Planck 2015 results. VIII. High Frequency Instrument data processing: Calibration and maps

TL;DR

This paper details the calibration, map-making, and systematic error analysis of the Planck HFI data over 2.5 years, producing high-precision temperature and polarization maps across multiple frequencies.

Contribution

It introduces a dual calibration scheme using planetary models and the CMB dipole, and presents the first polarization maps from the HFI with systematic error assessment.

Findings

CMB solar dipole amplitude measured as 3364.3 +/- 1.5 μK

High-quality temperature and polarization maps produced from 2.5 years of data

Identification and analysis of systematic effects like detector flux mismatch

Abstract

This paper describes the processing applied to the Planck High Frequency Instrument (HFI) cleaned, time-ordered information to produce photometrically calibrated maps in temperature and (for the first time) in polarization. The data from the entire 2.5 year HFI mission include almost five independent full-sky surveys. HFI observes the sky over a broad range of frequencies, from 100 to 857 GHz. To obtain the best accuracy on the calibration over such a large range, two different photometric calibration schemes have been used. The 545 and 857 GHz data are calibrated using models of planetary atmospheric emission. The lower frequencies (from 100 to 353 GHz) are calibrated using the time-variable cosmological microwave background dipole, which we call the "orbital dipole". This source of calibration only depends on the satellite velocity with respect to the solar system. Using a CMB temperature of 2.7255 +/- 0.0006 K, it permits an independent measurement of the amplitude of the CMB solar dipole (3364.3 +/- 1.5 \mu K) which is approximatively 1\sigma\ higher than the WMAP measurement with a direction that is consistent between both experiments. We describe the pipeline used to produce the maps of intensity and linear polarization from the HFI timelines, and the scheme used to set the zero level of the maps a posteriori. We also summarize the noise characteristics of the HFI maps in the 2015 Planck data release and present some null tests to assess their quality. Finally, we discuss the major systematic effects and in particular the leakage induced by flux mismatch between the detectors that leads to spurious polarization signal.