PRISM (Polarized Radiation Imaging and Spectroscopy Mission): A White Paper on the Ultimate Polarimetric Spectro-Imaging of the Microwave and Far-Infrared Sky
PRISM Collaboration, Philippe Andre, Carlo Baccigalupi, Domingos Barbosa, James Bartlett, Nicola Bartolo, Elia Battistelli, Richard Battye, George Bendo, Jean-Philippe Bernard, Marco Bersanelli, Matthieu Bethermin, Pawel Bielewicz, Anna Bonaldi, Francois Bouchet, Francois Boulanger, Jan Brand, Martin Bucher, Carlo Burigana, Zhen-Yi Cai, Viviana Casasola, Guillaume Castex, Anthony Challinor, Jens Chluba, Sergio Colafrancesco, Francesco Cuttaia, Giuseppe D'Alessandro, Richard Davis, Miguel de Avillez, Paolo de Bernardis, Marco de Petris, Adriano de Rosa, Gianfranco de Zotti, Jacques Delabrouille, Clive Dickinson, Jose Maria Diego, Edith Falgarone, Pedro Ferreira, Katia Ferriere, Fabio Finelli, Andrew Fletcher, Gary Fuller, Silvia Galli, Ken Ganga, Juan Garcia-Bellido, Adnan Ghribi, Joaquin Gonzalez-Nuevo, Keith Grainge, Alessandro Gruppuso, Alex Hall, Carlos Hernandez-Monteagudo, Mark Jackson, Andrew Jaffe, Rishi Khatri, Luca Lamagna, Massimiliano Lattanzi, Paddy Leahy, Michele Liguori, Elisabetta Liuzzo, Marcos Lopez-Caniego, Juan Macias-Perez, Bruno Maffei, Davide Maino, Silvia Masi, Anna Mangilli, Marcella Massardi, Sabino Matarrese, Alessandro Melchiorri, Jean-Baptiste Melin, Aniello Mennella, Arturo Mignano, Marc-Antoine Miville-Deschenes, Federico Nati, Paolo Natoli, Mattia Negrello, Fabio Noviello, Francesco Paci, Rosita Paladino, Daniela Paoletti, Francesca Perrotta, Francesco Piacentini, Michel Piat, Lucio Piccirillo, Giampaolo Pisano, Gianluca Polenta, Sara Ricciardi, Matthieu Roman, Jose-Alberto Rubino-Martin, Maria Salatino, Alessandro Schillaci, Paul Shellard, Joseph Silk, Radek Stompor, Rashid Sunyaev, Andrea Tartari, Luca Terenzi, Luigi Toffolatti, Maurizio Tomasi, Tiziana Trombetti, Marco Tucci, et al. (4 additional authors not shown)
PPolarized Radiation Imaging and Spectroscopy Mission
Spokesperson:
Paolo de Bernardis e-mail: [email protected] — tel: + 39 064 991 4271
PRISM
Probing cosmic structures and radiation with the ultimate polarimetric spectro-imaging of the microwave and far-infrared sky a r X i v : . [ a s t r o - ph . C O ] J un uthors and contributors Philippe Andr´e, Carlo Baccigalupi, Domingos Barbosa, James Bartlett, Nicola Bartolo, Elia Battistelli,Richard Battye, George Bendo, Jean-Philippe Bernard, Marco Bersanelli, Matthieu B´ethermin, Pawel Bielewicz,Anna Bonaldi, Fran¸cois Bouchet, Fran¸cois Boulanger, Jan Brand, Martin Bucher, Carlo Burigana, Zhen-YiCai, Viviana Casasola, Guillaume Castex, Anthony Challinor, Jens Chluba, Sergio Colafrancesco, FrancescoCuttaia, Giuseppe D’Alessandro, Richard Davis, Miguel de Avillez, Paolo de Bernardis, Marco de Petris,Adriano de Rosa, Gianfranco de Zotti, Jacques Delabrouille, Clive Dickinson, Jose Maria Diego, Edith Fal-garone, Pedro Ferreira, Katia Ferri`ere, Fabio Finelli, Andrew Fletcher, Gary Fuller, Silvia Galli, Ken Ganga,Juan Garc´ıa-Bellido, Adnan Ghribi, Joaquin Gonzalez-Nuevo, Keith Grainge, Alessandro Gruppuso, AlexHall, Carlos Hernandez-Monteagudo, Mark Jackson, Andrew Jaffe, Rishi Khatri, Luca Lamagna, Massimil-iano Lattanzi, Paddy Leahy, Michele Liguori, Elisabetta Liuzzo, Marcos Lopez-Caniego, Juan Macias-Perez,Bruno Maffei, Davide Maino, Silvia Masi, Anna Mangilli, Marcella Massardi, Sabino Matarrese, AlessandroMelchiorri, Jean-Baptiste Melin, Aniello Mennella, Arturo Mignano, Marc-Antoine Miville-Deschˆenes, Fed-erico Nati, Paolo Natoli, Mattia Negrello, Fabio Noviello, Francesco Paci, Rosita Paladino, Daniela Paoletti,Francesca Perrotta, Francesco Piacentini, Michel Piat, Lucio Piccirillo, Giampaolo Pisano, Gianluca Po-lenta, Sara Ricciardi, Matthieu Roman, Jose-Alberto Rubino-Martin, Maria Salatino, Alessandro Schillaci,Paul Shellard, Joseph Silk, Radek Stompor, Rashid Sunyaev, Andrea Tartari, Luca Terenzi, Luigi Toffolatti,Maurizio Tomasi, Tiziana Trombetti, Marco Tucci, Bartjan Van Tent, Licia Verde, Ben Wandelt, StaffordWithington
Coordination group
The preparation of this science case submitted to ESA has been coordinated by:
James Bartlett, Fran¸cois Bouchet, Fran¸cois Boulanger, Martin Bucher, Anthony Challinor, Jens Chluba,Paolo de Bernardis (spokesperson), Gianfranco de Zotti, Jacques Delabrouille (coordinator), Pedro Ferreira,Bruno Maffei
Supporters
As a result of the highly successful ESA Planck and Herschel missions, Europe has acquired considerablescientific and technical expertise in the scientifically fruitful and strategic field of microwave and far-infraredobservations and trained a new generation of young European astronomers in this area. The science themesoutlined in this proposal are the logical next step that will allow ESA to capitalize on these strengths.To demonstrate the breadth of support for PRISM, we are in the process of assembling a list of supportersthat can be found at the following website:
Scientists who believe that ESA should pursue as part of its programme the science themes presented in thisWhite Paper are strongly encouraged to visit the website and to sign on as supporters.
PRISM Steering committee
France:
Fran¸cois Bouchet, Martin Bucher, Jacques Delabrouille, Martin Giard
Germany:
Jens Chluba, Rashid Sunyaev
Ireland:
Anthony Murphy
Italy:
Marco Bersanelli, Carlo Burigana, Paolo de Bernardis
Netherlands:
Rien van de Weijgaert
Portugal:
Carlos Martins
Spain:
Enrique Mart´ınez-Gonz´alez, Jos´e Alberto Rubi˜no-Mart´ın, Licia Verde
Switzerland:
Martin Kunz
United Kingdom:
Anthony Challinor, Joanna Dunkley, Bruno MaffeiThis is a corrected version (10 June 2013) of the original document submitted on 24 May 2013 to ESA inresponse to
Call for White Papers for the definition of the L2 and L3 Missions in the ESA Science Programme ( http://sci.esa.int/Call-WP-L2L3 ) Executive summary
PRISM is a large-class mission that will carry out the ultimate survey of the microwave to far-infrared skyin both intensity and polarization as well as measure its absolute spectrum.
PRISM will consist of twoinstruments: (1) a high angular resolution polarimetric imager with a 3.5 m telescope cooled to around 4K toreduce thermal noise, particularly in the far-infrared bands; and (2) a low angular resolution spectrometer tocompare the sky frequency spectrum to a nearly perfect reference blackbody. The two instruments working intandem will enable
PRISM to make breakthrough contributions by answering key questions in many diverseareas of astrophysics and fundamental science. A few highlights of the new science with
PRISM include: (A) The ultimate galaxy cluster survey:
The Sunyaev-Zeldovich (SZ) effect is the method of choicefor assembling a catalog of clusters at high redshift, of particular interest for cosmology because of the tightcorrelation between integrated y -distortion and cluster mass. When PRISM flies, all-sky cluster samples (e.g.,from eROSITA, Euclid ) will likely count some 10 objects, mostly at z < . PRISM will find 10 times moreclusters extending to deeper redshifts, with many thousands beyond z = 2 . In fact,
PRISM will detect allclusters in the universe of mass larger than 10 M (cid:12) , and a large fraction of those with mass above 5 × M (cid:12) .Owing to its exquisite spectral coverage, angular resolution and sensitivity, PRISM will measure the peculiarvelocity of hundreds of thousands of clusters using the kinetic SZ effect, initiating a new research area: thecomplete mapping of the large-scale velocity field throughout our Hubble volume. In addition,
PRISM willalso be able to probe the relativistic corrections to the classic SZ spectral distortion spectrum, thus measuringthe gas temperature. This cluster sample will allow us to probe dark energy and better understand structureformation at large redshift. (B) Understanding the Cosmic Infrared Background:
Most star formation in the universe took placeat high redshift. Hidden from optical observations by shrouds of dust in distant galaxies, it is visible only inthe far infrared or in X-rays. Emission from these dusty galaxies constitutes the cosmic infrared background(CIB) which
PRISM , owing to its high sensitivity and angular resolution in the far infrared, is uniquelysituated to investigate. The survey will sharpen and extend to higher redshifts the determination of thebolometric luminosity function and the clustering properties of star-forming galaxies. Tens of thousandsof easily recognizable, bright, strongly lensed galaxies and hundreds of the very rare maximum starburstgalaxies, up to z >
6, will be detected, providing unique information on the history of star formation, thephysics of the interstellar medium in a variety of conditions up to the most extreme, and the growth of largescale structure, including proto-clusters of star-forming galaxies. The survey will also probe the evolution ofradio sources at (sub-)mm wavelengths and provide measurements of the spectral energy distribution (SED)of many thousands of radio sources over a poorly explored, but crucial frequency range. (C) Detecting inflationary gravity waves:
Present precision measurements of cosmic microwave back-ground (CMB) temperature anisotropies lend considerable support to simple models of inflation. Howeverthe most spectacular prediction of inflation—the generation of gravitational waves with wavelengths as largeas our present horizon—remains unconfirmed. Several initiatives from the ground and from stratospheric bal-loons are currently underway to attempt to detect these gravitational waves through the B-mode spectrumof the CMB polarization. However, they suffer from severe handicaps such as limited frequency coverage dueto atmospheric opacity, unstable seeing conditions, and far sidelobes from the ground. It is only from spacethat one may hope to detect the very low- (cid:96)
B-modes due to the re-ionization bump. Because of its broadfrequency coverage and extreme stability,
PRISM will be able to detect B-modes at 5 σ for r = 5 × − , evenunder pessimistic assumptions concerning the complexity of the astrophysical foreground emissions that mustbe reliably removed. Moreover, PRISM will be able to separate and filter out the majority of the lensingsignal due to gravitational deflections. (D) Probe new physics through CMB spectral distortions:
The excellent agreement between themicrowave sky emission and a perfect blackbody observed by the
COBE FIRAS instrument is rightfullyhighlighted as a crucial confirmation of Big Bang cosmology. However theory predicts that at higher sensitivitythis agreement breaks down. Some of the predicted deviations are nearly sure bets. Others provide powerfulprobes of possible new physics. The
PRISM absolute spectrometer will measure the spectrum more thanthree orders of magnitude better than
FIRAS. y -distortions from the re-ionized gas as well as from hotclusters constitute a certain detection. However µ -distortions and more general spectral distortions have thepotential to uncover decaying dark matter and to probe the primordial power spectrum on very small scalesthat cannot be measured by other means, being contaminated by the nonlinearity of gravitational clusteringat late times. 1 E) Probe Galactic astrophysics:
PRISM will have a major impact on Galactic astrophysics by providinga unique set of all-sky maps.
PRISM will extend Herschel dust observations to the whole sky and will mapemission lines key to quantifying physical processes. The survey will have the sensitivity and angular resolutionrequired to map dust polarization down to sub-arcminute resolution even at the Galactic poles. No projectwill provide a comparable perspective on interstellar components over such a wide range of scales. The
PRISM data will provide unique clues to study the interstellar medium, the Galactic magnetic field, and starformation, and will address three fundamental questions in Galactic astrophysics: What are the processesthat structure the interstellar medium? What role does the magnetic field play in star formation? What arethe processes that determine the composition and evolution of interstellar dust?These are but a few of the highlights of the rich and diverse physics and astrophysics that
PRISM will beable to carry out.
The hundreds of intensity and polarization maps of
PRISM will constitute a legacy archive useful for almost allbranches of astronomy for decades to come. Combining low resolution spectrometer data and high resolutionimages from the imager,
PRISM will deliver a full spectro-polarimetric survey of the full sky from 50 µ m to1 cm. The spectral resolution will range from about 0.5 GHz to 15 GHz at 1.4 ◦ angular resolution, and from δν/ν ≈ .
025 to 0 .
25 at the diffraction limit of a 3.5 m telescope (from ∼ (cid:48)(cid:48) to 17 (cid:48) ).We will make public full-sky maps of the absolute temperature of the CMB and of its polarization (ata resolution of about 2 arcminutes with a sensitivity of order 1 µ K or better per resolution element), of theemission of all galactic components in absolute intensity and polarization (including main spectral lines), andseveral catalogues of various galactic and extragalactic objects, among which a catalogue of about a milliongalaxy clusters and large groups up to redshift z = 3 or more. Figure 1:
Lower mass limits for detection of the in-dicated SZ effects at signal-to-noise
S/N >
The
PRISM mission will exploit the advantages of clustersurveying using the SZ effect in a spectacular way, sur-passing in depth any planned cluster survey and, in addi-tion, achieving an objective unattainable in any other way:measurement of the cosmic velocity field throughout theobservable universe. In short, we will detect cluster andgroup systems throughout our Hubble volume from themoment when they first emerge.
PRISM will also providecluster mass determinations out to high redshift throughgravitational lensing of the CMB in both temperature [96]and polarization [66], something only possible because of
PRISM ’s high angular resolution and frequency coverageextending into bands unreachable from the ground. The
Planck , ACT, and SPT experiments demonstrated the po-tential of the Sunyaev-Zeldovich effect for studying galaxyclusters and using them to constrain cosmological models.
PRISM will transform SZ cluster studies into arguably our most powerful probe of cosmic large-scale structureand its evolution.
The Cluster Catalog and its Applications:
We estimate the content of the
PRISM
SZ catalog by applyinga multi-frequency matched filter [70] to simulations of a typical field at intermediate Galactic latitude. Ourdetection mass remains below 10 solar masses at all redshifts (Fig. 1). Extrapolating from the observed Planck counts, we predict nearly 10 clusters with many thousands at z >
2. We already know from
Planck
SZ observations [79, 80] that the SZ signal in clusters scales as our adopted relation down to much smallermasses in the local universe, leaving as our main uncertainty poor knowledge of its redshift dependence. Thisis, of course, the primary motivation for studying the high redshift cluster population.Based on this calculation,
PRISM will surpass all current and planned cluster surveys, including eROSITA and Euclid—not just in total numbers, but most importantly in numbers of objects at z > .
5. Cluster2dentification will be vastly more robust for
PRISM than Euclid, which will suffer from the much highercontamination rate of optical/NIR cluster searches, especially at redshifts beyond unity. In all cases, only
PRISM has the ability to find significant numbers of clusters in the range 2 < z <
3, the critical epochthat current observations identify as the emergence of the characteristic cluster galaxy population on the redsequence.
PRISM will also enable us to explore the abundance of the intra-cluster medium (ICM) throughthe Y – M relation and its relation to the galaxy population at these high redshifts.At the time of operation, large imaging (e.g., DES, LSST, HSC ) and spectroscopic surveys (e.g., ) will have covered the entire extragalactic sky. We will easily beable to obtain redshifts, spectroscopic or photometric, for all objects to z = 2, and the two micron cutoffof Euclid’s IR photometric survey (H band) will be sufficient to detect the 4000˚ A break in brighter clustergalaxies at higher redshifts. Catalog as a cosmological probe:
As an example of the cosmology constraints that can be obtainedfrom the expected cluster catalog, we performed a standard Fisher analysis to constrain four parameters,Ω m , σ and the dark energy equation-of-state parameters w and w , in a standard flat ΛCDM cosmologicalmodel. The constraints on the latter dark energy parameters are w = − ± .
003 and w = 0 ± . PRISM cluster catalog as a dark energy probe.
Cosmic velocity field:
PRISM will initiate an untapped research area: study of the velocity field throughthe kinetic SZ effect [102, 90, 8], an independent probe of dark matter and large-scale structure evolution. InFig. 1 we show mass limit to which we expect to measure a velocity of 300 km s − to five sigma on individualclusters. This mass limit means that we will obtain velocity measurements for hundreds of thousands ofclusters out to the highest redshifts. In addition, by comparing measured velocities to mass concentrations,say from Euclid lensing or galaxy surveys, we can test the theory of gravity on cosmic scales and to highredshift. This science is unattainable by any other means. Relativistic and non-thermal effects:
We will determine the temperature of clusters down to a masslimit just above 10 solar masses by measuring the relativistic corrections to the thermal SZ spectrum[89, 15, 55, 95, 8]. These same characteristics allow us to search for non-thermal signatures in the spectrathat could signal the presence of highly energetic particles, perhaps dark matter annihilation products, andeven study the temperature structure of the most massive systems. Diffuse SZ and the cosmic web:
The diffuse, unresolved SZ effect probes a different mass and redshiftrange than observations of individually detected objects. We will study this diffuse effect through the powerspectrum and higher order moments of an SZ map of the sky.
Planck recently extracted the first Comptonparameter ( y -fluctuation) map [85], but the results are limited by foregrounds and noise. With many morespectral bands and much better sensitivity and resolution, PRISM will significantly improve the results,making possible attempts to directly map the cosmic web (i.e., its filaments) over large scales through itsdiffuse gas content.We will explore the gas content of dark matter halos down to very low masses, a research area pioneeredby
Planck by stacking SZ measurements based on known objects to detect the signal down below 10 solarmasses [79, 80]. The measurement over such a vast range is unique to the SZ effect and a highly valuableconstraint on the mysteries of the feedback mechanisms at the heart of galaxy formation. PRISM greatlyexpands this important science area by pushing to the lowest possible masses and by probing gas content asa function of object properties. Coupled with our lensing measurements, we have a new and exceptional toolto study the relation between luminous and dark matter.
Polarized SZ effect:
PRISM will enable searches for the polarized SZ effects, giving access to transversecluster velocities and measurements of the CMB quadrupole at distant locations.
Early evolution of galaxies:
Although
Herschel and
Spitzer made spectacular advances in our under-standing of early, dust enshrouded phases of galaxy evolution, our knowledge of star-formation history inthe distant universe is still very incomplete. The
PRISM mission will make essential progress thanks to itsunique properties: full sky coverage and unparalleled frequency range. As illustrated in Fig. 2,
PRISM ’s un-precedented frequency coverage provides direct measurements of the bolometric luminosities of star-forminggalaxies up to high redshifts. At z > ∼
2, i.e., in the redshift range where both the cosmic star formation and3igure 2:
SEDs of dusty galaxies (top panel) and of AGNs (bottompanel) at different redshifts compared with estimated 5 σ detection limits(solid black line) taking into account instrumental and confusion noisesummed in quadrature. The instrumental noise refers to the full mission.The 5 σ detection limits allowing for either component are shown by thedotted and the dashed black lines, showing that PRISM is confusionlimited above ≈
150 GHz. We have assumed that component separationtechniques, extensively validated both on simulations and on real data,can efficiently remove diffuse emissions such as the CMB (that wouldotherwise dominate the fluctuation field for ν < ∼
220 GHz) and Galacticemissions. In the top panel, at z = 0 . µ m) luminosity of 10 L (cid:12) . At z ≥ z ≈ . L IR = 10 L (cid:12) for z = 1 and z = 2, and to L IR = 3 · L (cid:12) [the luminosity of the z = 6 .
34 galaxy detected by
Herschel /SPIRE, 91]for z ≥
3. In the bottom panel, the solid colored lines represent SEDs ofa type-2 QSO (contribution of the host-galaxy subtracted) with L IR =10 L (cid:12) at several redshifts ≥
2, while the dashed colored lines showa schematic representation of the SED of the prototype blazar 3C 273shifted to redshifts from 1 to 5. the accretion rate onto supermassive black-holes are maximum, both the IR peak associated with the dustytorus around AGN ( λ p , AGN ≈ × (1 + z ) µ m) and the peak of dust emission in the host galaxy are withinthe covered range. Moreover, measurements of the complete far-IR to mm-wave SED will vastly improve theaccuracy of photometric redshift estimates that have a rms error of ≈ . z ) with SPIRE alone [65]. Thismeans that the PRISM survey will allow us to characterize to high statistical precision the evolution withredshift of the bolometric luminosity function. At z > ∼ Herschel /SPIRE will greatly improve identification of reliablecounterparts in other wavebands, necessary for a comprehensive understanding of the properties of detectedgalaxies.Its all-sky coverage makes
PRISM uniquely suited to study rare phenomena. Examples are the ‘maximumstarburst’ galaxy at z = 6 .
34, detected by
Herschel /SPIRE [91], or the most luminous star-forming hyper-luminous IR galaxies, such as the binary one, pinpointing a cluster of star-bursting proto-ellipticals at z = 2 . z = 6 .
34 galaxy was found when looking for ultra-red sources with fluxdensities S µ m < S µ m < S µ m . The PRISM survey will allow us to look for even redder sources,potentially at even higher redshifts, and will provide a test of our understanding of the interstellar mediumand of star-formation under extreme conditions.Strongly gravitationally lensed systems have long been very difficult to identify in sufficiently large num-bers to be statistically useful. This situation changed drastically with the advent of (sub-)mm surveys. Oneof the most exciting
Herschel /SPIRE results was the direct observational confirmation that almost all thegalaxies brighter than ≈
100 mJy at 500 µ m are either strongly lensed or easily identifiable low- z spirals[72]. The surface density of strongly lensed high- z galaxies above this limit is ≈ . − , implying that anall-sky survey can detect ∼ such systems. The fact that these sources are very bright makes redshiftmeasurements with CO spectrometers and high resolution imaging with millimeter interferometers relativelyeasy. This will allow us to get detailed information on obscured star formation in the early Universe and theon processes driving it in observing times hundreds of times shorter than would be possible without the helpof gravitational amplification and with an effective source-plane resolution several times higher than couldotherwise be achieved.Large numbers of strongly lensed galaxies are also expected from large area optical surveys. It should benoted, however, that sub-mm selection has important distinctive properties. The selected lensed galaxies arevery faint in the optical, while most foreground lenses are passive ellipticals, essentially invisible at sub-mmwavelengths so that there is no, or little, contamination between images of the source and of the lens. This4akes possible the detection of lensing events with small impact parameters. Also, compared to the opticalselection, (sub-)mm selection allows us to probe earlier phases of galaxy evolution.Optical spectroscopy of galaxies acting as lenses can be exploited to measure the mass distribution of theirdark matter halos as a function of redshift. Note that Euclid will directly provide redshifts for the majority ofthe lenses out to z ∼ PRISM will study the angular correlation function of detected sources with much better statistics thanwas possible with
Herschel ’s extragalactic surveys that, altogether, cover little more than 2% of the sky. Also,the accurate photometric redshifts will allow us to follow evolution with cosmic time. Clustering propertiesmeasure the mass of dark matter halos associated with galaxies and are a powerful discriminant for galaxyformation and evolution models. Studies of the correlation function of the power spectrum also establishoccupation numbers of star-forming galaxies, and therefore their environments. In particular this study willallow us to detect high- z proto-clusters of dusty galaxies. We thus investigate an earlier evolutionary phase ofthe most massive virialized structures in the Universe. This science is possible only in the wavebands coveredby PRISM.
The
PRISM clustering data will extend to much higher redshift than
Euclid , whose wide-area survey willaccurately map the galaxy distribution up to z ∼
1. The
PRISM data will provide information at higher z , and primarily over the redshift range 2 < z <
3, corresponding to the peak in star formation activity.Moreover, optical and near-IR data severely underestimate the SFR of dust obscured starbursts and mayentirely miss these objects, which are the main targets of far-IR/sub-mm surveys such as
PRISM . Only thecombination of
PRISM and
Euclid data will provide a complete view of the spatial distribution of galaxiesand of how star formation is distributed among dark matter halos.The
PRISM sensitivity and spectral coverage will allow substantially improved measurements of thecosmic infrared background (CIB) spectrum with an accurate removal of all contaminating signals.
PRISM will also measure in a uniform way the CIB power spectrum over an unprecedented range of frequencies andof angular scales (from ∼
10 arcsec to tens of degrees).
Radio sources:
PRISM will extend the counts of radio sources, both in total and in polarized intensity, byat least one order of magnitude downwards in flux density compared to
Planck . Above 217 GHz, the countswill be determined for the first time over a substantial flux density range with good statistics. This willmake possible the first investigation of the evolutionary properties of radio sources at (sub-)mm wavelengths.
PRISM will provide measurements of the spectral energy distribution (SED) of many thousands of radiosources and of multifrequency polarization properties for hundreds of them. The vast majority of these sourcesare expected to be blazars, and the accurate determination of their spectra will allow us to understand howphysical processes occurring along relativistic jets shape the SED. For steep-spectrum sources we will obtainthe distribution of break frequencies due to electron aging, allowing an unbiased estimate of the distributionof radio source ages. Moreover, these observations will shed light on the relationship between nuclear radioemission and star formation activity in the host galaxies.
At the heart of modern cosmology is a set of initial conditions generated at very early times by what is knownas cosmic inflation . During inflation, the Universe undergoes a period of ultra-rapid accelerated expansion,typically driven by a fundamental scalar field φ , with a potential energy V ( φ ) that dominates over its kineticenergy. Quantum fluctuations of spacetime and the scalar field are amplified and stretched to cosmologicalscales resulting in a quasi-Gaussian stochastic distribution of density perturbations with amplitude A S , anda scale dependence characterized by the scalar spectral index , n S ≡ d ln A S ( k ) /d ln k . Theory predictsthat A S and n S depend on the details of V and hence φ. Furthermore, interactions of φ with itself and withother fields induce cross-correlations between perturbation modes, leading to non-Gaussianity which can bedetected in higher order statistics (bispectrum, trispectrum). Inflation also produces a bath of primordialgravitational waves characterized by an amplitude A T and the tensor spectral index n T = d ln A T ( k ) /d ln k .Remarkably, in the simplest models of inflation, the ratio between the tensor and scalar perturbations, r , is adirect probe of V in the early Universe: r ≡ A T /A S ) ≈ M P l ( V (cid:48) /V ) . Present observations estimate that V / = 3 . × r / GeV, so that measuring r effectively translates into a measurement of the energy scaleof inflation. A measurement of r , n S , and n T can directly probe the physics of the early Universe for which5igure 3: Left: Constraints on inflationary potentials from
Planck and the predicted constraints from
PRISM (not assumingde-lensing) for a fiducial value of r = 5 × − (adapted from [86]). Right: distribution of inflationary model parameters generatedusing a model independent approach that Monte-Carlo samples the inflationary flow equations. While these simulations cannotbe interpreted in a statistical way (e.g., Kinney [63], Peiris et al. [77], Chongchitnan and Efstathiou [26]), they show that modelscluster around attractor regions (adapted from [107]). there is a very rich phenomenology. Single field inflation models can relate r directly with the evolution of φ at early times. Indeed, for an inflationary expansion lasting long enough to provide the observed level ofhomogeneity and isotropy, we have ∆ φ/m Pl (cid:39) ( r/ . / . Multiple field inflation models arising in stringtheory and other proposals for unification at high energies, as well as particle and string production duringthe inflationary period, can lead to even higher values of r .Primordial gravitational waves imprint a unique, as yet undetected, signature in the CMB polarization.CMB polarization is a spin-two field on the sky, and is decomposed into the equivalent of a gradient—theE-mode—and a curl—the B-mode. Gravitational wave fluctuations are visible as the B-mode polarizationof the CMB and are the only primordial contribution to B relevant at the time of recombination. Hence adetection of B-modes is a direct probe of r, and thus the energy scale of inflation and other primordial energeticprocesses. Furthermore, in the simple case of slow-roll inflation we have that r ≈ − n T . Additional detailedmeasurements of the shape of the temperature and polarization spectra will measure higher derivatives ofthe inflationary potential.The 2013 Planck data release has significantly improved previous constraints on inflationary models. Inparticular, and in the context of the simplest ΛCDM scenario,
Planck results provide n S = 0 . ± . r < .
12. These results are notable because exact scale invariance (i.e., n S = 1) of primordial perturbationsis ruled out at more than 5 σ . When specific inflationary models are considered, Planck imposes significantconstraints on the potential (Fig. 3), as discussed in Ref. [86]. Indeed
Planck has shown that it is possibleto test many inflation models using the CMB temperature data, yet even a forecast
Planck limit r < . (cid:96) , where themost constraining information on r is situated. If we propagate the uncertainties connected to foregroundcontamination into the parameter error forecasts [107, 6, 9], we find that the proposed experimental set-upwill enable us to explore most large field (single field) inflation models (i.e., where the field moves for ≥ M P )and to rule in or out all large-field models, as illustrated in the right-hand panel of Fig. 3.As the work by Smith et al. [98] indicates (see Fig. 8), the instrumental sensitivity, angular resolutionand, as a result, foreground control and subtraction will enable us to achieve a detailed mapping of thelensing signal, and in particular to implement de-lensing techniques for the measurement of r , improving bya factor of three our constraint on r . This implies that PRISM will detect r ∼ × − at more than 3 σ .This performance is very close, within factors O (1), to what an ideal experiment (i.e., with no noise and noforegrounds) could achieve, allowing PRISM to directly probe physics at an energy scale a staggering twelveorders of magnitude higher than the center-of-mass energy at the Large Hadron Collider (LHC).6igure 4: Reconstruction noise on the lensing deflectionpower spectrum forecast for the full
Planck mission (four sur-veys; left) and
PRISM (right) using temperature alone (red)and temperature and polarization (blue). For
Planck we alsoshow the approximate noise level for the temperature analysisof the nominal-mission data (red dashed) [81], and for
PRISM ,we also show the approximate noise level (green) for an im-proved iterative version of the reconstruction estimator. Thedeflection power spectrum is plotted based on the linear matterpower spectrum (black solid) and with non-linear corrections(black dashed).
The temperature anisotropies of the CMB have proved to be a remarkably clean probe of the high-redshiftuniverse and have allowed the standard cosmological model to be tested to high precision. However, theaccuracy of the recent results from
Planck , based on the temperature anisotropies, are now close to beinglimited by errors in modelling extragalactic foregrounds. Fortunately, further progress can be made with thepolarization anisotropies on small angular scales since the degree of polarization of the anisotropies is relativelylarger there (around 4% by l = 2000) than the foreground emission. By surveying the full polarized sky inmany frequency bands, and with uniform calibration, PRISM will fully exploit the small-scale polarization ofthe CMB, improving significantly on results currently obtained from the temperature and those conceivablyobtainable in the future with ground-based experiments.
Probing the dark universe with CMB lensing:
Gravitational lensing of the CMB provides a clean probeof matter clustering integrated to high redshift. Lensing can be reconstructed from the CMB anisotropiesvia specific non-Gaussian signatures imprinted by the lenses.
Planck has detected lensing via this routeat the 25 σ level using the temperature anisotropies, but with low S/N per lensing mode. Polarization-based reconstructions from
PRISM will be a major advance over
Planck , achieving
S/N (cid:29) l ≈
600 over nearly the full sky (see Fig. 4). Significantly,
PRISM can extract all of theinformation in the deflection power spectrum on scales where linear theory is reliable. To illustrate thepower of the lensing measurements from
PRISM in constraining physics that is inaccessible to the primaryanisotropies alone due to degeneracies, we consider the mass of (light) neutrinos. Oscillation data constrain(squared) mass differences, and provide only lower bounds on the total mass summed over eigenstates:0 .
06 eV and 0 . β -decay experiments. However, masses of these orders can be probed cosmologically via their effecton the clustering of matter. In w CDM models with massive neutrinos, we forecast a 1 σ error of 0 .
04 eV for thesummed mass. This constraint can be improved further by combining with near-future BAO measurements,for example by a factor of almost two using BOSS, at which point it becomes possible to distinguish betweenthe normal and inverted hierarchies (in the hierarchical limits) [43].Lensing constraints from
PRISM would be highly complementary to those from upcoming optical cosmicshear surveys, e.g.,
Euclid . The systematic effects are quite different with non-linearities being much less ofan issue for CMB lensing and there are no intrinsic-alignment effects. The combination of the two probesof mass is particularly promising, since it allows calibration of multiplicative bias effects such as due to PSFcorrections in the optical. Cross-correlating CMB lensing with other probes of large-scale structure, such asgalaxies, the Ly α forest or CIB clustering (see Sec. 4), also has exceptional promise, allowing self-calibrationof the tracer’s bias relation at the sub-percent level. Primordial non-Gaussianity:
Non-Gaussianity (NG) is now demonstrably a robust quantitative probeof cosmological physics [83].
Planck results dramatically improved previous NG analyses, offering the moststringent test to date of inflationary theory (with f locNL = 2 . ± .
8) while also detecting for the first time ISW-lensing and diffuse point source bispectra. Already
Planck offers enticing clues about the nontrivial ‘shape’of the CMB bispectrum of our universe (see Fig. 5), the origin of which is yet to be explained.
PRISM wouldoffer the highest precision reconstructions of the CMB temperature and polarization bispectra and trispectra,which will provide a decisive and unambiguous probe of primordial cosmology back to the Planck era. At thesame time,
PRISM
NG data will open new windows for investigating dark energy and gravitational physics,7
500 1000 1500 2000
0 500 1000 1500
0 500 1000 1500 2000
Figure 5:
Planck
CMB temperature bispectrum [83] (left) and primordial (right) and late-time (middle) non-Gaussian shapes [83,82]. Note the periodic CMB ISW-lensing signal (middle) in the squeezed limit along the edges, which is seen at the 2.5 σ level inthe Planck bispectrum on the left. Scale-invariant signals predicted by many inflationary models are strongly constrained by the
Planck bispectrum, although ‘oscillatory’ and ‘flattened’ features hint at new physics. An example of an inflationary ‘feature’model is shown on the right.
PRISM will probe these hints with an order of magnitude more resolved triangle configurations. as well as astrophysical sources, large-scale structure and galactic history.A unique advantage of the CMB for probing NG is its ability to recognize the distinct patterns that physicalmechanisms leave in the shape of higher-order correlators (Fig. 5).
PRISM will allow a vastly enhancedexploration of physically-predicted NG shapes compared to any other projected probe of NG. For example,the constraint volume in bispectrum space spanned by the local, equilateral and flattened bispectra will reduceby a factor of 75 compared to the current
Planck volume, and a factor of 30 over that predicted from thefull-mission
Planck data (including polarization). From polarization maps alone (which provide informationindependent of the temperature maps), we expect a volume reduction factor from the full-mission
Planck datato
PRISM of order 110. Moreover, local-model trispectrum parameters could be measured with a precision∆ g NL = 3 × and ∆ τ NL = 1 × [97]. These could investigate consistency conditions between polyspectra,which can be used to test large classes of multi-field inflation models in addition to single-field inflation. Thereare other alternative inflationary scenarios for which an observable non-Gaussian signal is quite natural, e.g.,those with features or periodicity in the inflationary potential (Fig. 5). Each of these models has a distinctfingerprint, many uncorrelated with the standard three primordial shapes and, in all cases, PRISM wouldsignificantly improve over present
Planck constraints, offering genuine discovery potential. Beyond searchesfor primordial NG,
PRISM is guaranteed to make important observations of late-time NG. For example, itwill decisively detect and characterize the lensing-ISW correlation, driven by dark energy, achieving a 9 σ detection, resulting in a new probe of dark energy physics from the CMB alone. Parameters from high-resolution polarization spectra:
PRISM will measure the CMB angular powerspectra with outstanding precision to small angular scales. In particular, in the 105–200 GHz frequency range,the relatively clean EE polarization spectrum is cosmic-variance limited to l = 2500 (and the BB spectrumfrom lensing to l = 1100). Such a remarkable measurement of the polarization of the CMB damping tail willbe an invaluable source of information on the shape of the primordial power spectrum and the fundamentalmatter content of the Universe. For example, in ΛCDM models, the spectral index and its running willbe measured more precisely than with current Planck data by factors of five and three, respectively. TheHubble constant (a point of tension between
Planck data and direct astrophysical measurements) will bemeasured a factor of 10 better than currently (and 2 . Planck data).Fundamental questions about the matter content include the effective number of relativistic species N eff ,for which a non-standard value (which can relieve the Planck – H tension) could be due to sterile neutrinos,as advocated in particle physics to explain certain anomalies in the neutrino sector, the helium abundance Y P , which provides a clean test of standard BBN, the neutrino mass, and the dark matter annihilationcross-section. In one-parameter extensions of ΛCDM, PRISM will measure N eff to 2% precision and Y P to 1%. These values indicate that a 2 σ anomaly hinted at by Planck could be confirmed decisively with
PRISM . Moreover, from its measurement of the B -mode power spectrum, PRISM should extend the rangeof sensitivity to cosmic strings by an order of magnitude over the recent
Planck constraints [82, 1].
COBE / FIRAS has shown that the average CMB spectrum is extremely close to a perfect blackbody, with8 n [GHz]10 -28 -27 -26 -25 -24 -23 D I n [ W m - H z - s r - ] R e i o n i z a t i o n & S t r u c t u r e f o r m a t i o n D ec ay i n g p a r t i c l e S il k d a m p i n g ( s t a n d a r d ) s Sensitivity R ec o m b i n a t i o n li n e s Monopole distortion signals S i l k d a m p i n g ( s t e p ) Figure 6:
Left: spectral distortions for different scenarios. Thick lines denote positive, and thinner lines negative signal. The1 σ sensitivities of PRISM for different designs are also indicated. Right: projected constraints on different metal ions. possible departures limited to ∆ I ν /I ν (cid:46) (few) × − [68, 37]. This places very tight constraints on thethermal history of our Universe, ruling out cosmologies with extended periods of significant energy releaseat redshifts z (cid:46) (few) × [108, 100, 54, 30, 12, 49, 16, 22, 58]. There are, however, a large number ofastrophysical and cosmological processes that cause (inevitable) spectral distortions of the CMB at a levelthat has only come within reach of present-day technology. With PRISM an unexplored window to the earlyuniverse will be opened, allowing detailed studies of (see Fig. 6 for illustration):
Reionization and structure formation:
Radiation from the first stars and galaxies [53, 2], feedbackby supernovae [73] and structure formation shocks [101, 14, 71] heat the IGM at redshifts z (cid:46) − y -distortion with averageamplitude ∆ I ν /I ν (cid:39) − − − . This signal will be detected at more than a 100 σ by PRISM , providing asensitive probe of reionization physics and delivering a census of the missing baryons in the local Universe.
PRISM furthermore has the potential to separate the spatially varying signature caused by the WHIM andproto-clusters [109]. It also offers a unique opportunity to observe the free-free distortion associated withreionization, providing a complementary way to study the late evolution of inhomogeneities [87].
Decaying and annihilating relics:
The CMB spectrum will establish tight limits on decaying and annihi-lating particles in the pre-recombination epoch [50, 29, 69, 17, 22]. This is especially interesting for decayingparticles with lifetimes t X (cid:39) − sec, as the exact shape of the distortion encodes when the decay oc-curred [22, 59, 18, 19]. PRISM therefore provides an unprecedented probe of early-universe particle physics,with many natural particle candidates found in supersymmetric models [36, 35].
Constraining inflation:
Silk damping of small-scale perturbations gives rise to CMB distortions [100,28, 3, 52] which directly depend on the shape and amplitude of the primordial power spectrum at scales0 . (cid:46) λ (cid:46) (cid:46) (cid:96) (cid:46) ) [25, 62]. This allows constraining the trajectory of theinflaton at stages unexplored by ongoing or planned experiments [24, 88, 60], extending our reach from 7 e -folds of inflation probed with the CMB anisotropies to a total of 17 e -folds. The signal is also sensitiveto the difference between adiabatic and isocurvature perturbations [3, 51, 31, 20], as well as primordial non-Gaussianity in the ultra squeezed-limit, leading to a spatially varying spectral signal that correlates withCMB temperature anisotropies as large angular scales [75, 38]. A competing monopole signal, characterizedby a negative µ - and y -parameter, is introduced by the adiabatic cooling of ordinary matter [16, 22, 61], towhich PRISM will also be sensitive.
Metals during the dark ages:
Any scattering of CMB photons after recombination blurs CMB anisotropiesat small scales, while producing new anisotropies at large scales. Electrons from the reionization epoch arethe dominant source of optical depth, causing a signature already detected by WMAP and Planck [7, 84].The resonant scattering of CMB photon by fine structure lines of metals and heavy ions produced by thefirst stars adds to this optical depth, making it frequency-dependent [5]. By comparing CMB temperatureand polarization anisotropies at different frequencies one can thus determine the abundances of ions such asOI, OIII, NII, NIII, CI, CII at different redshifts [46, 48]. Furthermore, UV radiation emitted by the first9tars can push the OI 63.2 µ m and CII 157.7 µ m transitions out of equilibrium with the CMB, producing adistortion ∆ I ν /I ν (cid:39) − − − due to fine structure emission [41, 47], providing yet another window toreionization within reach of PRISM . Cosmological recombination radiation:
The recombination of H and He [34] at redshifts z (cid:39) − ,corresponding to (cid:39)
260 kyr (
H i ), (cid:39)
130 kyr (
He i ), and (cid:39)
18 kyr (
He ii ) after the big bang [93, 21, 94]. Thesignal provides an independent determination of the cosmological parameters (such as the baryon density and pre-stellar helium abundance) and direct measurements of the recombination dynamics, probing the Universeat stages well before the last scattering surface [99]. The effect on the TT power spectrum introduced byresonance scattering of CMB photons by the first lines of the Balmer and Paschen series [92, 45, 48] willalso be detectable with
PRISM , providing an additional opportunity to directly constrain the recombinationhistory and obtain independent determinations of cosmological parameters (e.g. Ω b or Ω m ). Non-Gaussianity:
CMB spectral distortions can also provide a new probe of primordial NG [76]. We knowalmost nothing about NG on the small scales that can be probed via these observations. In particular, thecross-correlation between µ -type distortions and CMB anisotropies is naturally sensitive to the very squeezedlimit of the primordial bispectrum (probing scales as small as 50 ≤ k Mpc ≤ ). Also, the power spectrumof µ -distortions can probe the trispectrum of primordial fluctuations. Such measurements can be particularlyconstraining for models where the primordial power spectrum grows on small scales (see e.g. [23]), and values f locNL < µ -type distortions can shed light on non-standard initial states for thequantum fluctuations. For a large class of inflationary models characterized by a non-Bunch-Davies vacuum(whose bispectrum is enhanced in the squeezed limit with respect to the local form) a high S/N can beachieved [39].All these examples demonstrate that the CMB spectrum provides a rich and unique source of complementaryinformation about the early Universe, with the certainty of a detection of spectral distortions at a level withinreach of
PRISM ’s capabilities. The CMB spectrum will also establish interesting constraints on the powerspectrum of small-scale magnetic fields [57], cosmic strings [74, 103, 104], evaporating primordial black holes[13], decay of vacuum energy density [4, 11, 29], and other new physics [67, 10], to mention a few more exoticexamples. Deciphering all these signals will be a big challenge for the future. This area has great potentialfor new discoveries and for providing new independent constraints on unexplored processes that cannot beexplored by other means.
The data analysis is still on-going but it is already clear that
Herschel and
Planck will have a profoundand lasting impact on our understanding of the interstellar medium and star formation.
PRISM holds evengreater promise for breakthroughs. Dust and synchrotron radiation are the dominant contributions to the skyemission and polarization to be observed by
PRISM . Dust emission is an optically thin tracer of the structureof interstellar matter. Synchrotron radiation traces the magnetic field over the whole volume of the Galaxy,while dust polarization traces the magnetic field within the thin star forming disk, where the interstellarmatter is concentrated.
PRISM will image these two complementary tracers with unprecedented sensitivityand angular resolution. It will also provide all-sky images of spectral lines, which are key diagnostics of inter-stellar gas physics. No other initiative offers a comparable imaging capability of interstellar components overas wide a range of scales. In the following subsections we detail how
PRISM will address three fundamentalquestions of Galactic astrophysics: (1) What are the processes that structure the interstellar medium? (2)What role does the magnetic field play in star formation? (3) What are the processes that determine thecomposition and evolution of interstellar dust?
Herschel far infrared observations have provided astronomers new insight into how turbulence stirs up theinterstellar gas, giving rise to a filamentary, web-like structure within the diffuse interstellar medium andmolecular clouds.
PRISM will extend the
Herschel dust observations to the whole sky and provide uniquedata on emission lines key to quantifying physical processes. The spectral range of
PRISM includes atomicand molecular lines that serve as diagnostics of the gas density and temperature, its chemical state, andenergy budget.
Herschel has observed these lines along discrete lines of sight with very limited imaging. By10apping these lines and dust emission over the whole sky at an angular resolution comparable to that of
Herschel , PRISM will probe the connection between the structure of matter and gas cooling across scales.The
PRISM sky maps will provide multiple clues to characterize the physical processes that shape in-terstellar matter. The CII, CI, and OI fine structure lines and the rotational lines of CO and H O are themain cooling lines of cold neutral medium and molecular clouds and probe local physical conditions and theexchange of energy associated with the formation of molecular gas within the diffuse interstellar medium andof stars within molecular clouds. The NII lines at 122 and 205 µ m are spectroscopic tracers of the ionizedgas. These lines are essential for distinguishing the contribution of neutral and ionized gas to the CII emis-sion. PRISM will have the sensitivity to image the CII line emission at sub-arcminute resolution even at theGalactic poles. The CII map can be combined with HI and dust observations to study the formation of coldgas from the warm neutral medium through the thermal instability. This analysis will probe the expectedlink, yet to be confirmed observationally, between the small-scale structure of the cold interstellar mediumand gas cooling. The CII line emission is also key to studying the formation of molecular gas by tracing theCO-dark H gas [78]. In star forming molecular clouds, the CO, CI, OI, and H O lines are the key tracersof the processes creating the initial conditions of star formation and of the feedback from newly formed starson their parent clouds.
Star formation results from the action of gravity, counteracted by thermal, magnetic, and turbulent pressures[44]. For stars to form, gravity must locally become the dominant force. This happens when the turbulentenergy has dissipated and matter has condensed without increasing the magnetic field by a comparableamount. What are the processes that drive and regulate the rate at which matter reaches this stage? This isa long standing question to which theorists have over the decades offered multiple explanations, focusing oneither ambipolar diffusion, turbulence, or magnetic reconnection to decouple matter from the magnetic fieldand allow the formation of condensations of gas in which stars may form [27].
PRISM observations of the polarization in the far-IR and sub-mm will provide unique clues to understandthe role of the magnetic field in star formation. Compared to synchrotron radiation and Faraday rotation,dust polarization images the structure of the magnetic field through an emission process tracing matter.It is best suited to characterize the interplay between turbulence, gravity, and the Galactic magnetic field.The
PRISM data will provide unique data to study magneto-hydrodynamical turbulence because it willdrastically increase the spectral range of accurately probed magneto-hydrodynamical modes. The data willprovide unprecedented statistical information to characterize the energy injection and energy transfer downto the dissipation scales.Polarization data from the
PRISM survey will have the sensitivity and angular resolution required tomap continuously the Galactic magnetic field over the whole sky down to sub-arcminute resolution even atthe Galactic poles. The wide frequency range of the mission will measure polarization for separate emissioncomponents with distinct temperatures along the line of sight.
PRISM will provide a new perspective onthe structure of the magnetic field in molecular clouds, independent of grain alignment, by imaging thepolarization of CO emission in multiple rotational lines [40]. No project offers comparable capabilities.Planck has provided the first all-sky maps of dust polarization with 5’ resolution but the data is sensitivitylimited even at the highest Planck frequency (353 GHz). Ground based telescopes at sub-mm and millimeterwavelengths of bright compact sources at arcsecond resolution (for example with ALMA) complement thefull-sky survey of extended emission from the diffuse interstellar medium and molecular clouds that only
PRISM can carry out.
The combination of spectral and spatial information provided by
PRISM will provide new tools for study-ing the interstellar dust, in particular its nature and its evolution. Dust properties (e.g., size, temperature,emissivity) are found to vary from one line of sight to another within the diffuse interstellar medium andmolecular clouds. These observations indicate that dust grains evolve in a manner depending on their envi-ronment within the interstellar medium. They can grow through the formation of refractory or ice mantles,or by coagulation into aggregates in dense and quiescent regions. They can also be destroyed by fragmenta-tion and erosion of their mantles under more violent conditions. The composition of interstellar dust reflectsthe action of interstellar processes, which contribute to breaking and reconstituting grains over timescales11uch shorter than the timescale of injection by stellar ejecta. While there is broad consensus on this view ofinterstellar dust, the processes that drive its evolution in space are poorly understood [32]. Understandinginterstellar dust evolution is a major challenge in astrophysics underlying key physical and chemical processesin interstellar space. In particular, to fully exploit the
PRISM data we will need to characterize where in theinterstellar medium grains are aligned with respect to the Galactic magnetic field and with what efficiency.Large dust grains (size >
10 nm) dominate the dust mass. Within the diffuse interstellar medium,these grains are cold ( ∼ −
20 K) and emit within the
PRISM frequency range. Dipole emission fromsmall rapidly spinning dust particles constitutes an additional emission component, known as anomalousmicrowave emission. Magnetic dipole radiation from thermal fluctuations in magnetic nano-particles mayalso be a significant emission component over the frequency range relevant to CMB studies [33]. To achievethe
PRISM objectives on CMB polarization, it is necessary to characterize the spectral dependence of thepolarized signal from each of these dust components with high accuracy across the sky. This is a challengebut also a unique opportunity for dust studies. The spectral energy distribution of dust emission and thepolarization signal can be cross-correlated with the spectral diagnostics of the interstellar medium structureto characterize the physical processes that determine the composition and evolution of interstellar dust. Thesame data analysis will also elucidate the physics of grain alignment.
PRISM will also probe the zodiacal dust emission from within our solar system. The fact that PRISMscans a substantial portion of the sky each day allows for a three-dimensional tomographic mapping of thezodiacal emission. Understanding zodiacal emission is crucial both to understanding our solar system and tocarrying out a complete foreground separation.
Figure 7:
The PRISM spacecraft with its two instru-ments: PIM, with a 3.5-diameter telescope with a FOVat ∼ ◦ from the spacecraft spin axis, and ASP, alignedwith the spin axis. The science program above requires measuring the skybrightness and polarization at high angular resolution andin many frequency bands across a wide spectral range. Italso requires measuring the absolute spectrum of the skybackground with moderate angular and spectral resolution.As a baseline, we propose to perform the best possiblespectro-polarimetric sky survey in the 30-6000 GHz fre-quency range with two instruments optimized for best jointperformance sharing a single platform in orbit around theSun-Earth L2 Lagrange point: (1) a polarimetric imager (PIM) observing with about 30 broad and 300 narrow spec-tral bands with a diffraction limited angular resolution anda sensitivity limited by the photon noise of the sky emissionitself; and (2) an absolute spectro-photometer (ASP) thatwill measure sky emission spectra with a spectral resolutionbetween 500 MHz and 15 GHz and an angular resolutionof about 1.4 ◦ . These complementary instruments will mapsimultaneously the absolute sky intensity and polarization with high sensitivity and with high spectral orspatial resolution. The data from both instruments can be binned (in frequency) and smoothed to obtainmatching observations with δν/ν ≈ .
25 and 1.4 ◦ resolution, allowing on-sky inter-calibration on large scales(and hence absolute calibration of the PIM). This will also enable correction of the ASP spectra from fore-ground contamination using high resolution component maps extracted from PIM data (e.g., large clusters y -distortion in the ASP data and line emission from emitting regions unresolved in the coarse resolution ASPmaps).As the scientific outcome of this mission depends on the complementarity of both instruments and onthe control of systematic errors, a careful optimization of the ASP and the PIM (number and bandwidth ofspectral bands vs. sensitivity) and of the mission (scanning strategy, joint analysis tools) with comprehensivesimulations is an essential future phase of the mission study.The focal planes of both instruments will be cooled to 0.1K using a cryogenic system adapted from thatof Planck , with continuous recycling of the gases for an improved mission duration of 4 years (baseline) orlonger. 12 .1 Instruments ν Range ∆ ν/ν n det θ fwhm σ I per det σ ( Q,U ) per det Main molec. & atomic lines1 arcmin 1 arcminGHz GHz µ K RJ µ K CMB µ K RJ µ K CMB
30 26-34 .25 50 17’ 61.9 63.4 87.6 89.736 31-41 .25 100 14’ 57.8 59.7 81.7 84.543 38-48 .25 100 12’ 53.9 56.5 76.2 79.951 45-59 .25 150 10’ 50.2 53.7 71.0 75.962 54-70 .25 150 8.2’ 46.1 50.8 65.2 71.975 65-85 .25 150 6.8’ 42.0 48.5 59.4 68.690 78-100 .25 200 5.7’ 38.0 46.7 53.8 66.0 HCN & HCO + at 89 GHz105 95-120 .25 250 4.8’ 34.5 45.6 48.8 64.4 CO at 110-115 GHz135 120-150 .25 300 3.8’ 28.6 44.9 40.4 63.4160 135-175 .25 350 3.2’ 24.4 45.5 34.5 64.3185 165-210 .25 350 2.8’ 20.8 47.1 29.4 66.6 HCN & HCO + at 177 GHz200 180-220 .20 350 2.5’ 18.9 48.5 26.7 68.6220 195-250 .25 350 2.3’ 16.5 50.9 23.4 71.9 CO at 220-230 GHz265 235-300 .25 350 1.9’ 12.2 58.5 17.3 82.8 HCN & HCO + at 266 GHz300 270-330 .20 350 1.7’ 9.6 67.1 13.6 94.9320 280-360 .25 350 1.6’ 8.4 73.2 11.8 103 CO, HCN & HCO +
395 360-435 .20 350 1.3’ 4.9 107 7.0 151460 405-520 .25 350 1.1’ 3.1 156 4.4 221 CO, HCN & HCO +
555 485-625 .25 300 55” 1.6 297 2.3 420 C-I, HCN, HCO + , H O, CO660 580-750 .25 300 46” 0.85 700 1.2 990 CO, HCN & HCO + nK RJ kJy/sr nK RJ kJy/sr800 700-900 .25 200 38” 483 9.5 683 13.4960 840-1080 .25 200 32” 390 11.0 552 15.61150 1000-1300 .25 200 27” 361 14.6 510 20.71380 1200-1550 .25 200 22” 331 19.4 468 27.4 N-II at 1461 GHz1660 1470-1860 .25 200 18” 290 24.5 410 34.71990 1740-2240 .25 200 15” 241 29.3 341 41.5 C-II at 1900 GHz2400 2100-2700 .25 200 13” 188 33.3 266 47.1 N-II at 2460 GHz2850 2500-3200 .25 200 11” 146 36.4 206 51.43450 3000-3900 .25 200 8.8” 113 41.4 160 58.5 O-III at 3393 GHz4100 3600-4600 .25 200 7.4” 98 50.8 139 71.85000 4350-5550 .25 200 6.1” 91 70.1 129 99.1 O-I at 4765 GHz6000 5200-6800 .25 200 5.1” 87 96.7 124 136 O-III at 5786 GHz Table 1:
The 32 broad-band channels of the polarized imager with a total of 7600 detectors. Sensitivities are averages for skyregions at galactic latitude and ecliptic latitude both higher than 30 ◦ . A detector noise level equal to the sky photon noise isassumed. The mission sensitivity per frequency channel is the sensitivity per detector divided by √ n det . The polarimetric imager:
The optical configuration relies on a dual off-axis mirror telescope with a3.5 m projected aperture diameter primary and a 0.8 m diameter secondary coupled to a multi spectral bandpolarimeter. The broad-band PIM comprises 32 main channels of δν/ν ≈ .
25 relying on dual-polarized pixelarrays (Table 1). At frequencies below 700 GHz, the emphasis is on the sensitivity and control of systematicsfor CMB and SZ science.The whole frequency range will also be covered at higher spectral resolution ( δν/ν ≈ . ∼
300 frequency channels (not listed in Table 1) will be obtained using antenna coupled bolometersand channelizers to split the spectral band of each broad-band horn into 5-10 narrow frequency bands,with similar numbers of narrow-band and broad-band detectors. The sensitivity to continuum emission perdetector is reduced in the narrow-band channels as compared to the broad-band channels, but the sensitivityto spectral lines is better by a factor of about 2-3.
The absolute spectrophotometer:
A Martin-Puplett Fourier Transform Spectrometer (FTS) will allowfor a large throughput and sensitivity, differential measurements (the sky is compared to an internal blackbodycalibrator as in COBE-FIRAS), and a variable spectral resolution. Dichroics at the two output ports canoptionally split the full 30 - 6000 GHz range into sub-bands with reduced photon noise. The instrumentis cooled at 2.7K, so that the bolometric detector sensitivity is limited by photon noise from the sky. Twooperating modes are available: high-resolution (∆ ν ∼ . ν ∼
15 GHz). Thesensitivity of the high-resolution mode is 30 times worse than for the low-resolution mode. The instrumentbeam is aligned with the spin axis of the satellite, so that precession has a negligible effect during the13nterferogram scan ( ∼ Band Resolution A Ω Background NEP ν Global 4-yr mission(GHz) (GHz) (cm sr) (pW) (W/m /sr/Hz ×√ s) sensitivity (W/m /sr/Hz)30-6000 15 1 150 1 . × − . × − . × − . × −
500 - 6000 15 1 70 1 . × − . × − . × − . × − . × − . × − . × − . × − . × − . × − Table 2:
FTS performance of three possible configurations for photon noise limited detectors. With an entrance pupil 50 cmin diameter, the baseline throughput is ∼ sr and the angular resolution 1.4 ◦ . The theoretical monopole sensitivity for eachspectral bin is reported in the last column assuming 4 years of observation and 75% useful sky. The actual sensitivity, takinginto account efficiency factors can be 2-3 times worse. Line 1 is a configuration with an ultra-wide spectral coverage obtainedwith one detector in both output ports. In lines 2-3 the detectors at the output ports are sensitive to different bands. In lines4-7 each output port is split into two sub-bands using dichroics to minimize photon noise in the low-frequency bins. Using detectors with A Ω ∼ sr and angular resolution ∼ ◦ , we estimate that the CIB can bemeasured with S/N = 10 in a fraction of a second at 1500 GHz and in ∼
10 seconds at 140 GHz, while a y -distortion ∼ − can be measured with S/N = 10 at 350 GHz in two hours of integration. Recombinationlines could be measured integrating over the whole mission if the overall stability of the instrument and thequality of the reference blackbody are sufficient.The main issue for this instrument is the control of systematic effects. The instrument design allows fora number of zero tests and cross-checks on the data. The main problem is to control the blackness of thereference and calibration blackbodies with the required accuracy. Reflectivities lower than R = − / −
60 dBhave been obtained in the frequency range of interest in the Planck and ARCADE references. We plan toachieve
R < −
70 dB building on these experiences through a combination of electromagnetic simulations andlaboratory emissivity measurements on improved shapes and space-qualified materials.
The observing strategy must provide: (1) full sky coverage for both instruments; (2) cross-linked scan pathsand observation of all sky pixels in many orientations for all detectors of the PIM; (3) fast scanning of thePIM to avoid low-frequency drifts; (4) slow scanning for the ASP field of view to allow for few seconds longinterferogram scans with negligible depointing; (5) avoiding direct solar radiation on the payload. Theserequirements can be satisfied by a spinning spacecraft with the ASP aligned along the spin axis and the PIMoffset by θ spin ≈ ◦ (Fig. 7). During each spacecraft rotation (with ω spin of a few rpm), the PIM scanscircles of diameter ≈ θ spin while the APS rotates in place. A slow precession of the spin axis (with a periodbetween a few hours and one day) with a precession angle θ prec ≈ ◦ results in slow scans of the ASP onlarge circles of diameter 2 θ prec . Finally, the precession axis evolves by about 1 ◦ per day along the eclipticplane to keep the payload away from the Sun, and also slowly moves perpendicular to the ecliptic plane soas to map the ecliptic poles. Deployable screens isolate the payload from the heat from the Sun, providing afirst stage of passive cooling to ≈
40 K.
Telescope temperature:
Actively cooling the telescope to 4 K (mission objective) instead of 40 K (achiev-able by passive cooling) substantialy improves the sensitivity, especially for frequencies above 200 GHz.
PRISM will benefit from the development activities for the SPICA mission, the telescope of which is basedon a 3.5 m diameter primary cooled to 5 K.
Polarization modulation:
The baseline, similar to the solution proposed in the previous SAMPAN andEPIC studies, relies on the scanning strategy and the rotation of the entire payload. However alternativestrategies such as the use of a half-wave plate in front of the focal plane (the receivers being the major sourceof instrumental polarization) could be considered during a trade-off analysis.14 c ange Req. NEP Req. τ Focal plane technology[
GHz ] (cid:104) − W/ √ Hz (cid:105) [ ms ] Detector technology Optical couplingBaseline Backup Baseline Backup Table 3:
Required NEP and time constants for various frequency ranges and corresponding baseline and backup focal planetechnology. TES: Transition Edge Sensors (Technology Readiness Level 5); HEMT: High Electron Mobility Transistor (TRL 5);KID: Kinetic Inductance Detector (TRL 5); HEB: Hot Electron Bolometer (TRL 4); CEB: Cold Electron Bolometer (TRL 3);HA: Horn Array (TRL 9); LHA: Lithographed Horn Array (TRL 5); MPA: Multichroic Planar Antenna (TRL 4); CSA: CrossedSlot Antenna (TRL 5); POMT: Planar Ortho-Mode Transducer (TRL 5)
Detectors:
Direct detectors (such as TES bolometers, CEBs or KIDs) are the most sensitive detectors atmm wavelengths. Bolometers have achieved photon noise limited in-flight performance with the Planck [105]and Herschel [42] missions. Large bolometer arrays with thousands of pixels are currently used on largeground-based telescopes. They are currently not proven as a viable technology for 30 to 70 GHz but it islikely that their efficiency will improve in the next few years at low frequencies. For instance studies [64]have shown that 70 GHz CEBs could lead to NEPs of (few) × − W · Hz / . As an alternative solution,the PRISM instruments could take advantage of the recent breakthroughs in cryogenic HEMT technology,with sensitivities predicted to reach 2-3 times the quantum limit up to 150-200 GHz (instead of 4-5 timesup to 100 GHz so far). In addition, these devices allowing for cryogenically cooled miniaturized polarimeterdesigns will simplify the thermo-mechanical design. Hence, while a single detector technology throughout theinstruments would be preferable, the option of using a combination of HEMTs and bolometers remains open(Table 3).
Detector time constants:
The fast scanning of the
PRISM mission requires fast detector time constants,of order 1 ms at 100 GHz, down to ∼ µ s at 6 THz. These time constants are challenging (especially athigh frequencies), but have already been achieved with recent TESs, KIDs or CEBs. We propose that the mission include a small ancillary spacecraft serving the following functions:
Telecommunication:
The high resolution mapping of the full sky with the many detectors of
PRISM witha lossless compression of 4 gives a total data rate of ∼
350 Mbit/s (of which 300 Mbit/s is from the channelsabove 700 GHz). Further on-board reduction by a factor ∼ −
20 can be achieved by averaging the timelinesof detectors following each other on the same scan path (after automatic removal of spikes due to cosmic rays)to yield a total data rate <
40 Mbit/s (a few times greater than Euclid or Gaia). A phased-array antenna orcounter-rotating antenna on the main spacecraft is an option. Decoupling the communication function fromthe main spacecraft using an ancillary spacecraft as an intermediate station for data transmission will allowfor a maximally flexible scanning strategy for the best polarization modulation and full sky coverage.
In-flight calibration:
The hardest
PRISM design problem is ensuring that the performance is limited bydetector noise rather than systematic effects and calibration uncertainties. While pre-flight calibration isnecessary, an ancillary spacecraft fitted with calibrated, polarized sources could be used for precise in-flightcalibration of the polarization response and polarization angles of the detectors, and for main beams andfar sidelobe measurements down to extremely low levels (below -140 dB) at several times during the missionlifetime.
10 Competition and complementarity with other observations
B-mode experiments:
Searching for primordial gravitational waves through B-mode polarization is theprincipal science driver of numerous suborbital experiments (e.g.,
BICEP, QUIET, SPIDER, ACTPol, SPT-Pol, QUBIC, EBEX, PolarBear, QUIJOTE ) despite considerable limitations due to atmospheric opacity,far-side lobe pickup from the ground, and unstable observing conditions that make controlling systemic er-rors especially difficult, particularly on the largest angular scales where the B mode signal is largest. Forecastsof r from ground-based experiments are often impressive but assume very simple foregrounds. For this reasona detection of r from the ground would provide a strong motivation for a confirmation and more precise15haracterization from space. Moreover, two US space missions concepts, CMBPol and
PIXIE, and one inJapan,
LiteBird, have been proposed, but none has yet been funded. Among the current space mission con-cepts,
PRISM is the most ambitious and encompasses the broadest science case.
LiteBird is a highly-targeted,low-cost Japanese B-mode mission concept, in many respects similar to the
BPol mission proposed to ESAin 2007. With its coarse angular resolution and limited sensitivity,
LiteBird would be able to detect B-modesassuming that r is not too small and that the foregrounds are not too complicated. LiteBird, however, lacksthe angular resolution needed to make significant contributions to other key science objectives. The US
EPIC-CS mission is the most similar to the present proposal but has considerably less frequency coverage, fewerfrequency bands, and no absolute spectral capability. The US mission concept
PIXIE proposes an improvedversion of the
FIRAS spectrometer to measure B-modes and perform absolute spectroscopy simultaneously,but with an effective resolution of only 2 . ◦ . Cluster observations:
When
PRISM flies, the eROSITA
X-ray survey will likely be the only deeper all-skycluster survey available. 20–30 times more sensitive than ROSAT, eROSITA ’s principal goal is to explorecosmological models using galaxy clusters. Forecasts predict that eROSITA will detect ∼ clusters atmore than 100 photon counts, which is sufficient to provide a good detection and in many cases to detect thesource as extended in X-rays. The main survey provides a good sample of galaxy clusters typically out to z = 1 with some very massive and exceptional clusters at larger distance.The large majority of these clusters will be re-detected by PRISM and thus provide an invaluable inter-calibration of X-ray and SZ effect cluster cosmology, provide determinations of cluster temperatures bycombining the two detection techniques, and obtain independent cluster distances for many thousands ofclusters whose X-ray temperatures and shape parameters can be obtained from the X-ray survey. With ∼ clusters detected with PRISM , one can further exploit the eROSITA survey data by stacking in away similar to the analysis of the X-ray signals from the
ROSAT
All-Sky Survey for
SDSS detected clusters(Rykoff et al. 2008).
Other sub-millimeter/far-infrared initiatives:
Existing (
APEX, ASTE, IRAM 30m, LMT ) and future(
CCAT ) ground-based single-dish submillimeter observatories are not as sensitive above 300 GHz as
PRISM ,mainly because of the limitations of observing through the atmosphere. Interferometers (
ALMA, CARMA,PdB Interferometer, SMA ) are ill-suited to observing large fields. Moreover most interferometers are insen-sitive to large-scale structure.
SKA will span the radio range from 0.07 GHz up to 20 GHz, and will be theperfect complement to
PRISM , with more than 10 ( f sky / .
5) HI galaxies in a redshift range 0 < z < .
5, andmaps of the epoch of reionization above z ∼ PRISM will map the full-sky, large-scale continuum emission at higher sensitivities than ground basedsingle-dish telescopes. Bright compact sources found by
PRISM in its all-sky surveys can subsequently beobserved in more detail by interferometers. Observations can be combined to produce superior maps ofselected sky regions. The Atacama Large Millimeter Array (
ALMA ), operating in the range 30-1000 GHz,will complement
PRISM with follow-up of sources and clusters detected by
PRISM , mapping their structurein total intensity, polarization and spectral line at high angular and spectral resolution.
CCAT will initially have two imaging instruments. At low frequencies,
LWCam on CCAT will be ableto detect sources below the
PRISM confusion limit relatively quickly. However variations in atmospherictransmissivity and thermal radiation from the atmosphere make it difficult for
CCAT to map large scalestructures. At high frequencies,
SWCam will have difficulty mapping large areas to the confusion limit of
PRISM . Based on the specifications from Stacey et al. (2013),
CCAT can map an area of 1 square degree at857 GHz to a sensitivity of 6 mJy (the
PRISM confusion limit) within 1 hour. To map the entire southern skyto this same depth requires ∼
900 days (24h) with optimal observing conditions. Such large scale observationswill not be feasible with
CCAT. PRISM is needed to produce all-sky maps in these frequency bands.
PRISM will produce maps at the same resolution as
Herschel.
However
Herschel was able to map only a limitedportion of the sky.Few previous infrared telescopes have performed all-sky surveys in the bands covered by
PRISM . Akari was the last telescope to perform such observations, but the data is at much lower sensitivity and resolutionand is not yet publicly available. Several other prior telescopes (Spitzer, Herschel) as well as the airborneobservatory
SOFIA and the future mission
SPICA have observed or will observe in the 600-4000 GHz range,but only over very limited areas of the sky. Furthermore, except for a few deep fields, they observe objectsalready identified in other bands.
PRISM will be able to perform observations with sensitivities comparableto
Herschel or better, but covering the entire sky in many frequency bands.16 eferences [1] A. Avgoustidis, E.J. Copeland, A. Moss, et al.
Phys.Rev.Lett. , 107:121301, 2011.[2] R. Barkana and A. Loeb.
Phys. Repts. , 349:125-238,2001.[3] J. D. Barrow and P. Coles.
MNRAS , 248:52-57, 1991.[4] J. G. Bartlett and J. Silk.
ApJ , 353:399-405, 1990.[5] K. Basu, C. Hern´andez-Monteagudo, and R. A. Sun-yaev.
A&A , 416:447-466, 2004.[6] D Baumann et al.
AIP Conf.Proc. , 1141:10-120,2009.[7] C. L. Bennett, M. Halpern, G. Hinshaw, et al.
ApJS ,148:1-27, 2003.[8] M. Birkinshaw. Phys. Rep., 310:97-195, 1999.[9] F.R. Bouchet et al. arXiv:1102.2181 , 2011.[10] P. Bull and M. Kamionkowski.
ArXiv:1302.1617 ,2013.[11] C. Burigana. In G. L. Chincarini, A. Iovino, T. Mac-cacaro, & D. Maccagni, editor,
Observational Cos-mology , volume 51 of
Astronomical Society of thePacific Conference Series , 1993.[12] C. Burigana, L. Danese, and G. de Zotti.
A&A , 246:49-58, 1991.[13] B. J. Carr, K. Kohri, Y. Sendouda, and J. Yokoyama.
Phys.Rev.D , 81:104019, 2010.[14] R. Cen and J. P. Ostriker.
ApJ , 514:1-6, 1999.[15] A. Challinor and A. Lasenby.
ApJ , 499:1, 1998.[16] J. Chluba.
Spectral Distortions of the Cosmic Mi-crowave Background . PhD thesis, LMU M¨unchen,2005.[17] J. Chluba.
MNRAS , 402:1195-1207, 2010.[18] J. Chluba.
ArXiv:1304.6120 , 2013.[19] J. Chluba.
ArXiv e-prints: 1304.6121 , 2013.[20] J. Chluba and D. Grin.
ArXiv:1304.4596 , 2013.[21] J. Chluba and R. A. Sunyaev.
A&A , 458:L29-L32,2006.[22] J. Chluba and R. A. Sunyaev.
MNRAS , 419:1294-1314, 2012.[23] J Chluba, A Erickcek, and I Ben-Dayan.
Ap.J. , 758:76, 2012.[24] J. Chluba, A. L. Erickcek, and I. Ben-Dayan.
ApJ ,758:76, 2012.[25] J. Chluba, R. Khatri, and R. A. Sunyaev.
MNRAS ,425:1129-1169, 2012.[26] S. Chongchitnan and G. Efstathiou.
Phys.Rev.D , 72:083520, 2005.[27] R. M. Crutcher.
ARA&A , 50:29-63, 2012.[28] R. A. Daly.
ApJ , 371:14-28, 1991.[29] L. Danese and C. Burigana. In J. L. Sanz,E. Martinez-Gonzalez, and L. Cayon, editors,
Present and Future of the Cosmic Microwave Back-ground , volume 429 of
Lecture Notes in Physics,Berlin Springer Verlag , 1994.[30] L. Danese and G. de Zotti.
Nuovo Cimento RivistaSerie , 7:277-362, 1977.[31] J. B. Dent, D. A. Easson, and H. Tashiro.
Phys.Rev.D , 86:023514, 2012.[32] B. T. Draine. In T. Henning, E. Gr¨un, andJ. Steinacker, editors,
Cosmic dust: Near and far ,volume 414 of
Astronomical Society of the PacificConference Series , 2009.[33] B. T. Draine and B. Hensley.
ApJ , 765:159, 2013. [34] V. K. Dubrovich.
Sov. Ast. Lett. , 1:196, 1975.[35] J. L. Feng.
ARA&A , 48:495, 2010.[36] J. L. Feng, A. Rajaraman, and F. Takayama.
Phys.Rev.D , 68:063504, 2003.[37] D. J. Fixsen, E. S. Cheng, J. M. Gales, et al.
ApJ ,473:576, 1996.[38] J. Ganc and E. Komatsu.
Phys.Rev.D , 86:023518,2012.[39] J Ganc and E Komatsu.
Phys.Rev. , D86:023518,2012.[40] P. Goldreich and N. D. Kylafis.
ApJL , 243:L75-L78,1981.[41] Y. Gong, A. Cooray, M. Silva, et al.
ApJ , 745:49,2012.[42] M. J. Griffin et al.
A&A , 518:L3, 2010.[43] A. C. Hall and A. Challinor.
MNRAS , 425:1170-1184,2012.[44] P. Hennebelle and E. Falgarone. A&A Rev., 20:55,2012.[45] C. Hern´andez-Monteagudo and R. A. Sunyaev.
MN-RAS , 359:597-606, 2005.[46] C. Hern´andez-Monteagudo, L. Verde, andR. Jimenez.
ApJ , 653:1-10, 2006.[47] C. Hern´andez-Monteagudo, Z. Haiman, R. Jimenez,and L. Verde.
ApJL , 660:L85-L88, 2007.[48] C. Hern´andez-Monteagudo, J. A. Rubi˜no-Mart´ın,and R. A. Sunyaev.
MNRAS , 380:1656-1668, 2007.[49] W. Hu and J. Silk.
Phys.Rev.D , 48:485-502, 1993.[50] W. Hu and J. Silk.
Phys. Rev. Lett. , 70:2661, 1993.[51] W. Hu and N. Sugiyama.
ApJ , 436:456-466, 1994.[52] W. Hu, D. Scott, and J. Silk.
ApJL , 430:L5-L8, 1994.[53] W. Hu, D. Scott, and J. Silk.
Phys.Rev.D , 49:648-670, 1994.[54] A. F. Illarionov and R. A. Sunyaev.
AstronomicheskiiZhurnal , 51:1162-1176, 1974.[55] N. Itoh, Y. Kohyama, and S. Nozawa.
ApJ , 502:7,1998.[56] R. J. Ivison, A. M. Swinbank, I. Smail, et al.
ArXive-prints:1302.4436 , 2013.[57] K. Jedamzik, V. Katalini´c, and A. V. Olinto.
Phys-ical Review Letters , 85:700-703, 2000.[58] R. Khatri and R. A. Sunyaev.
JCAP , 6:038, 2012.[59] R. Khatri and R. A. Sunyaev.
JCAP , 9:016, 2012.[60] R. Khatri and R. A. Sunyaev.
ArXiv:1303.7212 ,2013.[61] R. Khatri, R. A. Sunyaev, and J. Chluba.
A&A , 540:A124, 2012.[62] R. Khatri, R. A. Sunyaev, and J. Chluba.
A&A , 543:A136, 2012.[63] W. H. Kinney.
Phys.Rev.D , 66:083508, 2002.[64] L. Kuzmin, G. Yassin, S. Withington, and P. Grimes.In A. Karpov, editor,
Eighteenth International Sym-posium on Space Terahertz Technology , 2007.[65] A. Lapi, J. Gonz´alez-Nuevo, L. Fan, et al.
ApJ , 742:24, 2011.[66] A. Lewis, J. Weller, and R. Battye.
MNRAS , 373:561-570, 2006.[67] K. Lochan, S. Das, and A. Bassi.
Phys.Rev.D , 86:065016, 2012.[68] J. C. Mather, E. S. Cheng, D. A. Cottingham, et al.
ApJ , 420:439-444, 1994.[69] P. McDonald, R. J. Scherrer, and T. P. Walker.
Phys.Rev.D , 63:023001, 2001.
70] J.-B. Melin, J. G. Bartlett, and J. Delabrouille.
A&A , 459:341-352, 2006.[71] F. Miniati, D. Ryu, H. Kang, et al.
ApJ , 542:608-621,2000.[72] M. Negrello, R. Hopwood, G. De Zotti, et al.
Science ,330:800, 2010.[73] S. P. Oh, A. Cooray, and M. Kamionkowski.
MN-RAS , 342:L20-L24, 2003.[74] J. P. Ostriker and C. Thompson.
ApJL , 323:L97-L101, 1987.[75] E. Pajer and M. Zaldarriaga.
Phys. Revs. Lett. , 109:021302, 2012.[76] E. Pajer and M. Zaldarriaga.
Phys.Rev.Lett. , 109:021302, 2012.[77] H. V. Peiris, E. Komatsu, L. Verde, et al.
ApJS , 148:213-231, 2003.[78] J. L. Pineda, W. D. Langer, T. Velusamy, and P. F.Goldsmith.
ArXiv e-prints , 2013.[79] Planck Collaboration, N. Aghanim, M. Arnaud, et al.
A&A , 536:A12, 2011.[80] Planck Collaboration, P. A. R. Ade, N. Aghanim,et al.
ArXiv e-prints:1212.4131 , 2012.[81] Planck Collaboration, P. A. R. Ade, N. Aghanim,et al.
ArXiv e-prints:1303.5077 , 2013.[82] Planck Collaboration, P. A. R. Ade, N. Aghanim,et al.
ArXiv e-prints:1303.5085 , 2013.[83] Planck Collaboration, P. A. R. Ade, N. Aghanim,et al.
ArXiv:1303.5084 , 2013.[84] Planck Collaboration, P. A. R. Ade, N. Aghanim,et al.
ArXiv:1303.5076 , 2013.[85] Planck Collaboration, P. A. R. Ade, N. Aghanim,et al.
ArXiv:1303.5081 , 2013.[86] Planck Collaboration XXII. arXiv:1303.5082 , 2013.[87] P. P. Ponente, J. M. Diego, R. K. Sheth, et al.
MN-RAS , 410:2353-2362, 2011.[88] B. A. Powell.
ArXiv:1209.2024 , 2012.[89] Y. Rephaeli.
ApJ , 445:33-36, 1995.[90] Y. Rephaeli and O. Lahav.
ApJ , 372:21-24, 1991.[91] D. A. Riechers, C. M. Bradford, D. L. Clements,et al.
Nature , 496:329-333, 2013.[92] J. A. Rubi˜no-Mart´ın, C. Hern´andez-Monteagudo,and R. A. Sunyaev.
A&A , 438:461-473, 2005.[93] J. A. Rubi˜no-Mart´ın, J. Chluba, and R. A. Sunyaev.
MNRAS , 371:1939-1952, (RMCS06), 2006.[94] J. A. Rubi˜no-Mart´ın, J. Chluba, and R. A. Sunyaev.
A&A , 485:377-393, 2008.[95] S. Y. Sazonov and R. A. Sunyaev.
ApJ , 508:1-5, 1998.[96] U. Seljak and M. Zaldarriaga.
ApJ , 538:57-64, 2000.[97] J. Smidt, A. Amblard, C. T. Byrnes, et al.
Phys.Rev.D , 81(12):123007, 2010.[98] K.M. Smith, A. Cooray, S. Das, et al.
AIP Confer-ence Proceedings , 1141:121, 2009.[99] R. A. Sunyaev and J. Chluba.
AstronomischeNachrichten , 330:657, 2009.[100] R. A. Sunyaev and Y. B. Zeldovich.
Astrophysics andSpace Science , 9:368-382, 1970.[101] R. A. Sunyaev and Y. B. Zeldovich.
A&A , 20:189,1972.[102] R. A. Sunyaev and Y. B. Zeldovich.
Comments onAstrophysics and Space Physics , 4:173, 1972.[103] H. Tashiro, E. Sabancilar, and T. Vachaspati.
Phys.Rev.D , 85:103522, 2012.[104] H. Tashiro, E. Sabancilar, and T. Vachaspati.
ArXiv:1212.3283 , 2012.[105] Planck HFI Core Team.
A&A , 536:A4, 2011.[106] T. Treu.
ARA&A , 48:87-125, 2010.[107] L Verde, H Peiris, and R Jimenez.
JCAP , 0601:019,2006.[108] Y. B. Zeldovich and R. A. Sunyaev.
Astrophysics andSpace Science , 4:301-316, 1969.[109] P. Zhang, U.-L. Pen, and H. Trac.
MNRAS , 355:451-460, 2004., 355:451-460, 2004.