Example 4: Removing Stellar Astrophysics and Quarter Stitching

The purpose of this tutorial is to flatten both systematic artifacts and astrophysical signal in order to fold candidate planet transits successfully upon their orbital period and characterize them with a physical model.

For this purpose, we download quarter 10 and 11 of KIC 12557548:

$ wget http://archive.stsci.edu/missions/kepler/lightcurves/0125/012557548/kplr012557548-2011271113734_llc.fits
$ wget http://archive.stsci.edu/missions/kepler/lightcurves/0125/012557548/kplr012557548-2012004120508_llc.fits

We can use kepstitch to stitch those light curves together:

$ kepstitch kplr012557548-2011271113734_llc.fits kplr012557548-2012004120508_llc.fits --outfile kplr012557548-kepstitch.fits

We can visualize the stitched light curve using kepdraw:

$ kepdraw kplr012557548-kepstitch.fits --datacol SAP_FLUX
$ kepdraw kplr012557548-kepstitch.fits --datacol PDCSAP_FLUX
../_images/kplr012557548-kepstitch-kepdraw.png ../_images/kplr012557548-kepstitch-kepdraw-pdcsap.png

With systematic artifacts already mostly-removed, the light curves need to be flattened but in a way that does not radically contaminate the transit profile with new systematic structure. The PyKE tool kepflatten fits piecemeal low-order polynomials on small timeslices aross the time-series. Fitting makes user of iterative σ-clipping that removes statistical outliers from a best-fit before refitting until no more time-stamps are rejected. In this way, transits profiles are conserved during the fit process and subsequent normalization by the final best-fit. The flattened are written to the output file in two new columns, DETSAP_FLUX and DETSAP_ERR_FLUX.

$ kepflatten kplr012557548-kepstitch.fits --nsig 3 --stepsize 1 --npoly 2 --niter 10 --plot

Now, we can apply kepfold to fold our final light curve:

$ kepfold kplr012557548-kepstitch-kepflatten.fits 0.653534 2455372.883 --threshold 3. --niter 5 --bindata --binmethod median --nbins 100 --plottype det