spacer link to MAST page spacer logo image spacer

EPIC Variability Extraction and Removal for Exoplanet Science Targets ("EVEREST")

Luger et al. 2017, AJ, submitted
See also: Luger et al. 2016, AJ, 152, 100

Source code available on GitHub. Documentation is available at MAST (applicable to data archived here) and at Washington (updated by the team as their software evolves).


Description of Data Products

Data Access

Download README

Latest Update: Version 2.0 files have replaced version 1.0. The previous version 1.0 files will be made temporarily available here for those who need to compare with work started using the older version, but will be taken offline in mid-2017.


An example of the uncorrected fluxes from K2 (black) and the corrected EVEREST version (blue).
EVEREST is an open-source pipeline for removing instrumental systematics from K2 light curves, using a combination of pixel-level decorrelations to remove spacecraft pointing error and Gaussian processes to capture astrophysical variability. Corrected light curves have precision comparable to the original Kepler mission for targets brighter than Kp = 13, and within a factor of 2 for fainter targets. Transit injection and recovery tests have been performed to validate the pipeline, and comparisons with other K2 detrended HLSPs is performed (consult the Luger et al. 2016 paper for further details).

NOTE: Since EVEREST performs least-squares fits to reduce the noise in K2 light curves, astrophysical features such as transits and eclipses can sometimes be slightly shallower in the de-trended dataset. In order to prevent this, EVEREST automatically masks outliers prior to computing the fits. However, low signal-to-noise transits are likely to be missed in this step. The EVEREST software (available on Github, see link at top of page) makes it easy to generate light curves with unbiased transit depths if the locations of the transits are known.

Description of Data Products

Each target is organized into a subdirectory structure, based on its EPIC ID, with the following format:


  • c<##> is the 2-digit, zero-padded Campaign number the target was observed in.
  • <####00000> are the first FOUR digits of the target's EPIC ID
  • <#####> are the last FIVE digits of the target's EPIC ID
For example, the star EPIC 202059070 was observed in Campaign 0. It's subdirectory path is therefore:

Several files are available for each target. Each file has the same prefix ("hlsp_everest_k2_llc_{EPICID}-c##_kepler_v2.0"), followed by an extension. Short cadence light curves are designated by "sc". The extensions and types for these files are:

  1. _lc.fits | _sc.fits = The FITS file has 5 extensions: the primary HDU with just a header, the data extension with the de-trended light curve, an extension containing the pixel-level light curve, an extension containing the aperture mask used, an extension that stores images of the full target postage stamp at three points in the light curve, and an extension that has a higher resolution image of hte target from the Palomar Observatory Sky Survey. The data extension (first extension) consists of the following columns:
    • TIME - the original timestamp (BJD - 2454833.0)
    • CADN - the original cadence number
    • FLUX - the de-trended flux, (e-/s)
    • FCOR - the CBV-corrected de-trended flux (e-/s)
    • FRAW - the original, raw SAP flux
    • FRAW_ERR - the observing errors on the raw flux
    • QUALITY - a 64-bit integer array of quality flags for each cadence
    • BKG - if present, the background flux subtracted from each cadence
  2. _dvs.pdf | _dvs_sc.pdf = The data validation summary (DVS) figures consist of PDF documents with two pages for each de-trended target. The first page shows three panels: the final, PLD-detrended and CBV-corrected flux (top), the uncorrected detrended flux (center), and the raw SAP flux (bottom). In the center plot, the red curve is the CBV fit. Miscellaneous information is displayed at the top (the K2 campaign number, the Kepler magnitude, the photometric precision, etc.). The second page shows the detailed PLD de-trending summary, showing the progression of the light curve as higher PLD orders are included in the de-trending. Snapshots of the postage stamp are shown at the top right, with a high resolution POSS image of the target. Below these, the cross-validation curves for each PLD order are plotted. More information on these summary files can be found in the EVEREST documentation.
  3. _dvs-0.png, _dvs-1.png | _dvs_sc-0.png, _dvs_sc-1.png = PNG versions of page 1 and 2 of the respective "dvs.pdf" file.

Data Access

MAST Web Form
MAST Discovery Portal
Bundles (.tar.gz for each
c## / #### / directory)
cURL or wget scripts
per target