Example 1: Aperture photometry on a K2 target

Step 1: Define a pixel extraction mask for a K2 target

We can use kepmask to generate a mask for our target of interest. This mask will be later used to define a photometric extraction aperture. The information about the aperture is stored in an ASCII file called maskfile.txt (this name can be changed using the --maskfile option). The required arguments for this tool are the name of the target pixel file (TPF) and the frame number. Let’s go ahead, open a terminal session, download a TPF using wget, and visualizing the 3000th cadence using kepmask:

$ wget https://archive.stsci.edu/missions/k2/target_pixel_files/c0/202000000/73000/ktwo202073445-c00_lpd-targ.fits.gz
$ kepmask ktwo202073445-c00_lpd-targ.fits.gz --frameno 3000

Step 2: Extract the light curve from the pixel aperture just defined

After creating a mask, we are able to extract a light curve for the target using the kepextract tool. The basic arguments needed for kepextract are the tpf name and a name for the output file. Optionally, we can subtract the background flux using the option --bkg and more importantly, we want to pass our mask created in the previous step.

$ kepextract ktwo202073445-c00_lpd-targ.fits.gz --maskfile maskfile.txt
$ kepdraw ktwo202073445-c00_lpd-targ-kepextract.fits

Step 3: Detrend the light curve for low-frequency structure

First, let’s clip our light curve to only include the part of our interest. That can be done with kepclip. To remove the low-frequency signals in our light curve, we can use the kepflatten tool as follows:

$ kepclip ktwo202073445-c00_lpd-targ-kepextract.fits 2456728.4110787315,2456771.907224878

$ kepflatten ktwo202073445-c00_lpd-targ-kepextract-kepclip.fits --datacol SAP_FLUX --errcol SAP_FLUX_ERR --stepsize 0.2 --winsize 3.0 --npoly 2 --niter 10 --plot --verbose

$ kepdraw ktwo202073445-c00_lpd-targ-kepextract-kepclip-kepflatten.fits --datacol DETSAP_FLUX

Step 4: Perform a photometric correction for motion systematics

To account for the motion of the spacecraft and remove those systematics, we can use the kepsff tool. This tool attempts to identify those centroids in the campaign time series that characterize the spacecraft motion most-accurately and identify those detrended flux measurements that yield the most consistent variability on the timescale of spacecraft roll. A parameterized relationship between the two provides a series of photometric correction factors.

$ kepsff ktwo202073445-c00_lpd-targ-kepextract-kepclip-kepflatten.fits --datacol DETSAP_FLUX --stepsize 5. --npoly_ardx 4 --sigma_dsdt 10. --overwrite

$ kepdraw ktwo202073445-c00_lpd-targ-kepextract-kepclip-kepflatten-kepsff.fits --datacol DETSAP_FLUX

If the option --plot is used in kepsff, a plot like the below is generated.


The four square panels measure the magnitude of motion, identifying motion outliers and thruster firings, before parameterizing the observed correlation between motion and detrended flux (lower-middle panel). The right-hand panels compare detrended aperture flux with it’s corrected counterpart. A series of such corrections and plots are made, one for each time window requested.