DR4 Release Notes - 2019/01/16 - DPC Updated: 2019/01/18 DPC Files in the DR4 distribution: ------------------------------- DR4_Release_Notes.txt (6 KB) - this file Field_Summary_Table.dat (266 KB) - summary of key quantities for each of the 3,234 DR4 fields, as one line per field Observing_Logs.zip (1.22 GB) - hand-written logs for each night of GPIPS observations, organized by observing run (YYYYMM) and then night (YYYYMMNN) POLCATs.zip (0.92 GB) - Zip (compressed) file containing 3,234 files of polarization information, one file for each of the fields. PHOTCATs.zip (0.91 GB) - same, for photometric information, one file for each of the fields. Images (14.0 GB) - Deep, summed image for each field, in FITS format. Typically 4.5 MB per image. The images have been grouped and compressed (zipped) into groups of 500 images each (typically 1.7 GB each). Plots_in_PS.zip (0.87 GB) - Zip file containing plots of each field, in postscript format, with overlaid polarization vectors or circles for upper limits. Note that these plots represent one interpretation of how to render the polarization information and are not necessarily the best approach. The vectors encode the polarization percentage via their lengths and their orientations indicate the equatorial polarization position angle. Vectors were drawn for those stars measured to have p/sp >= 2.5 and p < 12.0%. The circles indicate `significant upper limits' defined to be for those stars measured to have p/sp < 1.0 and sp < 0.33%. Plots_in_PDF.zip (0.86 GB) - Zip file of the previously described postscript plots, converted into PDF form, and grouped into sets of 100 plots, as noted in the file titles. Images_0001_to_0500.zip (1.7 GB) - Zip file containing the coadded (deep) FITS images for the first 500 GPIPS fields Images_0501_to_1000.zip - same, for next 500 Images_1001_to_1500.zip - same, next 500 Images_1501_to_2000.zip - ditto Images_2001_to_2500.zip - ditto Images_2501_to_3000.zip - ditto Images_3001_to_3237.zip - final 237 FITS images Absent: -------- For DR4, we have elected to not include DS9 region files. These can be computed from the POLCATs for a variety of purposes, and so best left to the user. Additional Notes and Explanations: ------------------------------------- 1. Three fields (GP0302, GP1218, and GP2698) failed quality control checks and their data are not included in the DR4 data directories. In the Summary Table, they are identified as having zero values for all parameters. These fields are planned for reobservation later in 2019 and will be included in the DR4.1 release. 2. An additional 18 fields have some serious, but less fatal issues that would benefit from reobservation and will likely be updated for the DR4.1 release. These include GP0224, 1415, 1517, 2182, 2215, 2219, 2247, 2255, 2350, 2443, 2494, 2526, 2538, 2695, 2806, 2858, 2894, and 2912. 3. The Observing_Logs.zip file contains JPEG scans of all of the hand-written logs created during each observing night. These may or may not be informational for any particular application of GPIPS data products and may be ignored if desired. 4. The Field_Summary_Table.dat file contains the final quality control check values for each of the 3237 GPIPS fields observed. 4.A. The L_c and B_c columns report the actual image field centers in galactic coordinates. 4.B. The Cen_Off values give the offset of these centers from the desired GPIPS position grid, in units of arcseconds. For the 3237-21=3216 fields that will not be reobserved, the mean Cen_Off is 12.7", with an rms of 8.3". 4.C. The Night designation and first image number may be matched to the hand-written logs to ascertain observing conditions or problems. 4.D. The FWHM values are in arcsec and have been approximately corrected for the detector pixel scale as FWHM_Cor = sqrt(FWHM_Raw^2 - 0.6^2). For the 3216 fields, the mean FWHM_cor is 1.46" with an rms of 0.25". 4.E. The SecAmp field reports the secular amplitude variation removed in Step 6 of the MSP-PPOL processing. High SecAmp values correlate with poor observing conditions. Most observations having SecAmp > 3% were redone, though a couple remain in the set of 21 fields to be reobserved. The SecAmp units are percentage of sky value and exhibit a mean of 0.45% with rms of 0.55%. 4.F. Nuse0 is the number of images rejected from each group, generally for poorly formed PSFs (wind shake, usually). The mean is 0.48 and rms is 0.98, so these are generally unimportant. 4.G. Nok is the number of good images making up each observation group. Older observations generally used a 7 position by 16 HWP image scheme, while newer ones skipped the central observation to have 6 by 16 = 96 images. Some observations were restarted, and so had somewhat more images. The mean Nok is 99 with an rms of about 8, which really reflects the bifurcation into observations mainly of 102 and 96 members. 4.H. The SN field is the remaining sky noise once the secular variation and the field polarization signals are removed. Fields with high SN values were flagged and reobserved for DR4. The mean SN is 0.53% of the sky mean brightness, with an rms of 0.15%. 4.I. The EB score relates to the degree to which the polarization vectors are parallel (high EB values) or tangential/spiral (low to negative EB values). A systematic problem led to strong spiral patterns which were not sustained upon reobservation, so low/negative EB values fields were reobserved. Two of the 21 reobserve fields (GP 2215 and 2806) sill contain spiral patterns. The mean EB value was about 6 with an RMS of 2, indicating nearly all fields show a predominance of parallel polarization patterns. 4.J. The PF value is the 'Problem Flag' formed by thresholding and weighting departures from acceptable values in the previous categories. PF values of zero obtain if the center offset is under 45", the FWHM is under 2.25", the SecAmp is 2% or less, Nok is 89 or higher, SN is under 2% and EB is 2 or greater. The mean PF is 0.07 with an rms of 0.36, hence nearly all fields pass the quality control tests.