Examples of Data Processing Glitches in Q3
Now that we’re in the thick of Quarter 3, there’s new data glitches popping up in the data unique to Quarter 3. Below are some examples of glitches to watch out for when classifying that were likely introduced during the processing of the raw data to remove instrumental effects. These glitches look like big V-shapes either getting brighter then fainter or getting fainter then brighter. Another type that you might see is what looks like a chunk of the light curve (more than about 20 points) that are offset either up or down compared to the rest of the light curve. These are not due to a transiting planets, and ignore that they are there when looking through the rest of the light curve for transits. A good clue that what you’re seeing might be a glitch due to the data processing is if the large V-shape or large offset datapoints comes before or after a gap in the data.
The gaps in the light curves are not caused by transiting exoplanets. They are spots where we have no data in the star’s light curve. This is caused by Kepler not observing (either because the spacecraft entered safemode or it is rotated back towards Earth to beam the observations to the ground) or by the data being bad (likely due to a cosmic ray hitting the detector and saturating it)
3 responses to “Examples of Data Processing Glitches in Q3”
Trackbacks / Pingbacks
- October 5, 2012 -
Great you confirmed what we suspected already. What kind of detrending algorithm is used by the way?
We use the Pre-Search Data Conditioning (PDC) light curves produced by the Kepler team – You can read more about it here in this paper by Jenkins et al. 2010 http://kepler.nasa.gov/Science/ForScientists/papersAndDocumentation/SOCpapers/7740-12-tps-jenkins-copyright.pdf