5
$\begingroup$

Last 65 days of Kepler's light curve for the infamous KIC 8462852 star is shown below from the published paper's graph. Skip the dips here, I wonder about two other features:

  • Day 1559, during a period which is calm but also the very brightest, the light curve instantly and permanently drops by half of a percent.
  • Days 1581 to 1587 the light curve is exactly flat (also days 803-809 not shown here). Is the real data for those periods lost?

Are such features commonplace or another particularity for KIC 8462852?

How wide would a confidence interval typically be for a finished Kepler light curve when all adjustments for telescope movements and whatnot have been made?

enter image description here

$\endgroup$
5
  • 1
    $\begingroup$ Can't be certain at this point, but I'd bet those are just data artifacts. My guess is the sudden drop is due to a recalibration and the flat line is just connecting actual data over missing data points. $\endgroup$
    – zephyr
    Mar 21, 2017 at 12:59
  • 1
    $\begingroup$ Do you know the actual dates that these days correspond to? That may be useful. $\endgroup$
    – zephyr
    Mar 21, 2017 at 13:04
  • $\begingroup$ @zephyr Kepler ended its (first) mission on July 14, 2012, so I suppose that the day numbering on the X-axis can be converted into real dates by counting backwards from last day 1592. NASA source I have looked at the Kepler databases, but it's a bit above my paygrade, and it's certainly safer to ask someone who lives with this thing. $\endgroup$
    – LocalFluff
    Mar 21, 2017 at 13:21
  • $\begingroup$ @zephyr Part of the question is if these apparently non-astrophysical features (not talking about the famous dips here) are special for this particular star, or if they are general for Kepler data. If the extremely sensitive Kepler can be recalibrated by ½% like this, how does that relate to the 2% dimming during Kepler's last 200 days, found in relative photometry? And what does it imply for the "calibration uncertainty" of the 100 year dimming found in historic photo plates? $\endgroup$
    – LocalFluff
    Mar 21, 2017 at 13:41
  • $\begingroup$ Hopefully my answer clears up your specific question. The questions you just asked however are also really good ones that would probably be worth asking another question on here. $\endgroup$
    – zephyr
    Mar 21, 2017 at 13:45

1 Answer 1

6
$\begingroup$

There are two separate points of interest you're looking at so I'll separate this into sections.

Sudden Drop at Day 1559

As near as I can tell, this is the result of a quarterly roll of the satellite, specifically the end of quarter 17. Every 90 days, which NASA calls a quarter, the space craft does a 90 degree roll to optimize the solar panel efficiency. What this means for the data is that when this roll occurs at the end of the quarter, the photons from the star now fall on a different CCD channel which is calibrated slightly differently, causing a slight discontinuity in the data. You can read about "Quarter Stitching" here. I've extracted some relevant text and an image for posterity.

After each roll maneuver, most Kepler targets fall on a different CCD channel and the target’s point-spread function will be distributed differently across neighboring pixels. Naturally this redistribution requires a new computation of the target mask and optimal aperture size within the Kepler pipeline, taking into account the point-spread function, new CCD characteristics, and new estimates of nearby source crowding. The operational outcome is often a different mask shape, with differing amounts of flux within the optimal aperture from both the target and contaminating sources.

enter image description here Light curves extracted from all single pixels within the quarter 4, 5 and 6 Target Pixel Files for KIC 9603833 (the symbiotic star StHA 169). Gray pixels comprise the optimal apertures that yield the archived light curves. The target point-spread function is distributed across neighboring pixels differently from quarter to quarter and hence the optimal aperture varies in size and shape from quarter to quarter.

Flat line between days 1581 to 1587

I'm less certain about this one. I believe the flat line is just missing data. To keep the line continuous, they just connected the data points they had with a straight line.

Why the data is missing, I can't be sure. I know that Quarter 17 ended with the failure of Kepler's second reaction wheel so my guess is that this missing data is in some way related to these reaction wheel troubles. I do know they tried,in the past, to put the spacecraft in 10-day "wheel rest" mode in hopes that it would give the reaction wheels a break, let them get lubed up again, and be better when they restarted. Perhaps, by this stage, they had already suspected a second reaction wheel problem and were trying out a wheel rest mode. I can't find any evidence a wheel rest occurred during Quarter 17 though, so this is just my guess.

$\endgroup$
2
  • $\begingroup$ Great! But wouldn't momentary vertical shifts occur at the end of the "wheel rests", if the star didn't happen to have the same brightness when the rests ended as when they started? Or maybe they normalized by assuming it was the same. Could very well be that the authors (Boyajian+) of the paper I picked that graph from didn't care about cleaning out all such details since they clearly are irrelevant to the dips they'd found. $\endgroup$
    – LocalFluff
    Mar 21, 2017 at 14:10
  • $\begingroup$ Perhaps the star just happened to have a nearly identical flux at the beginning and end of the missing data period, making the connecting line appear flat. Like I said, I'm less certain about the cause of the flat portion and I wasn't able to find anything definitive, so your guess is as good as mine. $\endgroup$
    – zephyr
    Mar 21, 2017 at 14:13

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .