Using FPI data to check our pointing for Mars methane obs. on Flight 291

Recall in the March 2016 we observed Mars on two flights, f288 and f291.
Things went well on 288, but on 291, for some reason we had to rush to fit in the methane setting.
The fluxes were hard to understand at the time. For the following, i compare everything to the values for file 40090, where we were centered on Mars.

File           padel offsets       NODN  Relative Continuum Flux

mars.sci.40089        0                3 pairs,   Flux pair 1: 85%, Flux pairs 2 & 3: 3%

mars.sci.40090        0                2 pairs,   Flux 100%.

mars.sci.40091        2.5" L         2 pairs    Flux 215%

mars.sci.40092        2.5" R        2 pairs    Flux 10%

I tried to sort this out using the archival FPI images.
I requested the set from Manuel and Enrico which covered the Mars methane observations on F291, by specifying the start and stop UTC data and time.

They gave me 1080 files of 1024 x 1024 x 1 fits files.
The file names include the UT date and time "block":
Ex: "fpi_node_b.ta_fpi.160324124938.ark.37579.fits"

160324 is 2016 3rd March, and time is 12:49:38. However, that "time" is blocked and doesn't get updated for a dozen or so files.
You can pull that info out of the header with SXPAR(h, "TIME"), which gives back the local time, which in this case is "05:49:55".

The pixel scale is 0.515 arcseconds , but Enrico notes that the last digit changes a bit with different focus values.

I made a subarray of the images (around ~300 x 300), and totaled the rows and columns into two sets of 1-d arrays. Call them xsmush, ysmush.
Then I cross-correlated the xsmush and ysmush with the boresite position and fit CCF with a Gaussian to measure the pixel shift.

Here's what a subimage and xsmush/ysmush of Mars looks like:

After finding the xoffsets and yoffsets from the boresite, I calculated the distance from boresite, and that's what's on the y-axis.
Remember we start on sky for each pair ( dist=44") and then go to Mars, which if centered is at dist=0".

So, we appear to have done what we thought we were doing, except for the 2nd/3rd pair drift in file 89. 

So is huge difference in flux between left/right padel motion a Day/Night thing? I googled and found a Day/Night temp of 200K and 273K at the equator, which comes out to a flux ratio of 0.046, which is exactly the ratio for files 91/92.  I don't want to take the time to look at the phase of Mars, but I guess I'm feeling better about these results being published.