[Paper Review] Planck 2018 results. III. High Frequency Instrument data processing and frequency maps

Planck Collaboration, N. Aghanim+153 more|arXiv (Cornell University)|Jul 17, 2018

Astrophysics and Cosmic PhenomenaPhysics and Astronomy37 references64 citations

TL;DR

This paper presents the Planck 2018 High Frequency Instrument (HFI) data processing pipeline, which significantly improves map-making accuracy through end-to-end simulations, refined calibration using the CMB dipole, and enhanced mitigation of systematic effects such as intensity-to-polarization leakage and bandpass mismatch. The key result is a 10−4-level absolute calibration accuracy and the first robust measurement of the reionization optical depth using HFI data, enabling sub-arcminute Solar dipole determination and reduced residual systematics across all frequency channels.

ABSTRACT

This paper presents the High Frequency Instrument (HFI) data processing procedures for the Planck 2018 release. Major improvements in mapmaking have been achieved since the previous 2015 release. They enabled the first significant measurement of the reionization optical depth parameter using HFI data. This paper presents an extensive analysis of systematic effects, including the use of simulations to facilitate their removal and characterize the residuals. The polarized data, which presented a number of known problems in the 2015 Planck release, are very significantly improved. Calibration, based on the CMB dipole, is now extremely accurate and in the frequency range 100 to 353 GHz reduces intensity-to-polarization leakage caused by calibration mismatch. The Solar dipole direction has been determined in the three lowest HFI frequency channels to within one arc minute, and its amplitude has an absolute uncertainty smaller than $0.35\mu$K, an accuracy of order $10^{-4}$. This is a major legacy from the HFI for future CMB experiments. The removal of bandpass leakage has been improved by extracting the bandpass-mismatch coefficients for each detector as part of the mapmaking process; these values in turn improve the intensity maps. This is a major change in the philosophy of "frequency maps", which are now computed from single detector data, all adjusted to the same average bandpass response for the main foregrounds. Simulations reproduce very well the relative gain calibration of detectors, as well as drifts within a frequency induced by the residuals of the main systematic effect. Using these simulations, we measure and correct the small frequency calibration bias induced by this systematic effect at the $10^{-4}$ level. There is no detectable sign of a residual calibration bias between the first and second acoustic peaks in the CMB channels, at the $10^{-3}$ level.

Motivation & Objective

• To improve the accuracy and reliability of Planck HFI frequency maps through advanced data processing and systematic error mitigation.
• To achieve sub-arcminute precision in the Solar dipole direction and 10−4-level absolute calibration accuracy using the CMB dipole.
• To reduce intensity-to-polarization leakage and bandpass mismatch effects through detector-specific calibration and end-to-end simulations.
• To validate the data processing pipeline using null tests, cross-spectra, and simulation-based consistency checks.
• To provide a robust, high-fidelity data product for cosmological analysis, particularly for CMB anisotropy and reionization studies.

Proposed method

• Employed end-to-end simulations to model and correct for instrumental systematics, including ADC non-linearities and bolometer drifts.
• Implemented a new map-making scheme (SRoll) that processes single-detector timelines with frequency-dependent bandpass corrections.
• Calibrated HFI data using the CMB dipole as a primary photometric reference, achieving 0.35 µK absolute uncertainty in dipole amplitude.
• Extracted frequency-dependent bandpass-mismatch coefficients per detector during map-making to correct for spectral response differences.
• Used cross-spectra and null tests (e.g., odd-even ring, survey nulls) to validate residual systematics and consistency with input simulations.
• Applied an empirical transfer function to correct for low-multipole residuals in the beam response, reducing leakage effects.

Experimental results

Research questions

• RQ1 What is the impact of ADC non-linearity residuals on HFI calibration, and can they be corrected at the 10−4 level?
• RQ2 How accurately can the Solar dipole be determined using HFI data, and what is its impact on absolute calibration?
• RQ3 To what extent do bandpass mismatch and calibration errors contribute to intensity-to-polarization leakage in the maps?
• RQ4 Can end-to-end simulations accurately reproduce detector gain variations and systematic effects in the data?
• RQ5 What is the residual level of systematics in the final frequency maps across different multipoles (ℓ = 4–5, 100, 2000)?

Key findings

• The Solar dipole direction was determined to within one arc minute in the three lowest HFI frequency channels, with an absolute amplitude uncertainty below 0.35 µK, achieving 10−4-level calibration accuracy.
• Intensity-to-polarization leakage was reduced to levels below 1 × 10−3 µK2 at multipoles ℓ ≈ 100 and below 5 × 10−2 µK2 at ℓ ≈ 2000, primarily through bandpass-mismatch coefficient fitting.
• End-to-end simulations reproduced detector gain calibration and drifts with high fidelity, enabling correction of ADC non-linearity-induced biases at the 10−4 level.
• No detectable residual calibration bias was found between the first and second acoustic peaks in the CMB power spectrum, at the 10−3 level.
• The total bandpass mismatch leakage was estimated at 5 × 10−2 µK2 at 353 GHz for multipoles ℓ ≈ 2000, with dust and CO contributions being the dominant components.
• The empirical transfer function reduced low-multipole residuals, with oscillations in the odd-even survey cross-spectra suppressed to below 1 × 10−5 µK2.