Delta/residual flats from post-cryo L1b frames

Approximately 30,000 L1b (instrumentally calibrated) frames were normalized by their modal pixel value, then stacked using trimmed averaging and renormalized to derive a residual pixel-gain map. Only frames > 35° from the moon and with |b| > 15° were used. This was done separately for an early set of post-cryo scans from v5 testing [08748a - 08925a] and a late set of scans [12181a - 12514a].

Figures 1 and 2 shows these residual gain maps for W1 and W2 for the early and late scans respectively. The uncertainty per pixel in these gain maps is ~ 0.15 - 0.25%. The color bars depict the range of pixel values. Corresponding histograms are shown in Figures 3 and 4 respectively. Bad pixels (NaNs from the input L1b frames) are colored magenta.

For the early post-cryo cryo phase, ~95% and ~99% of the pixels have residual "gain" values falling in the range 0.98 to 1.02 (i.e., within ±2%), for W1 and W2 respectively.

For the late post-cryo cryo phase, ~86% and ~95% of the pixels have residual "gain" values falling in the range 0.98 to 1.02 (i.e., within ±2%), for W1 and W2 respectively. This implies that the relative responsivity variations are being calibrated more accurately in the early phase.

Figure 1 - Left: W1 residual gain map for "EARLY" post-cryo scans. Right: Same for W2 with same stretch [Click to enlarge].


Figure 2 - Left: W1 residual gain map for "LATE" post-cryo scans. Right: Same for W2 with same stretch [Click to enlarge].


Figure 3 - Histograms corresponding to the residual gain maps in Figure 1 made from "EARLY" post-cryo scans. Tails are truncated.


Figure 4 - Histograms corresponding to the residual gain maps in Figure 2 made from "LATE" post-cryo scans. Tails are truncated.



Last update - 22 August 2012
F. Masci - IPAC/Caltech