"Galactic science and CMB foregrounds"
2nd Workshop of the JSPS Core to Core CMB Workshop series
Tenerife, Spain (Hybrid), 12-15 December 2022
"Galactic science and CMB foregrounds"
2nd Workshop of the JSPS Core to Core CMB Workshop series
Tenerife, Spain (Hybrid), 12-15 December 2022
↻
The wealth of cosmological information carried by the cosmic microwave background (CMB) is obscured by the foreground emission from the interstellar medium (ISM) of our Galaxy. Current and next-generation CMB experiments will scrutinize the millimeter sky with exquisite sensitivities, in search for the imprints of primordial gravitational waves generated during inflation, weak gravitational lensing, neutrino mass and hierarchy, light relics or cosmic birefringence. Yet, in order to achieve these thrilling scientific goals, the separation from the foregrounds must be achieved at a precision that require dedicated instrumental and data analysis strategies, as well as additional astrophysical knowledge. In turn, these unique millimeter observations provide the ISM science with abundant new data from which rich findings can be drawn. In this review talk, I will start by giving an overview of the most important scientific goals of CMB experiments. I will present our current understanding of the Galactic foregrounds as well as the remaining questions that have to be addressed. Finally, I will present some of the potential ISM science breakthroughs that can be foreseen with future CMB datasets.
I will review the current status of the QUIJOTE (Q-U-I JOint TEnerife) experiment, a project with the aim of characterising the polarisation of the Cosmic Microwave Background, and other galactic or extragalactic physical processes that emit in microwaves in the frequency range 10-42GHz, and at large angular scales (1 degree resolution). In particular, I will discuss the scientific results associated with the wide survey carried out with the first QUIJOTE instrument (MFI) at 11, 13, 17 and 19GHz, covering approximately 29000 deg$^2$ with polarisation sensitivities in the range of 35-40 $\mu$K/deg. I will present the final maps of the survey and their main characteristics (noise levels, calibration uncertainties, power spectra, etc), and I will summarize the list of scientific publications that constitute the MFI wide survey release. These QUIJOTE MFI maps are used to characterise the radio foreground components in the northern sky, with special emphasis on the synchrotron and AME components.
Interstellar dust grains impart a polarization on starlight that can be used to trace the magnetic field, similarly to polarized dust emission. With knowledge of stellar distances, stellar polarization can be used to disentangle 3D information about dust and magnetic field properties along the line of sight. I will review recent advances in the field that make use of this powerful tracer of the magnetized ISM. I will discuss the potential of combining stellar polarization datasets with polarized dust and synchrotron emission to improve foreground modeling, map 3D magnetic fields, and explore dust microphysics. Finally, I will present avenues for further investigation through existing and future datasets with focus on the upcoming survey PASIPHAE.
3D maps of dust extinction density in the Milky Way can be constructed by inversion of large catalogues of direction, distance and extinction for individual target stars. Progress is currently being made in this field thanks to the Gaia mission and massive ground-based stellar surveys. After a brief review of existing maps and their specificity, I will focus on the most recent inversion (Vergely et al, 2022) and show how high- latitude dust clouds compare with the Galactic dust opacity deduced from Planck. I will discuss the limitations and prospects for the detection of more tenuous high latitude clouds and the reconstruction of the cloud shape, as well as the inferences from the 3D map on the debated distance to the conspicuous Loop I structure.
As the oldest radiation permeating space, the cosmic microwave background (CMB) can be used to test the fundamental symmetries of nature. In particular, a violation of parity symmetry in electromagnetism would have rotated E modes into B modes in the observed CMB signal by an effect known as cosmic birefringence. In the past, attempts to measure cosmic birefringence have been limited by the precision in calibrating the instrument's polarization angles. Recently, a novel methodology has allowed us to bypass that limitation and simultaneously calculate birefringence and miscalibrated polarization angles by using Galactic foreground emission as our calibrator. Yet, the misalignment between the filamentary dust structures of the interstellar medium (ISM) and the plane-of-sky orientation of the Galactic magnetic field induces a non-null EB correlation on Galactic dust emission that can bias our measurement of birefringence. Thus, understanding the physics of the magnetic ISM and having a precise model of polarized dust emission are fundamental aspects of the search for cosmic birefringence. In this talk I will review the recent measurements of cosmic birefringence and current models for dust EB, highlighting the synergies between the Galactic science and CMB communities.
Magnetic fields in the Milky Way play a significant role in the Galactic ecosystem and thoroughly influence CMB polarization measurements, yet are hard to characterize. Observations of polarization in various wavelength regimes, caused by various physical processes involving magnetic fields, give observational constraints to models. These models typically consist of two components: a large-scale component which is coherent over galaxy scales, and a fluctuating part caused by turbulence. In this presentation, I will give an overview of recent modeling efforts, with special emphasis on the IMAGINE project.
This review covers the numerical efforts in simulating the multiphase interstellar medium (ISM) on different spatial scales. I will first give an overview of the processes that need to be included in present-day simulations like chemistry, star-formation, magnetic fields, radiation and cosmic rays. I show applications of the matter cycle in the ISM from the condensation of molecular clouds to the formation of stars and to the onset of galactic outflows driven by stellar feedback. After these general ISM simulations I will focus on setups that are closer to our local conditions. On small scales this covers simulations of the Local Bubble around the solar system, in which we are able to identify the relevant processes that influence our foreground view. On larger scales I will show current efforts in simulating a close analogue of the Milky Way including correct details of the Galactic magnetic field and spiral arm potential in addition to the aforementioned necessary ISM processes.
Three-dimensional magnetic field observations are required to determine the foreground Galactic contamination in cosmic microwave background polarization studies. To determine the 3D magnetic fields associated with giant molecular clouds, we first developed a novel technique for measuring the line-of-sight strength and direction of magnetic fields. To reconstruct the 3D magnetic fields, we then incorporated our line-of-sight and Planck's plane-of-sky magnetic field observations, along with models and statistical techniques. In this talk, we will discuss some of our recent work in mapping the three-dimensional interstellar magnetic fields and extending these techniques to more diffuse regions.
In the CMB frequency range, the Galactic foregrounds at high Galactic latitudes are known to be utterly complex, with contamination coming from the neutral and ionized phases of the interstellar medium. How the Galactic emission projects on the sky, especially in polarization, depends heavily on how matter and magnetic fields are organized in three-dimensions. In this talk I will present what we think we know on the nature of the diffuse interstellar medium structure, and how it relates to the observed properties of Galactic emission.
Many challenges of modern astrophysics are related to the study of non-linear processes with highly non-Gaussian statistics (interstellar medium, CMB galactic foregrounds, large scale structures of the Universe...). Although being an important lever arm, the efficient use of these non-Gaussian features remains however difficult, as shown by the increasing use of machine learning. Recently, the Scattering Transforms (ST) statistics have obtained promising results in the characterization of non-Gaussian processes. These statistics are inspired by neural networks, but can be computed without any learning step. One of their main advantages is that non-Gaussian generative models of a given process can be constructed directly from their ST statistics, even from a very small amount of data. ST also allowed to introduce new components separation methods that are able to separate different processes based on their non-Gaussian structure, even from data at a single frequency. In this talk, I will introduce ST statistics, and explain how they can be used to build realistic non-Gaussian models of Galactic foregrounds to the CMB directly from observations.
In the recent years it has become clear how the contamination coming from Galactic emission represents one of the main limiting factor for any new science achievable with Cosmic Microwave Background observations. Having a thorough understanding of the foreground properties is therefore fundamental in order to achieve a reliable reconstruction of the clean CMB signal and to constrain its cosmological properties. In this talk I will review the current status of knowledge and characterization of foregrounds, focusing in particular on their contamination to primordial CMB B-modes. I will summarize the main results from the latest analysis of available multi-frequency data, highlighting which are the main limitations in our current models and how we can overcome them in view of future CMB experiments.
Originally, the moment expansion method was developed for the modeling of the SZ effect in clusters with full 3D atmospheres, to take into account the spatial variations in the electron temperature and bulk velocity. However, the same method could be extended to the modeling of CMB foregrounds, where it now serves as a powerful tool for generalizing SED parametrizations in the presence of fluctuating spectral parameters. In this talk, I will review some of the background of the moment expansion method and its applications in cosmology.
A number of spectral distortions signals are predicted to exist in the standard model of cosmology. The amplitudes of these signals are very small (comparable to the primordial B-mode signal) and obscured by galactic and extra-galactic foregrounds brighter by a few orders of magnitude. Consequently the measurement is extremely challenging, requiring unprecedented frequency coverage, instrumental sensitivity and calibration. I will begin by briefly reviewing the current set of tools being used to forecast SD measurements with planned spectrometers, highlighting potential drawbacks. I will then discuss the component separation tool being actively developed for SD measurements - this exploits the best of blind as well as parametric component separation methods. Finally I will discuss some results and lessons learnt from recent work that constraints anisotropic spectral distortion using Planck data.
With unprecedented sensitivity, next-generation CMB experiments aim at detecting new faint cosmological signals, the most sought after being the primordial CMB B-mode polarization signal predicted by cosmic inflation models. The huge amplitude discrepancy between Galactic foregrounds and primary CMB B-modes makes component separation very challenging and much more sensitive to any foreground modeling errors. Overcoming the new foreground challenge in polarization requires innovative component separation methods. I will start reviewing a few typical parametric and non-parametric approaches to component separation and highlight their formal relationship and differences. I will then address some specific challenges on foreground modeling that we have to face for B-mode component separation: spectral mismodeling, spectral degeneracies and spectral distortions. I will then discuss new developments in semi-blind component separation relying on moment expansion and show how they offer an interesting avenue to overcome these new challenges.
The microwave emission from our own Galaxy is observed as a foreground contamination to detect the Cosmic Microwave Background (CMB) B-mode polarization. In the past decades, techniques (commonly known as component separation algorithms) have been proposed in the literature and have been shown to successfully separate and reconstruct the unpolarized CMB emission from the foreground one. This is also due to the availability of unprecedented templates from multi-frequency observations (e.g. WMAP, Planck). However, the situation in polarization is still challenged by the fact that the B-mode polarization is orders of magnitude weaker and that the Galactic polarized emission has just started to be probed. In this talk, we will show how novel techniques involving supervised and unsupervised learning encoding clustering techniques can be adopted in order to improve the performances of parametric component separation. Moreover, we have identified the number of clouds along the line of sight as a good tracer to infer the properties of Galactic dust. This is particularly relevant for future CMB experiments (e.g. SO, LiteBIRD, CMB-S4 ), where high sensitivities are expected to be achieved.
The C-Band All-Sky Survey (C-BASS) is a 5GHz radio survey at 45 arcminute resolution in both total intensity and polarisation. C-BASS provides a unique view on the Galactic synchrotron emission that will be critical to the success of future CMB missions such as Litebird and Simons Observatory, and also to the understanding of Galactic diffuse emission and discrete sources that includes supernova remnants, synchrotron loops, Hii regions, and the Galactic magnetic field. At present the C-BASS North survey has been completed and the northern survey data are close to release, while the southern half of the C-BASS survey has already begun. In this talk, I will describe the C-BASS North instrument and observations, the challenges involved in performing the northern survey, and present the latest total intensity and polarisation maps. Finally, I will give an overview of the total intensity results from C-BASS and how they impact on our current understanding of anomalous microwave emission in both our Galaxy and in the Andromeda galaxy, M31.
The C-Band All-Sky Survey (C-BASS) provides a unique view of the polarized Galactic synchrotron emission at 5 GHz, which will be critical to next-generation CMB surveys. In this talk I will present results from the analysis of polarized synchrotron emission in the northern sky, in combination with legacy data from WMAP and Planck. I will show map-based spectral index estimates for the polarized intensity, and derived estimates for polarized synchrotron contamination to CMB B-modes. I will also present results from a harmonic-space analysis, including C-BASS E and B power spectra, cross-spectra with WMAP and Planck, and constraints on the synchrotron EB correlation.
The South Pole Telescope (SPT) is a 10-meters diameter telescope observing the Cosmic Microwave Background (CMB) from the South Pole, with arcminute angular resolution. A third generation camera (SPT-3G) was mounted on SPT in 2018, showing the remarkable performances of this experiment, which is currently one of the most competitive for CMB science. The final goal of SPT-3G is to measure the lensing CMB B-modes with very high precision, in order to complement BICEP/Keck observation of large scales CMB polarisation. Simultaneously, the very high quality data of SPT-3G will allow us to achieve cosmological constraints at a level of precision comparable with Planck’s, or even better. For these goals, SPT-3G observes a baseline sky patch covering 1700 deg2 with very small foregrounds contamination, and three extended fields covering additional ~3000 deg2 with slightly larger foregrounds contamination. In this talk, I will describe the SPT-3G experiment, presenting current results and future prospects for cosmological constraints. Given the subject of this conference, I will also briefly focus on the level of foregrounds that we observe in the fields observed by SPT-3G.
Polarized Galactic dust emission encodes the plane-of-sky magnetic field orientation. We employ 1′ resolution intensity and polarization observations from the Atacama Cosmology Telescope (ACT) and 15′′ infrared photometry from the Wide-field Infrared Survey Explorer (WISE) to study the relationship between dust intensity structures and magnetic fields in ∼ 100 deg^2 patches of the sky, totalling O ∼ 10^4 deg^2, approximately 25% of the sky. The ACT × WISE intensity power spectrum is well fit by a power law from ℓ = 10^3 to ℓ = 10^4, with no evidence for a break, in regions across the sky, from dusty regions near the Galactic Plane to higher galactic latitudes. We measure a positive correlation between WISE intensity and ACT E-mode polarization at the smallest scales we observe. We interpret this as evidence that the magnetic field is statistically aligned parallel to the density structures on small scales. Both TT and TE are well-described by scale-invariant power laws in the range we are sensitive to. These results are consistent with prior studies, however we find there is significantly variability in the power law indices across regions of sky. This work constrains the TT and TE spectrum to smaller scales than previous analyses, indicating that the millimeter dust polarization at frequencies probed by CMB experiments is highly correlated with the 12 μm emission down to ℓ ∼ 10000 in TT and ℓ ∼ 5000 in TE.
We assess the uncertainty with which a balloon-borne experiment, nominally called Tau Surveyor (τS), can measure the optical depth to reionization σ(τ) with one mid-latitude flight and given realistic constraints of instrument noise and foreground emissions. Using a τS fiducial design with six frequency bands between 150 and 380 GHz with white and uniform map noise of 7 μK arcmin and including Planck 's 30 and 44 GHz data we assess the error σ(τ) achieved with three foreground models and as a function of sky fraction fsky between 40% and 54%. We carry out the analysis using both parametric and blind foreground separation techniques. We compare σ(τ) values to those obtained with low frequency and high frequency versions of the experiment called τS-lf and τS-hf that have only four and up to eight frequency bands with narrower and wider frequency coverage, respectively. We find that with τS the lowest constraint is σ(τ)=0.0034, obtained for fsky=54%. σ(τ) is larger, in some cases by more than a factor of 2, for smaller sky fractions, with τS-lf, or as a function of foreground model. The τS-hf configuration does not lead to significantly tighter constraints. Exclusion of the 30 and 44 GHz data, which give information about synchrotron emission, leads to significant τ mis-estimates. Decreasing noise by an ambitious factor of 10 while keeping fsky=40% gives σ(τ)=0.0031. The combination of σ(τ)=0.0034, BAO data from DESI, and future CMB B-mode lensing data from CMB-S3/S4 experiments could give σ(∑mν)=17 meV. More details can be found in our paper: https://arxiv.org/abs/2206.03389
GroundBIRD is a ground-based Cosmic Microwave Background (CMB) polarimeter for large-angular scale (6 < \ell < 300) observations at the Teide Observatory in Tenerife, Spain. GroundBIRD observes the CMB with fast rotation of the telescope (1 rotation every 3 seconds) in order to mitigate 1/f noise introduced by fluctuations of the atmosphere. We use Microwave Kinetic Inductance Detectors (MKID) made by TU Delft and SRON. These are a hybrid type MKID for high sensitivity and fast response times of O(100us) for observations with fast rotation. We aim to observe at a frequency of 145 GHz for the CMB with 138 detectors and 220GHz for thermal dust emission with 23 detectors. Although ground-based CMB experiments have been progressively improving accuracy of foreground sky maps, the sky areas covered by them are primarily in the southern hemisphere. Among them, GroundBIRD is an unique experiment that covers the Northern hemisphere with wide sky coverage in millimeter wave bands and with sufficiently high sensitivity. One of the science targets of GroundBIRD is a precise measurement of optical depth “\tau” in the reionization era of the universe. Space-based measurements of “\tau” have a systematic tendency for the central value to decrease. Therefore independent measurement of ‘\tau’ is important. GroundBIRD can achieve almost the same sensitivity as the Planck satellite for the \tau measurement by combined analysis with QUIJOTE, providing a complementary measurement with different systematics. QUIJOTE is also a CMB polarimeter next to GroundBIRD, which covers the 10-40GHz frequency range. The combined analysis with QUIJOTE provides unique spectral coverage, which allows us to remove foregrounds with sufficient accuracy. We are now commissioning the telescope, with early science observations using 23 MKIDs, which are almost the same design as those to be used for the full science observation. We are investigating methods for calibration by using the Moon: pointing, responsivity, polarization angle, etc. A remote observation system has been established to operate the telescope continuously and safely. In this talk, I will present the concept of the GroundBIRD experiment, initial calibration results, and the future prospects for GroundBIRD.
We present the first joint analysis of WMAP and Planck LFI data, presenting maps that have been generated from a fully consistent joint treatment, including the sampling of sky signals and instrumental properties. The joint analysis approach yields improved WMAP data with better treatment of poorly constrained modes, as well as the first fully optimal sampling of all nine years of data. We also improve on the BeyondPlanck analysis by reducing poorly measured modes in LFI polarization. In particular, we find a ~4 uK change in the 30 GHz channel as a result of including the higher signal-to-noise WMAP K-band data. The WMAP maps we present are free of previously documented systematic effects. As the first release of Cosmoglobe products, the maps from this analysis should be considered both a considerable improvement over previous analyses, as well as the first iteration of future joint analyses with other data, including, e.g., the ground-based QUIET experiment and the DIRBE instrument aboard COBE.
Employing an half-wave plate (HWP) as polarization modulator is a promising strategy to mitigate systematics in CMB experiments. However, if not calibrated or accounted for correctly in the analysis, its non-idealities represent an additional source of systematics, which can impact the science extracted from CMB. In this talk, I'll show how their effect on the observed angular power spectra is partially degenerate with cosmic birefringence. Also, I'll discuss how this issue intertwines with the study of foregrounds, given the frequency dependence of the HWP behaviour.
The accuracy of future CMB measurements depends critically on the foreground removal process, with some methods requiring different levels of prior knowledge about the foregrounds. Galactic synchrotron emission is the major foreground component at low frequency (below ~100 GHz). Away from the Galactic plane, bright emission originates mostly from filamentary structures, the nature of which remains poorly understood. We implement a Filament Finder algorithm which allows the detection of bright elongated structures in total polarised intensity maps. We analyse the sky at 23 and 30 GHz as observed respectively by WMAP and Planck. We identify 19 filaments, 13 of which have been previously observed either in total intensity or polarised intensity. For each filament, we study the polarisation fraction, finding a value of about 25% is reached in two filaments. Typically, we find that the polarization fractions of the filaments are larger than for the areas outside the filaments, excluding the Galactic plane. We study the polarisation spectral indices of the filaments, and find a spectral index consistent with the values found in previous analysis (about –3.1) for more diffuse regions. We then focus on understanding the statistical properties of the diffuse regions of the synchrotron emission at 23 GHz. Using Minkowski functionals and tensors, we analyse the non-Gaussianity and statistical isotropy of total polarised intensity maps. For a sky coverage corresponding to 80% of the fainter emission, and on scales smaller than 6 degrees (l>l_min=30), the deviations from Gaussianity and isotropy are significantly higher than 3 sigmas. Even if the level of deviation decreases for smaller scales (increasing l_min up to 130), it remains significantly high. When 60% sky coverage is analysed, we find that the deviations never exceed 3 sigmas. In conclusion, we present a simple data-driven model to generate non-Gaussian and anisotropic simulations of the synchrotron polarised emission. The simulations are fitted in order to match the spectral and statistical properties of the data.
The MIllimeter Sardinia radio Telescope Receiver based on Array of Lumped elements kids (MISTRAL) is a millimetric total power camera operating at 90 GHz. It will be installed on the Sardinia Radio Telescope (SRT) as a part of the SRT-HighFreq program, funded by a Programma Operativo Nazionale (PON), that aims at expanding the capabilities of the radio telescope up to the W-band. After technical and scientific commissioning (expected for 2023), MISTRAL will be open to proposal submission by scientists as a facility instrument. MISTRAL provides a wide 4’ field of view, sampled at a resolution of 12” with 400 Kinetic Inductance Detectors. The operational band of MISTRAL allows accessing a number of astrophysical processes in Galactic and extragalactic environments, at higher resolution than current CMB experiments. The study of galactic regions at 90 GHz is key to discriminate emission processes like free-free, synchrotron and thermal dust, with implications in star formation models. This band is close to the peak of the currently not well understood Anomalous Microwave Emission (AME), that sits at 20-60 GHz, whose understanding is crucial as it is a foreground for CMB B-modes experiments, if polarized. High resolution W-band observations are also interesting for extragalactic astrophysics and cosmology, spanning from the blind search for microwave point sources to the observation of galaxy clusters and the elusive cosmic web in high detail with the Sunyaev-Zel’Dovich effect.
The QUIJOTE-MFI Northern Hemisphere Wide-Survey provides maps of the sky above declinations −30◦ at 11, 13, 17 and 19 GHz. These data have been combined with ancillary data to produce Spectral Energy Distributions in intensity in the frequency range 0.4–3 000 GHz on a sample of 52 candidate compact sources harbouring anomalous microwave emission (AME). In this talk I will present the component separation analysis at 1◦ scale that has been applied on the full sample from which we identify 44 sources with high AME significance. I will show that the QUIJOTE-MFI data contribute to notably improve the characterisation of the AME spectrum, and its separation from the other components. In particular, ignoring the 10–20 GHz data produces on average an underestimation of the AME amplitude, and an overestimation of the free-free component. Once such data are included the average AME peak frequency is of 23.6 ± 3.6 GHz, about 4 GHz lower than the value reported in previous studies. Finally, a discussion will be given on some of the most relevant correlations between different fitted parameters on the sample of 44 sources with high AME significance, and their implication on our understanding of the AME carriers.
Intensity foregrounds in the microwave domain are specially complex to model, mostly due to the overlapping of synchrotron and free-free emissions. Besides, anomalous microwave emission (AME) begins to be important for frequencies higher than 10\,GHz. That is why having as many data as possible covering this spectral region is needed. In this talk, we will present how the addition of novel QUIJOTE-MFI data between 10 and 20\,GHz allows to better constrain the diffuse foreground components (mainly synchrotron, free-free and AME) at large angular scales ($\sim0.9^\degr$) along the Galactic Plane ($|b| < 10^\degr$). Together with QUIJOTE-MFI, we will be using ancillary low and high frequency data to build spectral energy distributions between 0.408 and 3000\,GHz. We will discuss how the parameters describing the AME (especially its peak emission frequency) vary along the Galactic Plane, and we will show how these parameters, and those describing the rest of components, correlate between each other. Some of the recovered correlations are the ones between the thermal dust and AME amplitudes (which is currently well-known), the AME peak emission frequency and the dust temperature, or the AME emissivity and the interstellar radiation field (ISRF) strength. Finally, we will show how lacking the 10-20\,GHz data from QUIJOTE-MFI can result in an underestimation (as large of 50\%) of the AME recovered signal. These results will be included in Fernández-Torreiro, M. et al. (2022, to be submitted).
Anomalous microwave emission (AME) is an additional component of diffuse foreground emission that cannot be easily explained by synchrotron, free-free, or thermal dust emission. AME has been observed by numerous experiments in the frequency range ∼10-60 GHz within specific galactic clouds and potentially even in external galaxies. It is found to be very closely correlated with far infrared (FIR) emission associated with thermal emission from dust grains. However, its polarized properties are still uncertain as physical models predict varying levels of potential polarized emission and even a very small polarization fraction in the AME could have a not negligible impact in the CMB B-modes detection. In this work we present new Q-U-I JOint Tenerife Experiment (QUIJOTE) 10-20 GHz maps in intensity and in polarisation with a sensitivity in polarisation of ~ 30 μK beam-1 and an angular resolution of ~ 1° for ρ Ophiuchi, improved QUIJOTE maps in W43 and W47 molecular complexes and W44 supernova remnant and a preliminary analysis of the Perseus molecular cloud (G159.6-18.5). For each region we provide spectra over the frequency range 0.4-3000 GHz after combination with other publicly available intensity data, which confirm that the emission in the range between 10 and 60 Ghz is dominated by AME (except for W44 where synchrotron is prevalent). We obtained for the first time upper limits for the AME polarization fraction at the QUIJOTE frequencies for ρ Ophiuchi, and the first ever constraints on the AME polarization degree in compact regions from Planck LFI data. This resulted in the most stringent constratints in Perseus and also in W43 in some specific frequencies, improving previously published results. Our best constraints on the AME polarization fraction for Planck are 1.2% in ρ Ophiuchi and 0.95% in Perseus at 28 GHz and 0.45% at 44 GHz in W43. In W43 we also obtained 0.81% at 28 GHZ and 0.37% at 17 GHz QUIJOTE frequency.
The search for the primordial B-modes in the cosmic microwave background (CMB) polarization depends on the separation from the brighter dust foreground signal. In this context, the characterisation of the spatial variations of the spectral energy distribution (SED) of thermal dust in polarization has become a critical subject of study. This contribution aims at presenting a power spectra analysis of Planck data (Ritacco, Boulanger, Guillet et al. 2022), which improves on previous studies by using the newly released SRoll2 maps that better correct systematic effects, and by extending the analysis to regions near the Galactic plane. Our analysis focuses on the lowest multipoles between l=4 , and 32, and three sky areas with sky fractions of fsky = 80%, 90%, and 97%. The mean dust SED for polarization and the 353 GHz Q and U maps are used to compute residual maps at 100, 143 and 217 GHz, highlighting spatial variations of the dust polarization SED. Residuals are detected at the three frequencies for the three sky areas. The power spectra show that models based on total intensity data are underestimating by a significant factor the complexity of dust polarized CMB foreground. This analysis emphasizes the need to include variations of polarization angles in simulations of the dust polarized CMB foreground and the importance to consider the geometrical properties of the polarization. The frequency dependence of the EE and BB power spectra of residual maps yields further insight, suggesting that a significant refinement to dust modeling will be needed to ensure an unbiased detection of the CMB primordial B-modes at the precision required by future CMB experiments.
We report on progress to determine the anomalous microwave emission (AME) mechanism near the S140 star-forming region G107.2+5.2. This talk will outline our methodology, progress to date, and plans for future targeted polarization observations. We mapped a circular region of radius 1.25 degrees centered on G107.2+5.2 using the Ku-band (12-18 GHz) receiver at the Green Bank Telescope (GBT). After performing aperture photometry on our maps of the region and on ancillary data from CGPS, Reich, Planck, and DIRBE, we will plot the spectral flux density and fit emission models to the spectrum. The region G107.2+5.2 was identified by the Planck satellite mission as a candidate AME region, with a spectrum consistent with both spinning dust and optically thick free-free emission models. In a previous study our group observed this region using the C-band (4-8 GHz) receiver at the GBT and found a spectrum consistent with either spinning dust or ultra-compact HII; additional data in the Ku-band (12-18 GHz) were needed to definitively constrain the emission mechanism. Based on our previous results, we expect to find a spectrum consistent with spinning dust as the emission mechanism and rule out the HII scenario. If this is the case, we will morphologically examine the maps to identify the region(s) of significant AME and target them for follow-up polarization observations. Though spinning dust is predicted to be polarized at levels less than ~1%, these levels could potentially be significant for CMB B-mode searches.
The detailed understanding of the Galactic emission processes in the frequency range from 1 to 3000 GHz is crucial for a state-of-the-art characterization of the Cosmic Microwave Background (CMB) anisotropies both in intensity and polarization. The Q-U-I JOint TEnerife (QUIJOTE) experiment is one of the current observational efforts devoted to understanding the microwave sky, which at these wavelengths is dominated by four Galactic mechanisms: synchrotron, free–free, thermal dust and Anomalous Microwave Emission (AME). In this talk, we want to provide new insights on the intensity and polarization properties of supernova remnants (SNRs) in the northern sky using the QUIJOTE-MFI wide survey (10-20 GHz). In particular, I will describe the physical properties of the spectral energy distribution (SED) for eight SNRs, with emphasis on new polarization measurements in this poorly explored frequency range. In addition, we will discuss the presence of AME in these SNRs, and we will we report the detection of intensity AME signal towards W49 and W51. The results presented in this talk are described in Tramonte et al. (2022, MNRAS accepted) and Lopez-Caraballo et al. (2022, in prep).
The CO Mapping Array Project (COMAP) Galactic Plane Survey is a radio continuum and radio recombination line survey between 26 and 34 GHz with 4.5 arcminute resolution that will span Galactic longitudes of 20 < l < 270 and latitudes of |b| < 2 with an expected completion date in 2024. The COMAP Galactic plane survey will compliment the main CO intensity mapping products produced by the rest of the collaboration. The survey aims to investigate star formation through observing Galactic Hii regions, producing the first Galactic AME catalogue on scales of a few arc minutes and creating diffuse emission templates for better subtraction of Galactic emission from CMB maps. In this talk we show our initial published results from Rennie et al. 2022 including some diffuse T-T analysis alongside spectral energy distributions of six supernova remnants and nine HII regions. Of these HII regions we find six to show significant excess at 30 GHz which we interpret as anomalous microwave emission (AME), fitting for the peaked AME spectrum using a log-normal model and finding an average peak frequency of ~44 GHz. We also show targeted, preliminary maps of AME and star forming regions to be published in upcoming works such as the SNR Westerhout 44.
The BeyondPlanck project has created a suite of frequency maps, each representing a sample from full end-to-end error propagation from Planck LFI time-ordered data. In this talk I will first describe how we utilized these data and a suite of novel Gibbs sampling techniques to set upper bounds on the maximum large-scale polarization fraction of anomalous microwave emission. For a reasonable set of synchrotron spectral indices, we find p < 2.5 %. Finally, I will preview how using these new techniques we are geared to test the feasibility of single vs. multi-component dust models.
The paths of cosmic rays are deflected upon passing-through the Galactic magnetic field structure. The strength of the deflections that these cosmic rays undergo is dependent on the strength and structure of the Galactic magnetic field. Unfortunately, our knowledge of the Galactic magnetic field is very limited, especially when considering the fields present out in the Galactic halo bubble region. In this talk, I wish to motivate the importance of the Galactic halo bubble magnetic fields not only from the point of view of radio observations but also for ultra high energy cosmic ray propagation. The observations from eROSITA and FERMI-LAT show clear evidence of large extended structures out in the Galactic halo region, with a total energy of up to ~1056ergs seen in thermal X-rays. However, most of the widely used Galactic magnetic field models focus predominantly on the Galactic disc rather than the halo. In Ref.~[1] we motivate a toy magnetic field model for the Galactic halo bubbles. We use this model, in combination with an analytic expression for the non-thermal electron distribution, to generate synthetic polarised synchrotron maps and compare them with 30 GHz Planck data. We obtain constraints on the parameters of our model and utilise these constraints to create arrival direction maps for ultra high energy cosmic rays. We conclude that present uncertainties in the field strengths can have major consequences on the arrival directions of the cosmic rays and thereby the source localisation. [1] V. Shaw, A. van Vliet, A. M. Taylor, arXiv:2202.06780
Observations of the Galactic synchrotron emission provide one of the most valuable tool to understand the large-scale Galactic magnetic field (B-field). However, there is a huge model degeneracy between B-field and cosmic-ray electron/positron density across the Galaxy. The consequence is that intensities and distribution of ordered and random components of the B-field are still very uncertain. To tackle this issue, in the past years we have developed a large effort to study simultaneously and consistently the B-field and cosmic rays with the help of cosmic-ray propagation models and multi-messenger data. Here we present the status of our study and recent updates in this context, underlying the importance of using available data and models in a consistent picture.
The Galactic magnetic field gives rise to polarized emission that constitutes a confusing foreground to the CMB. Modeling this foreground emission to the accuracy required for future CMB experiments requires new approaches in characterizing the interstellar magnetic field. In particular, the 3D nature of the magnetic field constitutes a key unknown in modeling efforts, as suggested by a growing body of work. I will present a promising method to measure 3D properties of the Galactic magnetic field through starlight polarization tomography. I will demonstrate the first autonomous Bayesian method for tomographic reconstruction of the magnetic field in dusty regions, with the use of stellar polarization and parallax. I will discuss the advantages and limitations of the method and present a first application on data from a PASIPHAE-pathfinder survey.
Has cosmic inflation occurred? If so, how? What is the fundamental physics behind inflation? How can we measure Inflation and what are observation challenges? Primordial gravitational waves are a privileged tool to answer these questions, since they allow to access the very Early Universe and discriminate between different physical mechanisms acting during inflation. In this talk I will emphasize how an exquisite control of astrophysical foregrounds in both future CMB observational campaigns (such as the LiteBIRD satellite) and direct detection experiments (such as space laser interferometers and pulsar timing arrays) will be required to disentangle the different possible origins of primordial gravitational waves. I will present results on gauge-fields contribution to inflation using realistic simulated component separated maps for the LiteBIRD satellite and realistic forecasts for future direct detection experiments, including the effect of astrophysical foregrounds marginalization through a new filter that we derived for cross-correlation in laser interferometers and pulsar timing array.
This talk will present the first large-scale 3 dimensional map of the temperature of the dust in the interstellar medium. Having a better understanding of the 3D dust can help us characterize the 3D structure of the magnetic field, and compute the six-dimensional phase-space density of the interstellar radiation field (the amount of light in every 3D voxel, in every direction, at every energy). In this talk I will show how large scale emission maps can be used to constrain the temperature of the dust in the interstellar medium in 3D. I build upon existing 3D reddening maps derived from starlight absorption (Bayestar19), covering 3/4 of the sky. Starting with the column density for each of 500 million 3D voxels, I propose a temperature and emissivity power-law slope (β) for each of them, and integrate along the line of sight to synthesize an emission map in five frequency bands observed by Planck and IRAS. The reconstructed emission map is constrained to match observations on a 10′ scale, and does so with good fidelity. I produce 3D temperature maps at resolutions of 110′, 55′,and 27′. I assess performance on Cepheus, a dust cloud with two distinct components along the line of sight, and find distinct temperatures for the two components. I thus show that this methodology has enough accuracy to constrain clouds with different temperature along the line of sight up to 1 − σ error. This could be an important result for dust frequency decorrelation analysis for cosmic microwave background experiments, which would be impacted by a line-of-sight with varying temperature and magnetic field components.
Where dust and gas are uniformly mixed, atomic hydrogen can be traced through the detection of FIR emission of dust. Discrepancies are observed between various direct and indirect tracers of gas outside the Galactic plane. We considered, for the origin of those discrepancies possible corrections to the zero levels of the Planck-HFI detectors. We set the zero levels of FIR skymaps from the correlation between FIR emission and atomic hydrogen column density excluding regions of lowest gas column density. When comparing the newly derived τ353 and NHI, we observed a uniform spatial distribution of the opacity outside regions with dark neutral gas and CO except in various large-scale regions of low NHI that represent 25% of the sky. In those regions, we observed an average dust column density 45% higher than predictions based on NHI with a maximum of 250% toward the Lockman Hole region.
I will present a recent investigation (arXiv:2208.07382) of misalignments between interstellar dust filaments and magnetic fields. Such misalignments can produce parity-violating signatures in polarized dust emission (observable, for example, as TB and EB cross-spectra). Parity violation of this kind would be an important input to magnetohydrodynamic models of the interstellar medium and also to measurements of cosmic birefringence in the cosmic microwave background, for which the dust polarization is a significant foreground. Using neutral hydrogen (HI) to identify filamentary structures and millimeter-wave polarization to infer the magnetic-field orientation, we find 1) a global misalignment angle of approximately 2 degrees, 2) a tendency for magnetic misalignment to be scale (multipole) independent and 3) that magnetic misalignment is predictive of the variation in dust TB and EB cross-spectra. Relatedly, I'll present a new HI-based dust-polarization template, which is constructed from the Hessian matrix of the HI intensity, i.e., without any explicit dust information, and which is found to correlate more strongly than previous templates with Planck dust B modes.
The neutral hydrogen HI exhibits complex morphology that encodes rich information about the physics of the interstellar medium (ISM). In this work, we explore the application of the scattering transform (ST) to characterize HI emission structure, and demonstrate the connection between the ST morphology measures and HI cold neutral medium (CNM) phase content. HI emission has been shown to be a promising tracer of foreground reddening. In this context, resolving HI phase structure has important implications for ISM physics in connection to CMB research. It is essential for quantifying dust-gas mixing variation due to the mixture of ISM phases along the line of sight (LOS), and for calibrating and constructing precise foreground maps. However, accurate determination has so far been limited by the availability of HI absorption measurements. Existing methods of indirect estimation from emission relies on complex spectral line decomposition. Here we present the first probe of CNM content using measures solely derived from HI emission morphology information using a powerful statistical technique called the scattering transform. We apply ST to GALFA-HI data at high Galactic latitude (b>30°). The resulting coefficients are consistent with the picture that small-scale filamentary structures are preferentially CNM. Model-independent correlation study using CNM fraction data derived from absorption measurements shows that ST morphology measures encode substantial CNM-correlating information, which is further corroborated by the enhancement of I857/NHI ratio with increasing ST measure of small-scale linearity. The results demonstrate the direct connection between HI morphology and phase content, and show that future phase decomposition methods can be improved by making use of both HI spectral and spatial information.
Synchrotron emission from our Galaxy constitutes the brightest source of photons in a wide range of frequencies ranging from radio to microwaves. Understanding its statistical properties is valuable from the perspective of observations targeting signals of cosmological origin, as well as from the perspective of understanding the physical processes in our Galaxy. With this objective, we have studied the statistical properties, namely the nature of non-Gaussianity and statistical isotropy, of the 408 MHz Haslam synchrotron map using morphological statistics, such as Minkowski functionals and tensors. We find that as we go to smaller scales, the field becomes less non-Gaussian. This is in agreement with the Gaussian approximations taken while modelling the synchrotron fluctuations at small scales. However, our results indicate that the field has non-negligible amount of non-Gaussianity at Haslam resolution. Our analysis shows that the nature of non-Gaussianity in the synchrotron field resembles that of the kurtosis type of non-Gaussianity. We have studied the anisotropy of the synchrotron emission using Minkowski tensors and found that they are increasingly isotropic at small scales. Next, we extend this formalism to other synchrotron maps, mainly the ones given by Planck and WMAP. Here, our motivation is twofold – to understand the frequency dependence of the morphological features of synchrotron emission and to see how well the synchrotron maps given by the component separation pipelines in Planck and WMAP reproduce the results we got previously. In this talk, I will summarize these results and give new directions in using morphological tools to improve our understanding of Galactic emissions.
Most of the efforts of the CMB scientific community are focused on the search for primordial B-modes. However, their detection is extremely challenging since they are obscured by other B-mode sources, e.g., astrophysical sources, E- to B-modes conversion due to gravitational lensing, systematic errors, etc. In this work, we addressed the issue of the astrophysical sources by improving our understanding of Galactic foregrounds using the final data from the QUIJOTE-MFI instrument along with ancillary data from WMAP and Planck. We provide linearly polarized astrophysical component maps of the Northern Sky obtained using the parametric component separation method B-SeCRET. The addition of QUIJOTE-MFI data significantly improves the parameter estimation of the low-frequency foregrounds, in particular the estimation of the synchrotron spectral index, $\beta_s$. We present the first detailed $\beta_s$ map of the Northern Celestial Hemisphere at a smoothing scale of 2$^{\circ}$. We find statistically significant spatial variability across the sky. We find that a power law model provides a good fit of the synchrotron emission outside the Galactic plane but fails to track the complexity of this region. On the other hand, we detect a uniform curvature when we fit the synchrotron using a model that includes curvature. However, the statistical significance is insufficient to determine which model, the power law or the power law with uniform curvature, best fits the data.
Dust polarization is important for both our understanding of astrophysical processes in the interstellar medium (ISM) and the search for primordial gravitational waves in the cosmic microwave background (CMB). In this talk, I will present our work on characterizing Galactic dust filaments by correlating BICEP/Keck and Planck data with polarization templates based on neutral hydrogen (HI) observations. In the diffuse ISM, HI is strongly correlated with the dust and partly organized into filaments that are aligned with the local magnetic field. The BICEP/Keck instruments have collected deep data at 95, 150, and 220 GHz, over the low-column-density region of sky where BICEP/Keck has set the best limits on inflationary theories. I will show how we separate the HI emission in that region into distinct velocity components and demonstrate a detection of dust polarization correlated with the local Galactic HI but not with the HI associated with Magellanic Stream I. Then, I will present a robust, multi-frequency detection of polarized dust emission correlated with the filamentary HI morphology template down to 95 GHz, which allows us to measure the spectral index of the filamentary dust component spectral energy distribution. I will also describe how we find no evidence for decorrelation in this region between the filaments and the rest of the dust field or from the inclusion of dust associated with the intermediate velocity HI. Finally, I will discuss our exploration of the morphological parameter space in the HI-based filamentary model and our efforts to improve this model.
Planck has mapped the entire sky between 30 and 857 GHz, and has produced maps of physical emission models fitted to the data, and catalogues of compact galactic and extragalactic sources extracted from the maps. Of specific interest to this conference, Planck has produced all-sky maps of polarized emission from 30 to 353 GHz, covering the two known mechanisms of polarized continuum emission from the Interstellar Medium (synchrotron and thermal dust). The Planck data are accurately calibrated and constitute a basic reference to understand the polarized sky. In addition, the polarized compact sources detected by Planck offer a potential means to calibrate future polarized data at similar wavelengths. The Planck Legacy Archive (PLA), developed and maintained by ESA, hosts and serves all the data products of the Planck mission. The PLA provides tools that allow to manipulate maps and catalogues, operate on them and extract data sets useful for astrophysical analysis. The PLA also provides an interface to the Planck Sky Model (PSM), a simulation tool based on Planck and other data which allows to predict the sky as observed by an instrument defined by the user. The PSM has recently been upgraded to include the latest (Legacy) Planck data. The PLA is also in the process of ingesting recent data products which result from new and optimized data pipelines. In this paper, we will describe the current contents and features of the PLA and PSM which make them a useful analysis interface to Planck data. We will also seek input from the participants as to what would be useful improvements to the PLA in terms of supporting future experiments and analysis efforts.
QUIJOTE-MFI polarization maps at 11, 13, and 17 GHz are used covering the North Polar Spur to derive synchrotron spectral indices and limits on spectral curvature, by the use of T-T plots carried out simultaneously in conjunction with WMAP and Planck data. The addition of the QUIJOTE-MFI data allows an improved spectral index determination over the whole region and the comparison with dust and X-ray maps is used to explore it's nature and distance.
In order to extract information about inflationary gravitational waves using $B$-mode patterns of cosmic microwave polarization anisotropy, it is necessary to remove the foreground radiation from the Milky Way. In our previous delta-map method, the number of observation bands depends on the number of parameters of the assumed foreground model, and therefore it was difficult to improve the sensitivity by increasing the number of observation bands. Here, we extend the delta-map method so that it can be adapted to an arbitrary number of observation bands. Using parametric likelihood and realistic foreground and CMB simulations, we show that our method can increase the sensitivity of the tensor-to-scalar ratio without inducing any significant bias.
The modeling and removal of foregrounds poses a major challenge to searches for signals from inflation using the cosmic microwave background (CMB). In particular, the modeling of CMB foregrounds including various spatial averaging effects introduces multiple complications that will have to be accounted for in upcoming analyses. In this work, we introduce the generalization of the intensity moment expansion to the spin-2 field of linear polarization: the spin-moment expansion. Within this framework, moments become spin-2 objects that are directly related to the underlying spectral parameter and polarization angle distribution functions. In obtaining the required expressions for the polarization modeling, we highlight the similarities and differences with the intensity moment methods. A spinor rotation in the complex plane with frequency naturally arises from the first order moment when the signal contains both SED distortions and polarization mixing. Additional dependencies are introduced at higher order, and we demonstrate on several illustrative examples how these can be accounted for. Our new modeling of the polarized SED reveals to be a powerful tool to model the frequency dependence of the polarization angle. As such, it can be immediately applied to numerous astrophysical situations.
Both Galactic dust emission and Cosmic infrared background anisotropies (CIB) act as a dominant foreground contaminant for the measurements of the CMB anisotropies (intensity and polarization) at the frequency range above 217 GHz. The Galactic dust and CIB share a similar form of spectral energy distribution, which makes it harder to disentangle between them. An estimate of the CIB map acts as an external tracer of the CMB lensing potential, which would help detect the primordial scalar-to-tensor ratio [r] measurements through delensing. In our study, we apply the Bayesian inference technique to disentangle the dust and CIB emission using an external tracer of the dust emission. One of the primary outcomes of this study is to set the correct zero offset levels of the Planck intensity maps taking into account the pixel-dependent dust emissivities. We show that our determined offset values are compatible with the official Planck offset values up to 353 GHz and differ at frequencies above 545 GHz.
Over the past decade, observations in polarization of the microwave sky have provided increasing evidence that the spectral properties of the polarised Galactic emission are characterised by a complex spatial variability across the sky. This poses a challenge for the quest of CMB primordial B-modes calling for the development of complex component separation methods. Among all the techniques, the Needlet-ILC (NILC) has great relevance because it is a blind method that requires no prior information on the Galactic emission. However, the expected level of spatial variability of the foreground spectral properties complicates the NILC estimation and mitigation of the Galactic contamination. Therefore, in this talk we propose an innovative version of this component separation method: the Multi-Clustering NILC (MC-NILC). It performs NILC variance minimisation on separate patches of the sky (clusters) properly chosen to have similar spectral properties of the B-mode foregrounds emission within them. Clusters are identified partitioning the ratio of B-modes maps at two separate frequencies which is used as a blind tracer of the spatial distribution of the spectral indices of the Galactic emission in B-modes. We consider ratios either of foregrounds-only simulated B-modes (ideal case) or of cleaned templates of Galactic emission obtained from more realistic simulations which also include CMB signal and instrumental noise. Different clustering techniques have been explored, specifically considering an application to the future JAXA-LiteBIRD satellite mission, which targets the observation of both reionization and recombination peaks of the primordial CMB B-modes angular power spectrum with a total error on the tensor-to-scalar ratio dr < 0.001. We find that in the case of an ideal ratio MC-NILC provides a CMB solution with residual foregrounds and noise contamination that is significantly reduced with respect to standard NILC at all angular scales and much lower than the primordial signal targeted by LiteBIRD. Even with a realistic ratio, MC-NILC foregrounds residuals are much lower at the reionisation peak and comparable at the recombination bump with respect to the NILC ones permitting to reach the LiteBIRD scientific target of the tensor-to-scalar ratio. These results demonstrate that the newly introduced MC-NILC method can play an important role in the quest for primordial CMB B-modes of the next decade.
PICO is a concept for a NASA probe-scale mission aiming to observe the microwave sky for a minimum of 5~years with 21~frequency bands between 21 and 799\,GHz. We conduct simulations of map-based component separation with input tensor-to-scalar ratio $r$ values $r_{in}=0$ and $r_{in} = 0.003$, with five foreground models, and with and without delensing. For four of the five foreground models, when $r_{in} = 0$ and \alens=27\%, we find $r < 1.3 \times 10^{-4}$ to $ < 2.7 \times 10^{-4}\, (95\%)$. For these four models $r_{in}=0.003$ is recovered without bias with confidence levels between $18\sigma$ and $27\sigma$. One foreground model, called model 98, gives strongly biased result with $r_{in}=0$ and a $3\sigma$ bias with $r_{in} = 0.003$. However, we show that by analyzing many small 2.5\% patches of the sky the bias with model 98 can be identified and mitigated without any prior knowledge of the sky model. We assess the need for broad frequency coverage by eliminating few low or high frequency bands. We find weaker and more biased upper limits in both cases. Two of the four aforementioned models give peak likelihoods for $r$ between $3.8\times10^{-4}$ and $5.6\times10^{-4}$ at a $3\sigma$ significance. While a majority of the results come from a blind component separation approach and a Gaussian approximation of the likelihood with multipole information up to $\ell=200$, one model is analyzed with a parametric component separation, a Blackwell-Rao likelihood model, and multipoles $\ell \leq 12$. We show that when only low multipoles $\ell \leq 12$ are used, the non-Gaussian shape of the true likelihood gives uncertainties that are on average 30\% larger than the Gaussian approximation. This paper gives the first demonstration of map-domain iterative delensing on data that have undergone component separation. Deep measurements over large portions of the sky are necessary to avoid misinterpretation of measurements of $r$, as are possible with model 98. The $r$ constraints PICO can achieve are the strongest of any foreseeable instrument.
Detecting the imprint of inflationary gravitational waves on the B-mode polarization of the Cosmic Microwave Background (CMB) is currently one of the most compelling fundamental science cases for cosmology. Arguably the largest source of systematic uncertainty for both space and ground experiments will be contamination by the astrophysical foreground signal from the Milky Way. The separation of CMB radiation and foregrounds can be performed either in real space or in Fourier space. Different CMB experiments have followed either approach, revealing their advantages and caveats. The effects of the different approaches will become crucial at the sensitivity of the forthcoming CMB experiments (e.g. CMB-S4, LiteBIRD). In this talk I will compare different foreground cleaning methods, including a novel two-step self-consistent hybrid method. This method combines some of the advantages of both approaches, by cleaning out the spatially-constant part of the foregrounds at the map level, assuming nothing about the scale dependence of foregrounds, and modelling the residual frequency maps at power-spectrum level. We validate the method using Simons-Observatory-like simulated observations, recovering an unbiased estimate of the tensor-to-scalar ratio r for a wide range of realistic foreground scenarios.
Artificial neural networks are machine learning models which can be trained to learn non-linear behaviors from data. Since the foregrounds are known to have a strong non-linear behavior, neural networks seem to be promising models for separating the CMB signal with respect to them. The goal of this talk is to present a new methodology based on a fully convolutional neural network which segmentate the CMB signal from the foregrounds in Planck realistic simulations (Casas et. al 2022b, https://doi.org/10.1051/0004-6361/202243450). The network is formed by a set of convolutional blocks, which make inference over total sky patches at 143, 217 and 353 GHz Planck channels, and they are connected to a set of deconvolutional blocks, which output a CMB signal patch at 217 GHz. Once trained, it is validated against patches not used for training at three latitude intervals and randomly at all sky. At all sky, we compute the mean power spectrum of both true and predicted CMB maps for all the patches in the validation dataset, formed by realistic simulations not used for training, reaching a mean difference less than 50 𝜇𝑘2 for multipoles up to 𝑙 = 4000. After computing the mean power spectrum of the residual patches, we obtain a power spectrum lower than 100 𝜇𝑘2 at the latitude intervals 5° < |𝑏| < 30° and 30° < |𝑏| < 90°, and lower than 700 𝜇𝑘2 at the Galactic plane, although the network was trained randomly at all sky. Upcoming work separating the CMB signal in polarization is now taken into account.
The COSmic Monopole Observer (COSMO) is a pathfinder, ground-based experiment, designed for the detection of the isotropic y-distortion of the Cosmic Microwave Background (CMB), due to comptonization of CMB photons along the thermal history of the Universe. The instrument consists of a cryogenic Fourier Transform Spectrometer (FTS) in the Martin-Puplett configuration, which measures the difference in brightness between the sky brightness and the one of an internal blackbody reference. The interferograms are obtained by modulating the length of one of the two arms of the interferometer by ±1.5 cm, using a cryogenic, frictionless mirror transport mechanism (MTM). The resulting spectral resolution is ~ 10 GHz, in two bands covering the ranges 110-170 GHz and 200-300 GHz. Even if COSMO will operate from the Concordia station, Antartica, arguably the best site on Earth for this kind of measurements, the atmospheric emission still represents a huge limitation. A fast sky scan is obtained by means of a spinning 600 mm diameter wedge mirror, obtaining fast sky dips along a 10° diameter circle around the central elevation. In this way we are able to evaluate and subtract atmospheric emission in real time. Fast detectors are required for this strategy to be effective, so we are developing two arrays of fast, multi-mode Kinetic Inductance Detectors (KIDs), to cover the two bands. A strong limitation to spectral distortions measurements is represented by Galactic foreground emission. The spectral capabilities of COSMO will allow for efficient components separation. In particular, the absolute value of ISD brightness will be measured in different patches at Galactic latitudes ranging from ~ 10° to ~ 60°, with an angular resolution of ~ 1° providing accurate base levels for ISD maps obtained by differential instruments.
In this work we present a study to assess the potential scientific impact of bolometric interferometry in the removal of Galactic foregrounds from Cosmic Microwave Background (CMB) polarization data. The Q&U Bolometric Interferometer for Cosmology (QUBIC) is the first bolometric interferometer that aims at measuring the primordial B-mode polarization of the CMB. A Technological Demonstrator working in the 150 GHz channel will observe the sky from Alto Chorrillo, Argentina, starting from the end of 2022. Subsequently, the full instrument will be operational and observe in two frequency bands, centered at 150 GHz and 220 GHz. Bolometric interferometry is a novel technique that combines the sensitivity from bolometric detectors with the control of instrumental systematic effects from interferometry. Furthermore, a unique feature of bolometric interferometry is spectral imaging: the ability to recover the sky signal in several sub-bands within the physical band, providing a spectral resolution unattainable for a traditional imager (∆ν/ ν ~ 0.04). Specifically, the spectral imaging is totally software-based: the number of sub-bands, and therefore the spectral resolution, is chosen during data analysis. This allows us to reanalyze the data with different configurations and can help us detecting biases in our results. In this study we investigate how the increased spectral resolution provided by Bolometric Interferometry can resolve Galactic foreground complexity and provide robustness to foreground mitigation for primordial B-mode studies. For this purpose, we addressed the component separation procedure for two different experimental configurations. The first one consists of the anticipated CMB-S4 sensitivities and frequency coverage; for the second one, we extended the CMB-S4 set up with QUBIC spectral resolution (corresponding to a bolometric interferometry version of the same experiment). We generated 500 different CMB polarization and noise realizations in the case of a tensor-to-scalar ratio r = 0, combined to a power-law synchrotron emission and a thermal dust emission modelled to include frequency decorrelations. We performed a maximum likelihood component separation, assuming a power law parametrization for synchrotron and a modified blackbody for dust, and we looked for biases in the estimation of r from resulting CMB maps. Our results indicates that unaccounted dust frequency decorrelations lead to a biased result of the order of r ≃ 10--3. However, when changing the spectral resolution, the bias also changes thus hinting that the estimated r is not due to primordial CMB B-modes. Although we considered a specific dust model, this study illustrates the complementarity of Bolometric Interferometry to traditional imaging techniques for CMB polarimetry. This poster will be presented by E. Manzan on behalf of M. Regnier, S. Paradiso, L. Zapelli and the QUBIC Collaboration.
We describe Lumped Elements KID arrays suitable for balloon-borne measurements at mm and sub-mm wavelengths. These have been validated in the test flight of the OLIMPO experiment in 2018, and are designed to operate in four photometric bands centered at 150, 250, 350 and 480 GHz. After a description of the instrument, we show that in the low background conditions of the polar stratosphere, with a room temperature telescope, these detectors reached a sensitivity per-detector of the order of ~100 μK_{RJ} √s. This allows for deep multi-band mapping of the Sunyaev Zeldovich in clusters of galaxies, but also for high resolution (2 arcmin at 480 GHz) maps of interstellar dust (ISD) in selected areas, both at low and high Galactic latitudes. These data will provide a deeper understanding of the morphology and temperature distribution of ISD clouds.
In this poster, I will present the current status of the LSPE-STRIP experiment that will operate at the Observatorio del Teide. This is one of the two instruments of the Large Scale Polarization Explorer (LSPE) collaboration, which is a project devoted to the measurement of the CMB polarization in five frequency bands from 45GHz up to 240GHz, with the goal to constrain the B-mode signal from inflation down to a sensitivity in the tensor-to-scalar ratio of r=0.03 (99.7% confidence level). The second LSPE instrument is the SWIPE (Short-Wavelength Instrument for the Polarization Explorer), designed to fly on a winter arctic stratospheric long-duration balloon. STRIP will observe approximately 25% of the Northern sky using an array of 49 coherent polarimeters at 43GHz, coupled to a 1.5m fully rotating crossed-Dragone telescope, similar to the QUIJOTE telescopes. A second frequency channel with six elements at 95GHz will be exploited as an atmospheric monitor. Here I will present the current status of the project and will discuss some laboratory tests with the polarimeters.
In this talk, I will present a new, machine-learning-based technique designed to extract cosmological information from the thermal and kinetic Sunyaev-Zel'dovich (tSZ & kSZ) power spectra in the small scales of the CMB TT power spectrum. Indeed, the physically-motivated models used for the SZ spectra, based on the halo model and on an asymmetric reionisation history, respectively, are expensive to compute. With trained random forests, we are able to infer the SZ spectra from a set of cosmological and reionisation parameters. We use this method to analyse simultaneously the small- and large-scale data, namely SPT and Planck data. We show that accounting for the cosmological information enclosed in the SZ spectra in this analysis breaks the degeneracy between the two spectra amplitudes: We report a 9σ (5σ) measurement of the tSZ (kSZ) signal at l = 3000. Tighter constraints are obtained on the cosmological parameters and the tSZ scaling relation parameters. Additionally, we find the SPT data favour slightly larger optical depths, and so earlier reionisation scenarios, than Planck. However, these results are dependent on the modelling of other small-scale foregrounds and reliable constraints on cosmological parameters can only be achieved once, e.g., the cosmic infrared background (CIB) is also properly accounted for.
The thermal Sunyaev-Zeldovich (tSZ) effect is crucial for measuring cluster properties such as the pressure profile, shocks, and temperature but also for cosmology through clusters count and the proportionality of the tSZ power spectra with sigma 8 and Omega matter. Therefore, having a clean, unbiased map of the tSZ is essential. However, one of the main contaminants of the tSZ effect is the Cosmic Infrared Background (CIB). Upcoming ground-based experiments such as the Simon Observatory (SO) telescope and the Fred Young Submillimeter Telescope (FYST), both placed in the Atacama desert in Chile will yield unprecedented clean maps of the tSZ effect. With simulations of the sky as will be seen in those two experiments, we quantify the CIB-leakage into an ILC-extracted tSZ map. In particular with shows that FYST and its high-frequency coverage reduce the CIB residual noise in the recovered tSZ map by ~20% when combined with SO data compared to SO alone. And that combining SO+FYST increases the Signal-to-Noise ratio of the tSZ power spectrum by ~20% compared to SO alone.
One of the primary challenges facing future CMB polarization experiments is the modeling and removal of polarized galactic foregrounds such as thermal dust and synchrotron emission. Thermal dust emission is often modeled as a single modified blackbody. However, total emission along a given line of sight is likely to consist of a superposition of modified blackbody functions, and overly simplistic foreground models can bias measurements of the tensor-to-scalar ratio r. Future CMB experiments must employ more complex dust models while making minimal assumptions about the underlying dust physical parameters. We explore the use of a moment expansion technique to model the effects of spatial averaging of a continuous distribution of dust temperatures along the line of sight. We find that the use of this method can reduce bias while also shedding light on the distribution of dust properties. While here we focus on expanding around temperature, this technique can be extended to include a distribution in dust spectral index and can also be applied to the modeling of the synchrotron foreground.
It is known that the majority of interstellar dust is an amorphous material with no crystalline structure. The frequency dependence of the absorption coefficient of amorphous materials at sub-millimeter varies widely. Therefore, it is significant to model the amorphous dust emission with a physics motivation in order to separate the amorphous dust emission from observed signals in CMB experiments and to extract information about the amorphous dust from the observation. In this study, the emission spectrum from amorphous dust is modeled by assuming that the atoms composing the amorphous dust are trapped in a soft potential and solving the electromagnetic interaction, called the soft-potential (SP) model. The SP model is an extension of the two-level state (TLS) model, which has been proposed as an amorphous dust emission model. This presentation demonstrates the SP model, shows what kind of amorphous dust produces the what spectrum, and compares it with the TLS model. We also discuss anomalous microwave emission and the relationship between spectral indices and temperature, for which amorphous dust is proposed as one possible candidate of the origins.
We present ZodiPy, a modern and easy-to-use Python package for modeling the zodiacal emission seen by an arbitrary Solar System observer, which can be used for the removal of both thermal emission and scattered sunlight from interplanetary dust in astrophysical data. ZodiPy implements the COBE Diffuse Infrared Background Experiment (DIRBE) interplanetary dust model and the Planck extension, which allows for zodiacal emission predictions at infrared wavelengths in the 1.25–240μm range and at microwave frequencies in the 30–857 GHz range. In the near future, ZodiPy will also support the Rowan-Robinson and May (2013) model. The predicted zodiacal emission may be extrapolated to frequencies and wavelengths not covered by the built-in models to produce forecasts for future experiments. ZodiPy attempts to enable the development of new interplanetary dust models by providing the community with an easy-to-use interface for testing both current and future models. We showcase a few use-cases of ZodiPy, including simulated timestreams and binned maps corresponding to DIRBE observations and compare these to the official DIRBE calibrated individual observations.
A physical constraint which a dust grain cannot radiate photon with energy larger than its internal energy enforces a strict upper limit on the photon frequency radiated by the dust grain. We call this constraint the Internal Energy Bound (IEB). However, IEB was missed in previous studies. We have calculated the temperature distribution and SED of interstellar dust grains with the IEB for the first time. A component with a several-hundred Kelvin appeared as in the previous studies. On the other hand, we found that majority of VSGs stay at 10 K. The 10-K component makes an excess emission at the millimeter wavelength. Our result predicts that there is a correlation between the intensity of the millimeter excess and the mid-infrared excess.
We extend the ability of ForSE (Foreground Scale Extender) package, which is based on generative adversarial neural networks (GANs) and allows to simulate non-Gaussian polarized thermal dust maps. In this work we perform additional training to produce maps at arcminute angular scales along with the capacity to generate random realizations of small scales. These new realistic maps will be used to understand the impact of non-Gaussian foregrounds on the measurements of the cosmic microwave background (CMB) signal, in particularly on the lensing reconstruction, delensing and the detection of primordial CMB B-modes. With the input of the large-scale polarization maps from observation, ForSE is trained to produce realistic polarized small scales following the statistical properties, mainly the non-Gaussianity, of observed intensity small scales, which are evaluated through Minkowski functionals. Furthermore, by adding different realization of random noise to the large-scale foregrounds, we expect ForSE to be able to generate small scales in a stochastic way. In this study, we demonstrated the performance of ForSE to generate small scales at 3’ from polarization maps at 80’ and to generate stochastic small scales from different levels of noise on top of large-scale signals. These results will be integrated into PySM3, which is the next version of the commonly used foregrounds simulation package.
A parity-violating phenomena in the early universe creates a correlation between E-mode and B-mode signals in CMB though in the standard cosmology, there is no origin that creates EB-mode spectrum. We expect that the Sparse Wire Grid Calibrator in SO enables us to not only the primordial B-mode but non-zero EB-mode spectrum if it exits. In this poster presentation, I will introduce the development of the calibrator for the observation of primordial B-mode and EB-mode polarization signals.
In this work we compare the performance of bolometric interferometry (BI) and standard imagers in the separation of astrophysical foregrounds from CMB polarization measurements aimed at the detection of primordial B-modes. Bolometric Interferometry (BI) is a measurement strategy, complementary to direct sky imaging, which deals more effectively foreground cleaning thanks to its ability to resolve sub-frequency bands in the main frequency interval. This ability has been called “spectral imaging”. The Q&U Bolometric Interferometer for Cosmology (QUBIC) is polarimeter based on bolometric interferometry that has been recently installed in the high altitude (5000 m) site of Alto Chorillo, Argentina. The first QUBIC prototype (named Technological Demonstrator, TD) will start observing the sky at 150 GHz during early 2023; between 2023 and 2024 the final instrument (named Full Instrument, FI) will be operating at 150 and 220 GHz. Being a bolometric interferometer, QUBIC has spectral imaging capabilities thanks to the frequency-dependent shape of its synthesized beam. In our work we have studied the impact of spectral imaging in foreground analysis by exploring the likelihood of the B-modes power spectrum using the Bayesian component separation code “Commander”. A companion poster describing a similar work carried out with the parametric code FGBuster has been proposed for this meeting by E. Manzan et al. Our work allowed us to perform this analysis with a comparative and independent methodology, still relying on a Bayesian approach. Furthermore, Commander allows one to draw samples of the model parameters from the joint posterior distribution of the CMB and foreground amplitudes and their spectral indices in order to characterize the full posterior.
It has been pointed out that the spurious cosmic microwave background (CMB) B-mode polarization signals caused by the absorption of the CMB monopole component by the Galactic interstellar matter, called the CMB shadow, degrade the accuracy of extracting the CMB B-mode polarization imprinted by primordial gravitational waves. We have made realistic estimation using simulated maps of how the CMB shadow affects forthcoming high-precision CMB B-mode experiments at the first time. The Delta-map method, an internal template method taking into account first order spatial variation of foregrounds' spectral parameters, is applied as a foreground removal method. Then, we estimate tensor-to-scalar ratio r and sky-averaged foreground parameters by maximum likelihood estimation. We have shown that the false detection of the CMB B-mode polarization originating from primordial gravitational waves is led when the existence of the CMB shadow effect was missed. We have also shown that the effect of the CMB shadow is able to be mitigated by our revised Delta-map method and we are able to target r=0.001.
The analysis of the Cosmic Microwave Background (CMB) is a powerful tool to explore the primitive Universe. The detection of primordial B-modes is an important milestone in cosmology, because it would confirm the presence of gravitational waves in the primordial Universe generated in the inflationary epoch. Up to now, the B-modes induced by primordial gravitational waves remain undetected. In this context, the foreground emission of the Galaxy is a dominant element that limits the determination of the signal of interest. Therefore, a comprehensive characterization of the galactic and extragalactic foregrounds is essential to clean and recover the primordial B-mode signal out of the CMB observations. In this contribution, we present the implementation and testing of the HIT Fitting method (Hybrid Internal combination with Template Fitting). This new approach is based on a combination of the internal linear combination (ILC) and internal template fitting (ITF) methods, which is devoted to recovering the CMB signal. In particular, we discuss their bias properties using multifrequency simulations of CMB experiments.
Planck satellite left a great legacy in terms of constraints on properties of Galactic dust, thanks to the full-sky high-frequency coverage of its submillimetre emission. The aim of this project is to study the correlation between the thermal dust polarization from Planck PR4 data with the starlight polarization data in the visible at high Galactic latitudes, also extending the result published by Planck Collaboration in 2018 at 353 GHz (Planck 2018 results XII), up to smaller frequencies: 217 and 143 GHz. It is known that dust grains that emit a submillimetre polarized radiation also extinguish and polarize starlight in the visible, and a comparison of both in selected lines of sight (los) provides important diagnostics of properties of dust, and therefore strong constraints for models of Galactic dust in the diffuse interstellar medium (ISM). We correlated properties in the submillimetre and in the visible for those los through the diffuse ISM and at high Galactic latitudes with polarization directions close to orthogonal and comparable values of the estimated column density, finding the polarization emission-to-extinction ratio RP/p = PS /pV. It is a diagnostic which focus directly on the polarization properties of the aligned grain population, defined as the ratio between the polarized emission and the visible degree of polarization. Regarding the optical data, we use the same set of Berdyugin catalogs of interstellar polarization and extinction of starlight presented in Planck 2018 results XII. As regards the emission data, we decided to use a new approach, applying the photometric aperture method which consists in taking the average values of Q and U Planck 2020 data release (PR4) of a circular aperture, centered in the star’s position. Moreover, we studied the behavior of the emission-to-extinction polarization ratio with respect to a range of aperture’s radii, ranging from 5 arcminutes up to 1 degree, finding that the minimum residuals with respect to the expected values are in correspondence of a radii’s range between 15-16 arcminutes, 22-48 arcminutes and 45-47 arcminutes for 353, 217 and 143 GHz channels, respectively. The final estimates of the emission-to-extinction polarization ratios we found are: RP/p(353GHz) = 5.57 ±0.05 MJy/sr; RP/p(217GHz) = 1.19 ±0.01 MJy/sr; RP/p(143GHz) = 0.300 ±0.004 MJy/sr. These new results of the diagnostic RP/p are slightly higher than the one found in the Planck 2018 results XII at 353 GHz, and than the other ratios at 217 and 143 GHz that we can find scaling with a dust model law. This indicates that changes to the optical properties in the models of the aligned grain population could be required.