Optimal bolometer transfer function deconvolution for CMB experiments through maximum likelihood mapmaking

TL;DR

This paper introduces a maximum likelihood mapmaking method that simultaneously deconvolves bolometer transfer functions and pixelizes CMB data, reducing beam asymmetry and improving data quality for experiments like Planck HFI.

Contribution

It presents a novel integrated approach for transfer function deconvolution and mapmaking, improving beam symmetry and accuracy over traditional two-step methods.

Findings

Effective beam ellipticity reduced by 64%

Beam FWHM decreased by 2.3%

Method demonstrated with simulated Planck HFI data

Abstract

We revisit the impact of finite time responses of bolometric detectors used for deep observations of the cosmic microwave background (CMB). Until now, bolometer transfer functions have been accounted for through a two-step procedure by first deconvolving an estimate of their Fourier-space representation from the raw time-ordered data (TOD), and then averaging the deconvolved TOD into pixelized maps. However, for many experiments, including the Planck High Frequency Instrument (HFI), it is necessary to apply an additional low-pass filter to avoid an excessive noise boost, which leads to an asymmetric effective beam. In this paper we demonstrate that this effect can be avoided if the transfer function deconvolution and pixelization operations are performed simultaneously through integrated maximum likelihood mapmaking. The resulting algorithm is structurally identical to the artDeco algorithm introduced by Keih\"anen & Reinecke (2012) for beam deconvolution. We illustrate the relevance of this method with simulated Planck HFI 143 GHz data, and find that the resulting effective beam is both more symmetric than with the two-step procedure, resulting in a sky-averaged ellipticity that is 64 % lower, and an effective beam full-width-at-half-maximum (FWHM) that is 2.3 % smaller. Similar improvements are expected for any other bolometer-based CMB experiments with long time constants.