Description of the EPOCh Stellar Transit Observations and Photometry By Drake Deming August 23, 2009 Revised December 20, 2010, to include summary of calibration steps and error levels. EPOCh Stellar Transit Observations ---------------------------------- The EPOCh stellar transit observations utilized a broad optical band, extending from 350 to 1000 nanometers (clear filter number 6), of the HRIV CCD. Each resulting image data file represents a single 50-second exposure, using a subarray mode. The units of the archived raw images are raw data counts while units of the archived calibrated images are Watts per square meter per steradian per micron. (Both raw and calibrated data files are archived at Multi-mission Archive at STScI (MAST) and the NASA Planetary Data System (PDS).) In most instances, the subarray size is 128x128 pixels. But sometimes during transit and/or secondary eclipse, a 256x256 subarray size was used. The larger subarray size ensured that pointing jitter did not cause the star to fall beyond the edges of the subarray. The stellar images are defocused (a permanent on-orbit property of the telescope) to about 10 pixels (4 arcsec) FWHM. The calibrated HRIV CCD images have been bias-subtracted with removal of electronic crosstalk and transfer smear and nominally flat-fielded, using flat-field calibration exposures of an integrating sphere made on the ground before launch; these radiance images are archived at the PDS (http://pds.nasa.gov) under data set with an ID of DIF-X-HRIV-3-EPOXI- EXOPLANETS-V1.0. Raw images are archived at the PDS with a data set ID of DIF-X- HRIV-2-EPOXI-EXOPLANETS-V1.0. However, the CCD detector has changed in space, and the ground flat-field calibration is not adequate for precision stellar photometry. The EPOCh team used an independent bootstrap procedure, fitting aperture photometry to a 2-D spline to define corrections to the flat-field calibration (that is, the archived radiance images). Table 1 lists the steps that the team applied to the calibration and photometry of each EPOCh stellar target. The resulting photometry data are archived at the PDS under data set with an ID of DIF-X-HRIV-5-EPOXI-EXOPLANETS-PHOT-V1.0. +----------------------------+------------------------+----------------------+ |Calibration Step |Reference |Comment | |----------------------------|------------------------|----------------------| |bias and dark current |Klassen et al. 2008 |pipeline at Cornell | |subtraction | | | |----------------------------|------------------------|----------------------| |preliminary flat-fielding |Klassen et al. 2008 |pipeline at Cornell | |----------------------------|------------------------|----------------------| |corrections for central rows|Ballard et al. 2010 | | |and columns, subarray size, |Christiansen et al. 2010| | |and quadrant-dependent bias | | | |----------------------------|------------------------|----------------------| |construct PSF using drizzle |Barry et al. 2010 |each target star used | | | |to make its own PSF | |----------------------------|------------------------|----------------------| |secondary flat-fielding |Christiansen et al. 2010| | |using stimulator lamp | | | |data | | | |----------------------------|------------------------|----------------------| |fit PSF to each stellar |Ballard et al. 2010 |determines stellar X,Y| |image | |position | |----------------------------|------------------------|----------------------| |aperture photometry, |Ballard et al. 2010 |circular aperture | |10-pixel radius | |centered on star | |----------------------------|------------------------|----------------------| |correct aperture |Ballard et al. 2010 |spline surface based | |photometry using 2-D | |on random subset of | |spline surface | |points | |----------------------------|------------------------|----------------------| |lower right quadrant |Ballard et al. 2011 |only needed for | |correction | |contingent TrES-2 and | | | |HAT-P-4 data | |----------------------------|------------------------|----------------------| |point-by-point correction |Christiansen et al. 2010|for points within a | | | |3-hour window centered| | | |transit | +----------------------------+------------------------+----------------------+ Table 1. Listing of the steps applied to the calibration and photometry of each EPOCh stellar target. Error levels for EPOCh stellar photometry are given in the following tables (2-7), one table for each transiting planet system. The EPOCh CCD detector has a spatially non- uniform response. Coupled with spacecraft pointing jitter, this leads to noise levels that are a function of the time scale, i.e. the number of points that are averaged. The ideal photon-limited case would produce standard deviations for binned data that decrease as the inverse square root of the total number of binned points (equivalently, as the square root of the total time interval). The tables give the number of averaged points (NP), the time interval (Ntime, in minutes) and the standard deviation for that time interval in units of the stellar flux. Each point that is binned represents 50-seconds of exposure time, plus a few seconds of overhead. Overhead is included in the Ntime values. HAT-P-4 NP Ntime (minutes) Standard Deviation --- --------------- ----------------- 1 0.843667 0.0011568513 2 1.68733 0.00090268208 4 3.37467 0.00073773344 8 6.74933 0.00055135903 16 13.4987 0.00042166669 32 26.9973 0.00034057043 64 53.9947 0.00026740742 128 107.989 0.00019248507 256 215.979 0.00012183909 512 431.957 5.9978127e-05 Table 2. Photometry error levels for HAT-P-4. HAT-P-7 NP Ntime (minutes) Standard Deviation --- --------------- ----------------- 1 0.843667 0.00075210299 2 1.68733 0.00064274314 4 3.37467 0.00054440071 8 6.74933 0.00050449965 16 13.4987 0.00043062094 32 26.9973 0.00036409640 64 53.9947 0.00029516709 128 107.989 0.00020787818 256 215.979 0.00017629711 512 431.957 0.00014063646 Table 3. Photometry error levels for HAT-P-7. GJ 436 NP Ntime (minutes) Standard Deviation --- --------------- ----------------- 1 0.843667 0.00050875102 2 1.68733 0.00043097194 4 3.37467 0.00036651007 8 6.74933 0.00032530082 16 13.4987 0.00022578942 32 26.9973 0.00018538466 64 53.9947 0.00012555730 128 107.989 0.00011223905 256 215.979 8.4118445e-05 512 431.957 3.1448330e-05 Table 4. Photometry error levels for GJ 436. TrES-2 NP Ntime (minutes) Standard Deviation --- --------------- ----------------- 1 0.843667 0.0015681651 2 1.68733 0.0012356165 4 3.37467 0.00093257416 8 6.74933 0.00070808700 16 13.4987 0.00055674871 32 26.9973 0.00044851966 64 53.9947 0.00034450906 128 107.989 0.00031302523 256 215.979 0.00027201473 Table 5. Photometry error levels for TrES-2. TrES-3 NP Ntime (minutes) Standard Deviation --- --------------- ----------------- 1 0.843667 0.0025323839 2 1.68733 0.0019495520 4 3.37467 0.0014570355 8 6.74933 0.0010532849 16 13.4987 0.00079664454 32 26.9973 0.00062819093 64 53.9947 0.00048610280 128 107.989 0.00041822722 256 215.979 0.00027489904 512 431.957 0.00019008991 Table 6. Photometry error levels for TrES-3. WASP-3 NP Ntime (minutes) Standard Deviation --- --------------- ----------------- 1 0.843667 0.00083660282 2 1.68733 0.00073821470 4 3.37467 0.00064862869 8 6.74933 0.00054654607 16 13.4987 0.00045596197 32 26.9973 0.00037014380 64 53.9947 0.00030564962 128 107.989 0.00020876195 256 215.979 0.00016026215 512 431.957 0.00010886080 Table 7. Photometry error levels for WASP-3. Finally it is important to note that the spacecraft clock is affected by a systematic drift relative to ground clocks, due to the changing thermal environment of the spacecraft. The EPOCh team has calibrated the spacecraft clock versus ground clocks, and the corrected times were computed by the EPOCh team for each transiting system (giving Julian Date, and Barycentric Julian Date, corresponding to each file name). However, the computation was eventually implemented in the mission's automated data pipeline, and the EPOCh team verified the resulting values were consistent with theirs. Thus the Barycentric Julian Date (when light from the target reaches the solar system barycenter) is provided by the KPKSSBJT keyword in the FITS headers of raw and calibrated data files archived at MAST and PDS. References ---------- Ballard, S., Christiansen, J. L., et al. 2010, ApJ, 716, 1047. Ballard, S., Christiansen, J. L., et al. 2011, submitted to ApJ. Barry, R. K., Lindler, D., et al. 2010, SPIE 7731, 77313D. Christiansen, J. L., Ballard, S., et al. 2010, in press for ApJ, astro-ph/1011.2229. Klassen, K., A'Hearn, M. F., et al. 2008, Rev. Sci. Instr., 79, 77.