Accessing Kepler and CoRoT data
Based on materials developed by Peter Plavchan, NExScI/IPAC
First, a necessary aside on file formats. Modern time-series data sets are served in one of two formats:
- Binary FITS Tables
- Standard data format
- Compact file size
- Standard headers (FITS keywords)
- Supports inclusion of ancillary information
- Difficult to work with other visualization and manipulation tools -- e.g., how to plot, load into Excel?
- Fixed precision in data values – float or double – vs. formatting specified in somewhat obscure keywords
- 64-bit vs. 32-bit wonkiness from heritage tools (e.g. long doubles) (e.g., some tools assume 64-bit numbers and some assume 32-bit)
- Real life example: Original Kepler public data release introduced machine precision errors into time and position values larger than the other noise sources. (e.g., the real number should have been, say, 6.0045123056 and what got written to the file was shorter, say, 6.00451 -- that truncation introduced errors into the data that were much larger than any other noise source).
- (relatively) Easy to apply multiple tools -- IDL, Excel, mathematica, other plotting utilities can easily grab these files
- Precision tailored to data set -- if you want 6.0045123056 then you write 6.0045123056.
- Easier to directly investigate data values with simple file viewers
- Lack of standardized formatting -- e.g. tab vs. space vs. comma vs. fixed width, etc.
- Lack of standardized headers -- e.g. IPAC ASCII keywords and formatting
- Larger file sizes
Kepler stares at one field continuously. It produces light curves at a specific cadence for selected objects in its field of view.
MAST is the official archive for Kepler. MAST=Multimission Archive at STScI, http://archive.stsci.edu This serves FITS light curves.
The NASA Exoplanet Archive also serves Kepler data. http://exoplanetarchive.ipac.caltech.edu This serves FITS and ASCII light curves. We also have a "value-added" statistics and periodogram tool (e.g., this is something you can't get from MAST)
e.g., the interactive data search, the target search, the public light curves, and the data release notes.
You can download all the public light curves at once in one GIGANTIC tarfile -- like 11 GB.
NASA Exoplanet Archive
Unsurprisingly (because we developed it), we like this option better. :) We think it's easier to understand what is going on. Note that these screenshots are from an earlier version of the interface to the archive.
Of course, for Kepler data, hit the Kepler button.
The NASA Exoplanet Archive also allows you to compute a periodogram from the light curve. Here is an example of what happens when you do that. The orange boxes in this figure indicate how to download the results of the periodogram analysis:
CoRoT stares at a field for a long periods of time (up to 6 months), and then picks another field to observe.
CoRoT has two channels -- Asteroseismology (~10 targets/run) & Exoplanet (~10k targets/run) Exoplanet light curves have two categories -- CHRomatic & MONochromatic, in separate binary FITS table files Asteroseismology light curves come in three flavors --RAW, HELiocentric, & HELiocentric with REGularized time-sampling, all in one binary FITS file with 3 table extensions.
The NASA Exoplanet Archive is the official US portal for CoRoT data -- http://exoplanetarchive.ipac.caltech.edu -- It serves FITS light curves (ASCII coming soon).
IAS Data and Operations Center also serves the data, but in binary FITS table format -- http://idoc-corotn2-public.ias.u-psud.fr/index.jsp
Finally, there is also the CoRoT archive at LAEFF (in Spain); they serve binary FITS tables & ASCII format -- http://sdc.laeff.inta.es/corotfa/jsp/frontpage.jsp
NASA Exoplanet Archive
Here is the search page at IAS:
You can download whole runs, but note that (a) this is a European site, so "3,3 GB" means "3.3 GB" (note use of commas rather than decimal points), and (b) some of these files are quite large, like 18 GB.
Note: you can reasonably frequently encounter database errors when using this site. It's not you, it's them...