Periodic Reporting for period 1 - CMB-INFLATE (Advanced Methodologies for Next Generation Large Scale CMB Polarization Analysis)
Período documentado: 2021-10-01 hasta 2024-08-31
Around 10^-37 seconds after the Big-Bang, the Universe went through a rapid expansion phase called inflation with extreme density and temperature conditions such that the known laws of physics do not apply and require a unified theory of quantum gravity. The best probe of this period is a relic emission called Cosmic Microwave Background (CMB) from 380000 years after the big-bang, when the Universe state changed from a hot and opaque plasma to a transparent phase. This emission shows small fluctuations which are the seeds of structures we observe today, and inflation provides a mechanism for the generation of those perturbations: they are quantum fluctuations brought to cosmic scales. This process generates density perturbations measured with high precision experiments, in particular the Planck satellite of ESA. It also produces a background of gravitational waves. The CMB emission is polarized and primordial gravitational waves produce a parity-odd CMB polarization pattern on the sky called B-modes, as illustrated by Figure 1, undetected as of today. The positive parity patterns called E-modes are generated by both types of perturbations. Full sky polarization B-modes varying at roughly 20-degree scales on the sky contains precious information about the inflation physics and can uniquely be probed with future satellite mission such as LiteBIRD of the Japanese space agency JAXA.
This is a challenging measurement because of the presence of polarized foreground emissions at frequencies of the CMB emission: polarized thermal Galactic dust and synchrotron emissions. Our capacity to subtract those foreground emissions is one of the current limitations hampering the accurate measurement of large angular scales. Improved modelling of foreground emission, as well as the development of analysis methods incorporating all the complexity of the sky and the instrument are crucial steps for an accurate determination of the primordial signal from inflation with future experiments.
Another difficulty is the understanding and mitigation of instrumental effects in data. The success of future experiments will require accurate modelling of those effects in close link with instrumental developments that are now performed over the world. It will also require the development of new innovative analysis techniques to estimate the tiny gravitational wave signal. Analysis techniques will require mathematical developments to reduce efficiently a large volume of data.
The main objective of CMB-Inflate is to improve communications and foster collaborations within a community of researchers for advanced and innovative analysis of CMB polarization data focused on the large angular scales (10 degrees and more) with the ultimate goal of measuring in a robust way the tensor modes and the fundamental parameters that describe them to distinguish between the different scenarios of inflation. The network is gathering scientists with diverse expertise and backgrounds from instrumentalists to theoreticians in order to reach the next steps in CMB polarization data analysis consisting in providing a complete and advanced modelling of the data including instrumental effects and all the complexity of the sky emission; and in the development of global analysis of the data, from the raw time-streams to the sky component separation and likelihood analysis for the extraction of science.
Moreover, the training of the young community of researchers are an important objective of the CMB-Inflate project.
This is a challenging measurement because of the presence of polarized foreground emissions at frequencies of the CMB emission: polarized thermal Galactic dust and synchrotron emissions. Our capacity to subtract those foreground emissions is one of the current limitations hampering the accurate measurement of large angular scales. Improved modelling of foreground emission, as well as the development of analysis methods incorporating all the complexity of the sky and the instrument are crucial steps for an accurate determination of the primordial signal from inflation with future experiments.
Another difficulty is the understanding and mitigation of instrumental effects in data. The success of future experiments will require accurate modelling of those effects in close link with instrumental developments that are now performed over the world. It will also require the development of new innovative analysis techniques to estimate the tiny gravitational wave signal. Analysis techniques will require mathematical developments to reduce efficiently a large volume of data.
The main objective of CMB-Inflate is to improve communications and foster collaborations within a community of researchers for advanced and innovative analysis of CMB polarization data focused on the large angular scales (10 degrees and more) with the ultimate goal of measuring in a robust way the tensor modes and the fundamental parameters that describe them to distinguish between the different scenarios of inflation. The network is gathering scientists with diverse expertise and backgrounds from instrumentalists to theoreticians in order to reach the next steps in CMB polarization data analysis consisting in providing a complete and advanced modelling of the data including instrumental effects and all the complexity of the sky emission; and in the development of global analysis of the data, from the raw time-streams to the sky component separation and likelihood analysis for the extraction of science.
Moreover, the training of the young community of researchers are an important objective of the CMB-Inflate project.
After two years of the project, work has been carried out on many fronts. Regarding our first objective, the modeling of instrument components, we have developed a general and innovative model of a half wave plate (HWP), a key optical device to measure the CMB polarization and envisaged by the future satellite mission LiteBIRD. The model, available on a repository, predicts the signal produced by HWP imperfections for different incidence angle and observation frequencies. It is based on calculations of electromagnetic wave propagation through a multi-layer sapphire and anti-reflection coating on the surface.
Moreover, we developed optical model for a satellite mission to accurately account for features producing artefacts in data such as the co- and cross-polarization components of the optical response (called beam) in presence of an HWP. The prediction of the coupling of the optical response with a moving component, the HWP, is the originality of the project.
We also addressed the problem of modelling transition edge sensor (TES) responses using as a baseline the LiteBIRD satellite configuration and we have started a study of the coupling with other components such as the focal plane and the HWP. Emphases have been put on the modeling of non-linear response and the coupling with the HWP signal.
Regarding the second objective: the development of innovative data analysis methods, we have developed new and efficient map-making methods handling systematics, in particular HWP, beam and gain systematics, using models of the instrument components developed within the project. Map-making is an important data analysis step for the CMB measurement, reducing the data from measurements acquired continuously in time by each individual detector to the pixelized maps of the estimated sky signal, also removing instrument effects. One aspect of the work concerned the generalization of the map-making method for the inclusion of other maps than the sky temperature and polarization capturing specific systematics effects: band-pass mismatches, pointing errors, gain errors and HWP spurious systematics. Other methods developed include: joint estimation of maps and instrument parameters (HWP in particular) in a maximum likelihood approach and a Wiener filtering approach.
We have developed astrophysical component separation methods adapted for CMB B-mode searches in light of the most recent foreground emission modeling. Hybrid (or semi-blind) methods are under development and more flexible blind method have been tested in presence of instrumental effects modeled in the project. The inclusion of such effects is the originality of CMB-Inflate. Consequent work has taken place for the comparison of methodologies.
Two simulations frameworks including instrumental and foreground emission models developed in the project were put in place with the objectives of testing our models and methodologies. Instrumental effects that are being integrated in the simulation frameworks include: optical effect convolution; HWP Mueller matrix; noise models and detector response including the interaction with the focal plane.
The last aspect concerns the global end-to-end analysis. Work is performed in synergy with the Cosmoglobe and BeyondPlanck EU funded projects: an analysis pipeline for LiteBIRD is put in place; young researchers participating to the project have been trained for high-level end-to-end data analysis during several secondments; advanced likelihood techniques to estimate science parameters have been developed and tested with our data model.
Moreover, we developed optical model for a satellite mission to accurately account for features producing artefacts in data such as the co- and cross-polarization components of the optical response (called beam) in presence of an HWP. The prediction of the coupling of the optical response with a moving component, the HWP, is the originality of the project.
We also addressed the problem of modelling transition edge sensor (TES) responses using as a baseline the LiteBIRD satellite configuration and we have started a study of the coupling with other components such as the focal plane and the HWP. Emphases have been put on the modeling of non-linear response and the coupling with the HWP signal.
Regarding the second objective: the development of innovative data analysis methods, we have developed new and efficient map-making methods handling systematics, in particular HWP, beam and gain systematics, using models of the instrument components developed within the project. Map-making is an important data analysis step for the CMB measurement, reducing the data from measurements acquired continuously in time by each individual detector to the pixelized maps of the estimated sky signal, also removing instrument effects. One aspect of the work concerned the generalization of the map-making method for the inclusion of other maps than the sky temperature and polarization capturing specific systematics effects: band-pass mismatches, pointing errors, gain errors and HWP spurious systematics. Other methods developed include: joint estimation of maps and instrument parameters (HWP in particular) in a maximum likelihood approach and a Wiener filtering approach.
We have developed astrophysical component separation methods adapted for CMB B-mode searches in light of the most recent foreground emission modeling. Hybrid (or semi-blind) methods are under development and more flexible blind method have been tested in presence of instrumental effects modeled in the project. The inclusion of such effects is the originality of CMB-Inflate. Consequent work has taken place for the comparison of methodologies.
Two simulations frameworks including instrumental and foreground emission models developed in the project were put in place with the objectives of testing our models and methodologies. Instrumental effects that are being integrated in the simulation frameworks include: optical effect convolution; HWP Mueller matrix; noise models and detector response including the interaction with the focal plane.
The last aspect concerns the global end-to-end analysis. Work is performed in synergy with the Cosmoglobe and BeyondPlanck EU funded projects: an analysis pipeline for LiteBIRD is put in place; young researchers participating to the project have been trained for high-level end-to-end data analysis during several secondments; advanced likelihood techniques to estimate science parameters have been developed and tested with our data model.
Several developments in CMB-Inflate are beyond the start of the art in the field. They include:
- Complete model and characterization of a HWP and the its impact on the optical response of a satellite
- New map-making methods including instrument effect processing
- New hybrid component separation methods
- Complete model and characterization of a HWP and the its impact on the optical response of a satellite
- New map-making methods including instrument effect processing
- New hybrid component separation methods