Getting Started with Kepler

1
Getting Started with Kepler

• Kepler Science Guest Observer Offices
• Steve Bryson
• Douglas Caldwell
• Jessie Christiansen
• Michael Fanelli
• Michael Haas
• Jon Jenkins
• Karen Kinemuchi
• Jeffrey Kolodziejczak
• Pavel Machalek
• Fergal Mullally
• Jason Rowe
• Martin Still
• Susan Thompson
• Jeffrey Van Cleve

January 11, 2011
e-mail kepler-scienceoffice_at_lists.nasa.gov
keplergo_at_mail.arc.nasa.gov
2
Agenda

• What Kepler provides
• Spacecraft Idiosyncrasies Doug Caldwell
• Pixels to Photometry
• Light Curve Systematics Detrending
• How you can be involved
• Opportunities and Support
• How to Find Interesting Targets
• Archival Products
• User Documentation

3
Kepler Data from Stars to Ground
Module 3, consisting of 4 channels, died in Jan
2010.
Kepler is a Schmidt telescope with a focal plane
of 42 back-thinned CCDs each with two read-out
amplifiers for a total of 84 detector channels.
The photometer stares continuously for 3
months, then rotates 90º about the boresight. The
CCD layout is four-fold symmetric, except for the
center module, so rows and columns maintain their
orientation on the sky.

• Kepler is continuously monitoring gt150,000 stars
  in a 110 sq. degree field in Cygnus/Lyra. The
  mission is planned for 3.5 years. Three main
  types of data are available
• Long cadence 30-minute samples of gt150K stars
  from 5 18th mag
• Short cadence 1-minute samples of 512 stars
• Full-Frame Images monthly snap shots of all
  active pixels in the focal plane

4
Kepler Data from Stars to Ground
Image motion differential velocity aberration
introduces target star motions up to 0.5 pixel
at the edges of the focal plane (top). Pointing
drift has been greatly reduced since Q3 (bottom)
1) Stars to CCDs Seasonal focus variations
change the PSF width up to 10 on some channels
DVA motion over focal plane
Scattered light Optical ghosts affect only
1-3 of focal plane.
Pointing drift attitude error
Attitude error (mpix) 10 30 50 70
5
Kepler Data from Stars to Ground
2) CCDs to memory Electronic ghosting (video
crosstalk) occurs within modules. Peak signal
level is about 1/200 of source signal
Undershoot/overshoot occurs during readout with a
median level of 0.3. Undershoot is corrected in
SOC calibration
Fine guidance sensor clocking crosstalk occurs on
all channels, affecting 15 of targets. FGS
xtalk is corrected in calibration. Moiré pattern
artifacts affect 10-20 of the FOV at a level
significant for Earth-size transit detections
(0.02 DN/pixel/read)
6
Kepler Data from the Stars to the Ground
3) Memory to storage to ground Pixel values are
requantized based on intrinsic noise level to
reduce range from 23 bits to 16 bits. Requantized
step size is designed such that quantization
noise is lt25 of intrinsic pixel noise (3 noise
increase when RSSd)
Subsequent compression is lossless baseline
image difference followed by Huffman coding and
reduces range to 5 bits/pixel/long cadence. DMC
decompresses and un-requantizes
7
Agenda

• What Kepler provides
• Spacecraft Idiosyncrasies
• Pixels to Photometry Steve Bryson
• Light Curve Systematics Detrending
• How you can be involved
• Opportunities and Support
• How to Find Interesting Targets
• Archival Products
• User Documentation

8
Pixel Level Data
Module 17 Output 2
Zoomed Image near HAT-P-7b
HAT-P-7b LC pixels
6.6 x 6.6 millideg 28 pixels collected Black no
data
0.09 x 0.09 degrees 80 x 80 pixels 6400 pixels
total
Scaled to show faint detail
1.13 (h) x 1.22 (w) degrees
9
Pipeline Pixels to Planets
CAL Pixel Level Calibrations
TAD Select Pixels
Raw Data
Pixel Specification
Calibrated Pixels
PA Photometric Analysis Sums Pixels Measures
Centroids
PDC Pre-search Data Conditioning Removes
Systematic Errors
Raw Light Curves Centroids
Corrected Light Curves
TCEs Threshold Crossing Events
TPS Transiting Planet Search
Diagnostic Metrics
DV Data Validation
10
TAD Find the Optimal Pixels that Maximize SNR

• TAD Pixels are selected that maximize the SNR of
  the calibrated sum
• Use the measured PSF to create synthetic images
  with and without the target, so we know the
  signal and the noise of each pixel

11
TAD Get the Optimal Pixels and More

• TAD Pixels are selected that maximize the SNR of
  the calibrated sum
• Optimal aperture embedded in a halo for margin
• Optimal aperture halo define the pixels
  downlinked for each target

12
Pipeline Pixels to Planets
CAL Pixel Level Calibrations
TAD Select Pixels
Raw Data
Pixel Specification
Calibrated Pixels
PA Photometric Analysis Sums Pixels Measures
Centroids
PDC Pre-search Data Conditioning Removes
Systematic Errors
Raw Light Curves Centroids
Corrected Light Curves
TCEs Threshold Crossing Events
TPS Transiting Planet Search
Diagnostic Metrics
DV Data Validation
13
CAL Pixel Level Calibrations
LDE Undershoot
FGS Clocking Crosstalk
Identify CR in Black Black Level Correction
2D Black Correction
Gain Nonlinearity Correction
LDE Undershoot Correction
Raw Pixels
ID CR Remove from Smear
Smear Dark Current Correction
Flat Field Correction
Calibrated Pixels Uncertainties
14
CAL Pixel Level Calibrations
Raw FFI
Calibrated FFI
15
Pipeline Pixels to Planets
CAL Pixel Level Calibrations
TAD Select Pixels
Raw Data
Pixel Specification
Calibrated Pixels
PA Photometric Analysis Sums Pixels Measures
Centroids
PDC Pre-search Data Conditioning Removes
Systematic Errors
Raw Light Curves Centroids
Corrected Light Curves
TCEs Threshold Crossing Events
TPS Transiting Planet Search
Diagnostic Metrics
DV Data Validation
16
PA From Pixels to Photometry

• PA Create the flux light curve by summing pixels
  in the optimal aperture
• First remove background from the pixels
• Then remove cosmic rays from the pixels
• Then sum optimal pixels to create flux light
  curve
• Also compute pixel centroids at each observation
• Systematics in the data
• Focus variations due to thermal events
• Pointing drift early in the mission

30 days
16 months
17
PA Special Issues

• Sometimes conventional photometry has problems
• Saturated targets
• Saturation is very complex and poorly modeled
• So in early quarters some targets were poorly
  captured
• Galaxies misclassified as stars
• Solution do your own photometry via target pixel
  files

Example of a galaxy misclassified as a bright star
TAD pixel selection
Custom pixel selection
18
Agenda

• What Kepler provides
• Spacecraft Idiosyncrasies
• Pixels to Photometry
• Light Curve Systematics Detrending Jeff Van
  Cleve
• How you can be involved
• Opportunities and Support
• How to Find Interesting Targets
• Archival Products
• User Documentation

19
Pipeline Pixels to Planets
CAL Pixel Level Calibrations
TAD Select Pixels
Raw Data
Pixel Specification
Calibrated Pixels
PA Photometric Analysis Sums Pixels Measures
Centroids
PDC Pre-search Data Conditioning Removes
Systematic Errors
Raw Light Curves Centroids
Corrected Light Curves
TCEs Threshold Crossing Events
TPS Transiting Planet Search
Diagnostic Metrics
DV Data Validation
20
How PDC Works

• PDC is designed to remove systematic errors by
  removing correlations (cotrending) between light
  curves and known system state variables such as
• Electronics board temperatures
• Pointing and local image motion
• Thermometers on or near optics (not currently
  used, but available)
• Method used is Least Squares fit to the principal
  components of this set of time series
• Simple filtering (detrending) cannot remove
  systematic noise on the same time scale as a
  transit without removing most of the transit,
  while cotrending can
• PDC also removes light curve discontinuities
  caused by persistent damage of individual pixels
  by cosmic rays

21
What PDC Does Well

• PDC works well on gt70 of stars gt90 of the time
  for its primary purpose cleaning up light
  curves for transit detections
• Quarters with several flight system anomalies
  (Safe Mode, Loss of Fine Point, pointing tweaks)
  are difficult because of the number and diversity
  of discontinuities

22
PDC Emphasizes Transits While Sometimes
Distorting Astrophysical Signals

• A Least-squares (LS) fitting of systematic errors
  (cotrending) to an incomplete model can minimize
  the bulk RMS of a light curve by transferring
  signal power from real astrophysical signals into
  false high-frequency noise
• Users will need to be cautious when their
  phenomena of interest are much shorter (lt1 h) or
  much longer (gt5 d) than a transit, or have
  complex light curves with multiple extrema on
  transit time scales (such as eclipsing and
  contact binaries)

23
PDC Sometimes Adds Noise
PDC-induced high frequency noise is more
noticeable for bright stars, where it can
dominate shot noise
24
Present Expedients and Future Progress

• For diverse science, how you process the data can
  depend on what you are looking for
• In the near term, users are invited to use the
  engineering data and intermediate Pipeline
  products provided as the Data Release Notes
  Supplement to do their own cotrending of
  uncorrected light curves
• In the middle term, users may also wish to
  cotrend data using principal components derived
  from an ensemble of reference light curves for
  each channel, after a Quarter goes public
• In the long term, we are developing the Maximum A
  Posterior (MAP) method described in Mondays talk
  (103.02) and poster (140.08) to provide PDC with
  constraints on the magnitudes and signs of the
  fit coefficients to prevent overfitting and
  high-frequency noise
• The legacy solution would be to make an off-line
  Pipeline available to users, pending availability
  of resources

25
The Joy of MAP An Example

• Blue line is mean-removed, normalized flux
• Green line is a robust fit to the blue line,
  showing signal artifact and high-frequency
• noise in the fit
• Red line is the MAP fit to the blue line,
  showing gradual trend to be removed
• Both robust fit MAP use light curve ensemble
  as cotrending basis in this example

26
Agenda

• What Kepler provides
• Spacecraft Idiosyncrasies
• Pixels to Photometry
• Light Curve Systematics Detrending
• How you can be involved
• Opportunities and Support Martin Still Mike
  Fanelli
• How to Find Interesting Targets
• Archival Products
• User Documentation

27

• GUEST OBSERVER
• PROGRAM
• Annual Program
• PURPOSE provide the whole community with
  competitive access to Kepler through a
  peer-reviewed science competition
• 3,000 long cadence targets available every
  quarter
• 25 short cadence targets are available every
  month
• Successful US proposals are funded by NASA grants
• URL keplergo.arc.nasa.gov
• E-mail keplergo_at_mail.arc.nasa.gov

• Martins Slides go here

28

• GUEST OBSERVER PROGRAM
• Directors Discretionary Targets
• PURPOSE provide rapid response to community
  demand for Kepler targets
• Targets of Opportunity
• Pilot studies for the main GO program
• Follow-up of old Kepler targets
• Reinstatement of dropped targets
• 100 targets available every quarter
• Successful proposals are not funded by NASA
  grants
• URL keplergo.arc.nasa.gov
• E-mail keplergo_at_mail.arc.nasa.gov

• Martins Slides go here

29
KEPLER ASTEROSEISMOLOGY SCIENCE CONSORTIUM

• KASC is an unfunded consortium of (currently
  440) asteroseismologists.
• KASC receives 1700 targets per quarter to
  investigate stellar seismic activity
• Solar-like oscillations
• Pulsations in open clusters
• ß Cephei stars
• d Scuti stars
• roAp stars
• Cepheid variables
• B stars
• Red giants
• Pulsations in binary stars
• g Doradus stars
• Compact pulsators
• Miras and semi-regulars
• RR Lyrae stars
• KASC is an open consortium, membership is free
• apply online at astro.phys.au.dk/KASC/

30
PARTICIPATING SCIENTIST PROGRAM

• Participating Scientists serve as members of the
  Kepler Science Team and participate in Science
  Team activities, such as exoplanet data
  processing and analysis, transit candidate
  follow-up and characterization, and publication
• Exoplanet statistics and theory
• Sub-Neptune size planets
• Multi-planet systems
• Planet atmospheres
• Asteroseismology
• Stellar activity
• Eclipsing and interacting binaries
• Cluster astrophysics
• The deadline for cycle 2 PSP proposals is Feb 11,
  2011. Proposal instructions and element details
  are provided in the NASA Research Announcement
• URL keplerscience.arc.nasa.gov/PSP.shtml
• E-mail douglas.m.hudgins_at_nasa.gov

31
ASTROPHYSICS DATA ANALYSIS PROGRAM (ADAP)

• From Feb 1, 2011 the Kepler archive at MAST
  (http//archive.stsci.edu/kepler) will have
  available for the whole community 165,000 light
  curves with 130 days of near-continuous
  monitoring (Q0, Q1 Q2)
• Analysis of archived Kepler data can be funded
  through NASAs Astrophysical Data Analysis
  Program, designed to support continued science
  using NASA mission data sets. Possible topics
  include
• Exoplanets
• Asteroseismology
• Stellar activity and evolution
• Binary stars
• Stellar and extragalactic accretion
• The expected deadline for 2011 ADAP proposals is
  May 14, 2011. Proposal instructions and element
  details for the 2011 ADAP cycle will be released
  in early February 2011. Details about the 2010
  ADAP program are provided in the NASA Research
  Announcement NNH10ZDA001N-ADAP on NSPIRES.
• URL http//nspires.nasaprs.com
• E-mail douglas.m.hudgins_at_nasa.gov

32
(No Transcript)
33
KEPLER COMMUNITY SERVICES

• Kepler Guest Observer Office
• URL http//keplergo.arc.nasa.gov
• E-mail keplergo_at_mail.arc.nasa.gov
• Kepler Science Office
• E-mail kepler-scienceoffice_at_lists.nasa.gov
• Kepler archive at MAST
• URL http//archive.stsci.edu/kepler
• E-mail archive_at_stsci.edu
• Kepler Users Panel
• URL http//nspires.nasaprs.com
• E-mail kepler-users-panel_at_lists.nasa.gov

34
Agenda

• What Kepler provides
• Spacecraft Idiosyncrasies
• Pixels to Photometry
• Light Curve Systematics Detrending
• How you can be involved
• Opportunities and Support
• How to Find Interesting Targets Mike Fanelli
• Archival Products
• User Documentation

35
Selecting Targets For Observation
The Kepler field-of-view, as seen in a full-frame
image taken during commissioning

• The challenge
• Classify catalog sources
• Identify appropriate targets for a wide range of
  astrophysical investigations

36
Creating a Source Catalog
? Need a comprehensive catalog describing
source positions, motions, brightness,
colors classifications ? Primary
classification source optical observing
program of the FOV carried out by the Kepler
Science Team ? Photometry obtained in Sloan
g,r,i,z broad-band filters D51
narrow-band filter (a surface gravity
diagnostic) ? Calibrated data compiled into the
Kepler Input Catalog ? 4.4 million
cataloged sources located within the
detector footprint
37
The Kepler Input Catalog
Contents of the KIC ? Unique source
identifications expressed as KIC (sometimes
KepID) ? Federated with 2MASS JHK photometry,
also uses USNO-B catalog ? Astrometry from
2MASS, USNO-B ? Star/galaxy separation using the
2MASS extended source catalog ? Estimation of
source contamination (crowding) at the Kepler
pixel scale ? Calibrated Kepler magnitudes (Kp)
flux as seen through the Kepler photometer ?
Source locations in the FOV, defined using focal
plane models ? Estimates of Teff, log_g (for
dwarf / giant separation), Fe/H, E(B-V) ?
Single epoch data NO information on source
variability KIC algorithms document
www.cfa.harvard.edu/kepler/kic/algorithms/algori
thms.ps
38
MAST Target Search
KIC search engine archive.stsci.edu/kepler/keple
r_fov/search.php input parameter
search criteria Example All
sources with 15.0 lt Kp lt
16.0 Teff lt 4000 K
Log g gt 4.0 multiple output
formats
39
Mining the Kepler Input Catalog
? Primary information source for target
identification ? Excellent search engine at
MAST ? Users can also download the KIC as a text
file from MAST, be aware of caveats as
described on the MAST page ? Users must carefully
check selected target lists artifacts exist in
the KIC ? Use the KIC overlay tool provided by
skyview.gsfc.nasa.gov Details on Guest
Observer webpage keplergo.arc.nasa.gov under
Tools
Diffraction spikes flagged as KIC entries 2x2
arcminute extracts from the DSS2 red images, with
KIC entries overlain using the tool noted
above Multiple KIC entries in galaxy core
40
Expanding the Target Knowledge Base
? The KIC is optimized to identify cool (FGKM)
dwarf stars for transit searches ? Users wanting
to explore a broader range of science need
sources e.g., eclipsing binaries,
pulsators, hot stars, accretors, compact stars,
galaxies, active nuclei ? Kepler full-frame
images provide a user resource for new source
identification One FFI is obtained each
month all FFIs are available at MAST ? A number
of community efforts are underway to expand
knowledge of sources within the FOV
GALEX (UV) / Kepler Cross Match Catalog
developed by MAST UKIRT deep J-band
survey includes Kepler FOV
Courtesy Phil Lucas ROE, see keplergo.arc.nasa.gov
/Tools All Sky Automated Survey - Kepler
variable stars in the Kepler field
www.astrouw.edu.pl/asas/kepler/kepler.html
Variable Star Catalog derived using 8
commissioning FFIs see 201.06 Observing
Campaigns spring/summer 2011 U-band deep (to
21 mag) optical g,r,i, bands Ha
Other multiwavelength surveys visit
HEASARC, IPAC archives (NStED)
41
GALEX - Kepler Source Match
42
Agenda

• What Kepler provides
• Spacecraft Idiosyncrasies
• Pixels to Photometry
• Light Curve Systematics Detrending
• How you can be involved
• Opportunities and Support
• How to Find Interesting Targets
• Archival Products Susan Thompson
• User Documentation

43
Introduction to the Kepler Archive
Light Curve Files
Target Pixel Files
Full Frame Image
Data at MAST -- Multimission Archive at
STScI http//archive.stsci.edu/kepler
44
Introduction to the Kepler Archive

• Data Products at the MAST
• Light Curve Files (_llc and _slc)
• LC is a quarter long
• SC is a month long
• Calibrated Aperture photometry (PA)
• Detrended light curve (PDC)
• Centroids
• Target Pixel Files
• Full Frame Images

45
Introduction to the Kepler Archive

• Data Products at the MAST
• Light Curve Files (_llc and _slc)
• Target Pixel Files (_lpd-targ and _spd-targ) NEW!
• Contains pixel information for one target.
• RAW counts
• Calibrated Pixels
• Background Pixels
• Cosmic Rays
• Quality Flags
• Aperture used by PA
• Full Frame Images (_ffi-cal)

46
Target Pixel Files

• Contains the pixels for each cadence.
• Shows aperture used for photometry
• Barycentered time (BKJD)
• Flux in e-/sec

47
Introduction to the Kepler Archive

• Data Products at the MAST
• Light Curve Files (_llc and _slc)
• Target Pixel Files (_lpd-targ and _spd-targ)
• Full Frame Images (_ffi-cal)
• Full readout of the CCDs each month
• 84 extensions, one for each mod/out
• https//archive.stsci.edu/kepler/ffi/search.php

http//archive.stsci.edu/kepler
48
Getting Kepler Data

• FTP download of gzipped tar files
• Each contains subset of public data
• Dropped Targets for Quarters 0-3
• Public Data from Quarters 0-3
• other public data
• http//archive.stsci.edu/pub/kepler/lightcurves/ta
  rfiles/
• Anonymous ftp
• archive.stsci.edu
• cd /pub/kepler/lightcurves/tarfiles
• Data Search Page
• Good to search for individual or groups of data.

49
Kepler Data Search

• http//archive.stsci.edu/kepler/data_search/sea
  rch.php

50
(No Transcript)
51
Retrieving Data

• To obtain a username and password (or use
  anonymous for public data)
• http//archive.stsci.edu/registration/registration
  _form.html
• You must highlight target pixel files or FFI if
  that is what you wish to retrieve

52
Future Data Releases

• February 1, 2011 All Q2 exoplanet data will go
  public
• April 2011 Q3 GO targets will be public

You are here
  May-09 Jun-09 Jul-09 Aug-09 Sep-09 Oct-09 Nov-09 Dec-09 Jan-10 Feb-10 Mar-10 Apr-10 May-10 Jun-10 Jul-10 Aug-10 Sep-10 Oct-10 Nov-10 Dec-10 Jan-11 Feb-11 Mar-11 Apr-11 May-11 Jun-11 Jul-11 Aug-11 Sep-11 Oct-11 Nov-11 Dec-11 Jan-12 Feb-12 Mar-12 Apr-12 May-12 Jun-12 Jul-12 Aug-12 Sep-12 Oct-12 Nov-12 Dec-12 Jan-13 Feb-13 Mar-13 Apr-13 May-13 Jun-13 Jul-13 Aug-13 Sep-13 Oct-13 Nov-13 Dec-13 Jan-14 Feb-14 Mar-14 Apr-14  
Data Taken on Space-craft Q0,Q1 Q0,Q1 Q2 Q2 Q2 Q3 Q3 Q3 Q4 Q4 Q4 Q5 Q5 Q5 Q6 Q6 Q6 Q7 Q7 Q7 Q8 Q8 Q8 Q9 Q9 Q9 Q10 Q10 Q10 Q11 Q11 Q11 Q12 Q12 Q12 Q13 Q13 Q13 Q14 Q14 Q14 Q15 Q15                                    
Data at MAST, FFIs           Q0, Q1 Q0, Q1 Q0, Q1 Q2 Q2 Q2 Q3 Q3 Q3 Q4 Q4 Q4 Q5 Q5 Q5 Q6 Q6 Q6 Q7 Q7 Q7 Q8 Q8 Q8 Q9 Q9 Q9 Q10 Q10 Q10 Q11 Q11 Q11 Q12 Q12 Q12 Q13 Q13 Q13 Q14 Q14 Q14 Q15 Q15                        
GO Data Release                                               Q2, Q3 Q2, Q3 Q2, Q3 Q4 Q4 Q4 Q5 Q5 Q5       Q6, Q7 Q6, Q7 Q6, Q7 Q8 Q8 Q8 Q9 Q9 Q9 Q10 Q10 Q10 Q11 Q11 Q11 Q12 Q12 Q12 Q13 Q13 Q13 Q14 Q14 Q14 Q15 Q15
Planet Data Release                           Q0, Q1 Q0, Q1 Q0, Q1 Q0, Q1 Q0, Q1 Q0, Q1 Q0, Q1 Q0, Q1 Q2 Q2 Q2 Q2 Q2 Q2 Q2 Q2 Q2 Q2 Q2 Q2 Q2 Q2 Q2 Q2 Q3, Q4 Q3, Q4 Q3, Q4 Q3, Q4 Q3, Q4 Q3, Q4 Q3, Q4 Q3, Q4 Q3, Q4 Q3, Q4 Q3, Q4 Q3, Q4 Q5, Q6 Q5, Q6 Q5, Q6 Q5, Q6 Q7 - Q15 Q7 - Q15 Q7 - Q15          
53
Agenda

• What Kepler provides
• Spacecraft Idiosyncrasies
• Pixels to Photometry
• Light Curve Systematics Detrending
• How you can be involved
• Opportunities and Support
• How to Find Interesting Targets
• Archival Products
• User Documentation Jessie Christiansen

54
User Documentation

• The following documentation can be obtained from
  the Kepler MAST Documents page
  http//archive.stsci.edu/kepler/documents.html
• Kepler Instrument Handbook photons to pixels
• Description of hardware design, performance,
  operational constraints
• Updated bi-annually
• Kepler Data Processing Handbook pixels to
  photometry
• Description of science processing pipeline,
  including theoretical basis of the algorithms
  used
• Updated with each new release of the SOC Pipeline
• Kepler Archive Manual photometry to you
• Description of data available through MAST,
  including file formats and availability
• Updated with each new release of the SOC Pipeline

55
User Documentation

• The following documentation can be obtained from
  the Kepler MAST Data Release page
  http//archive.stsci.edu/kepler/data_release.html
• Kepler Data Characteristics Handbook phenomena
  in photometry
• Description of all instrumental/systematic
  phenomena identified in the data
• Updated with identification of new phenomena
• Kepler Data Release Notes brief newsletter
  accompanying each data release
• Dynamic set of tables and figures documenting
  phenomena identified during each processing of
  each quarter of data
• Supplement tar ball of files for use in user
  analysis and cotrending of the data, including
  ancillary engineering data