Name | Last modified | Size | Description | |
---|---|---|---|---|
Parent Directory | - | |||
experiment-1/ | 2022-09-13 08:09 | - | ||
This page contains simulated data and parameters for direct imaging sequences with various injected planets following simulated orbits. The goal is to provide data for certain benchmark cases to compare different image orbit modelling approaches.
In this experiment, we begin with real ADI Keck/NIRC2 data taken in the L' filter (Lp), or 3.8 microns. The data is available calibrated & registered, as well as SNAP reduced.
These data are present under experiment-1/sequences/
.
For experiment 1, we use the SNAP pipeline to inject the simulated companion on a known orbit. Note: the epochs we simulate do not match the dates the real data was captured on. Conisder the simulated data to be captured at these epochs in MJD:
58849 + 0
58849 + 365.25/2
58849 + 2*365.25/2
58849 + 3*365.25/2
58849 + 4*365.25/2
That is, in 6 month increments.
The simulated companion is injected in "backwards" mode such that it is not visible in the standard reduction, and only appears in the "backwards" reduction. The noise statistics are identical between these two modes.
The files are as follows:
data/SNAP.stacked.fits.gz
data/SNAP.prepared.fits.gz
data/SNAP.prepared_opp.fits.gz
data
subdirectory.
Other files you may find interesting:SNAP.toml
: instructions to the SNAP pipeline data/SNAP.median.fits
: simple median stack of North-up imagesdata/SNAP.median.psf.fits
: simple median stack of star PSFdata/SNAP.median.sub.fits
: simple stack of subtractions of median from each imagedata/SNAP.psf.fits
: planet PSF templatedata/SNAP.subtracted.fits
: individual "batches" of frames processed with SNAP. They are groups of 10 frames processed jointly to reduce correlated noise.Note: currently the headers are missing from the prepared files, I will aim to rectify this shortly.
If you would like a bigger challenge, you can try to find the planet in the *.median.sub.fits
HDU 1 images! These have only the most basic of ADI processing applied so the SNR is lower.