependent, increasing with increasing wavelength. Most filters fall fairly close in a tight band between 0.05 and 0.1 while the shortest and longest wavelengths fall outside this range. The clean spots (Figure 15a) and secondary horizontal target (Figure 15d) are quite similar, while the grayscale rings (Figure 15c) are characterized by a more dramatic divergence in the near infrared filters (R4-R6, from 900 nm to 1000 nm), though also in this case the L6 is consistently well below the other filters. The average increase of the offsets for the grayscale rings was 14% and for the secondary horizontal target 25%, on average. In particular, for the secondary horizontal target the increase rate became flatter towards the longer wavelengths (from 56% in L6 to 7.7% in R6). The grayscale rings present a more random distribution, with the filters R4-R6 (from 900 nm to 1000 nm) following a slight decrease (-8% in R6). The offsets are quite stable in time showing a limited increase over the 350 sols. Figure 15b shows the offsets of the clean spots as function of the wavelengths of the 12 narrow-band filters, where the offsets in each filter have been averaged over the first 100 sols (blue circles) and the last 100 sols (red circles) considered in our analysis (sols 0 to 350). The spectral shape of the offsets showed a smooth trend with first increase between L6 and L2 (442 nm and 754 nm, respectively), followed by a slight decrease and a final increment from R2 (866 nm) to R6 (1022 nm). In addition, the mean offsets in the 250-350 sol interval were 14% higher than those in the first 100 sols of the mission, with a minimum increase in the near infrared (6.3% in R6) and larger increase at shorter wavelengths (20.7% in L6).