5 Euclid & SKA Synergies 1 0 2 n a J 6 1 ThomasD. Kitching∗1, DavidBacon2, MichaelL.Brown3,Philip Bull4, JasonD. McEwen1,MasamuneOguri5, RobertoScaramella6,Keitaro Takahashi7, Kinwah ] O Wu1, DaisukeYamauchi8 C 1UniversityCollegeLondon,DepartmentofSpaceandClimatePhysics;2ICG,Universityof h. Portsmouth;3DepartmentofPhysics,UniversityofManchester;4UniversityofOslo,Institute p forTheoreticalAstrophysics;5GraduateSchoolofScience,UniversityofTokyo,Japan; o- 6INAF-OsservatoriodiRoma,I-00040MonteporzioCatone,Italy;7GraduateSchoolofScience r andTechnology,KumamotoUniversity,Japan;8ResearchCenterfortheEarlyUniverse, t s UniversityofTokyo,Japan; a [ E-mail: [email protected] 1 v Overthepastfewyearstwoofthelargestandhighestfidelityexperimentsconceivedhavebeen 8 approvedforconstruction:EuclidisanESAM-Classmissionthatwillmapthree-quartersofthe 7 9 extra galactic sky with Hubble Space Telescope resolution optical and NIR imaging, and NIR 3 spectroscopy,itsscientificaims(amongstothers)aretocreateamapofthedarkUniverseandto 0 . determine the nature of dark energy. The Square Kilometre Array (SKA) has similar scientific 1 0 aims (and others) using radio wavelength observations. The two experiments are synergistic 5 in several respects, both throughthe scientific objectivesand through the control of systematic 1 : effects.SKAPhase-1andEuclidwillbecommissionedonsimilartimescalesofferinganexciting v i opportunitytoexploitsynergiesbetweenthesefacilities. X r a AdvancingAstrophysicswiththeSquareKilometreArray June8-13,2014 GiardiniNaxos,Italy ∗Speaker. (cid:13)c Copyrightownedbytheauthor(s)underthetermsoftheCreativeCommonsAttribution-NonCommercial-ShareAlikeLicence. http://pos.sissa.it/ Euclid&SKASynergies ThomasD.Kitching 1. Introduction In this Chapter we will describe the ESA Euclid mission, and discuss some of the synergies that Euclid has with the SKA. Euclid probes the low redshift Universe, through both weak lens- ing and galaxy clustering measurements. The SKA has the potential to probe a higher redshift regime and a different range in scales of the matter power spectrum, linear scales rather than the quasi-non-linear scalesthatEuclidwillbesensitiveto,thatwillmakethecombination particularly sensitive tosignatures ofmodified gravity and neutrino mass. Incombination alonger baseline in redshiftwillimproveexpansionhistoryandgrowthofstructuremeasurementsleadingtoimproved measurements oftheredshift behaviour ofthedarkenergy equation ofstate. Thecrosscorrelation between Euclid and SKA weak lensing (shape as well as size and flux magnification), Euclid and SKA galaxy clustering (BAO and redshift space distortions), SKA 21cm intensity mapping and Planckdata(CMB,SZandISW)willprovideamultitude ofcross-correlation statistics. As an example of the benefit of combining the experiments for systematic control for the primary science, the shape measurement from Euclid and SKA will be affected by systematics in different ways, meaning that a cross-correlation of the weak lensing data from both data sets will be less prone to shape measurement biases (see Brown et al., in this volume). Furthermore the intrinsic (un-lensed) ellipticityalignmentsmeasuredbybothEuclidandtheSKAwillhavemutual benefit. Specifically, in addition to Euclid, for weak lensing SKA contributes i) depth, providing more source counts for weak lensing, ii) a way to cross-calibrate shape systematics, iii) a way to cross-calibrate intrinsic alignments, and iv) precise redshifts through 21cm line observations. Andpossiblytheuseofpolarisation and21cmrotational velocitiesforintrinsicshapesandlensing tomography ofhighredshift21cmfluctuations. Inaddition totheprimary science objectives ofEuclidthere isaplethora ofadditional legacy andcosmologysynergies. Wecanexpect>105 stronglensdetectionsfromEuclidandSKA,many with redshifts for the lenses and sources and high-resolution images. The combination of Euclid andSKAwillbetterinvestigatetheobscuredcosmicstarformationhistoryasafunctionofredshift and environment thanks to their combination of areal coverage and sensitivity. Forpopulation III, hypernovae andTypeIIsupernovae NIRemissionwillbedetectable byEuclidfromsupermassive PopIIIsupernovaeindense107-108 M haloesouttoz=10−15,andradiosynchrotron emission ⊙ fromPopIIIsupernovaremnantsdetectable bySKAouttoz=20. A further synergistic aspect is that each of these areas will require post-operation infrastruc- ture development in terms of a very large number of hydrodynamical and N-body simulations to understand the effects of non-linear clustering and baryonic feedback. Finally the data analysis of giga-scale catalogues, peta-bytes ofdata, and post-operation simulations require astrostatistical andastroinformatics synergies forastronomers toaccessandvisualise thesedatasets. 1.1 EuclidOverview Euclid1 is one of the European Space Agency’s (ESA) medium (M) class missions that has beenselectedaspartofESAs“cosmicvisions”programme. EuclidwasselectedbyESAinOctober 2012, totakethesecondoftheM-classmissionplaces, M2,whichmeansthatthereisascheduled 1http://euclid-ec.org 2 Euclid&SKASynergies ThomasD.Kitching launch date for 2020. The science objective of Euclid is primarily to determine the nature of the the phenomenon that is causing the expansion rate of the Universe to accelerate; so-called ‘dark energy’. Howeverthetop-level science objectives areinfact four-fold and coverallcurrent major openquestions incosmology: • Dynamical Dark Energy: Is the dark energy simply a cosmological constant, or is it a field thatevolvesdynamically withtheexpansion oftheUniverse? • Modification of Gravity: Alternatively, is the apparent acceleration instead a manifestation of a breakdown of General Relativity on the largest scales, or a failure of the cosmological assumptions ofhomogeneity andisotropy? • DarkMatter: Whatisdarkmatter? Whatistheabsolute neutrino massscaleandwhatisthe numberofrelativistic speciesintheUniverse? • Initial Conditions: What is the power spectrum of primordial density fluctuations, which seeded large-scale structure, andaretheydescribed byaGaussianprobability distribution? More quantitatively the objective of Euclid is to achieve a dark energy Figure of Merit (Albrecht etal.2006)of≥400andtodeterminethegrowthindexg ,thatprovides asimpleparameterisation for deviations from general relativity, to an accuracy of 0.02. Euclid is designed to achieve these science objectives with support from ground-based observations. In combination with the SKA it is expected that the combined strength of these experiments will enable both to far exceed their singular goals. Inordertoachievethesescienceobjectives Euclidwillusetwoprimarycosmolog- ical probes: weaklensing andgalaxy clustering. Thesearegivenequal priority inthemission and indeed it is only through the combination of weak lensing and galaxy clustering that the science objectives can be achieved. In addition Euclid will provide the astronomy community with a rich datasetthatwillenablemanyastrophysical studies;areasinEuclidthatarereferredtoas‘Legacy’ science. The SKA science goals are synergistic with those of Euclid. Whilst Euclid is a focussed ex- periment to address a particular goal, the SKA design allows for flexibility in the type of science questions that canbeaddressed. Inthiscase proposals forSKAobservations can betailored tobe inthebestsynergy withEuclid. TheEuclidmissionhasbeendesignedgivenparticularconstraints on the telescope size and overall cost of the mission and instrumentation. An optimisation during the Euclid Assesment Phase (Laureijs 2009) that wasrefined and elaborated during theDefinition Phase(Laureijsetal.2011)ofthemission. ThisresultedinanominaldesignofEuclidbeinga1.2 meter Korsch, 3 mirror anasigmat telescope. There will be two instruments on board: the VISi- ble focal plane instrument (VIS) (Cropper et al. 2012) and the Near Infrared SpectroPhotometric (NISP)instrument. VIS will provide high-resolution optical imaging over a field of view of size 0.787×0.709 square degrees, with a resolution of 0.18 arseconds, over a single broad-band wavelength range of 550-900 nm (an optical band equivalent to an RIZ filter, although Euclid in fact does not have an optical filter on board). It will consist of 36 4k×4k CCDs. The primary purpose of the VIS instrument is to provide imaging to enable the measurement of galaxy ellipticities to sufficient 3 Euclid&SKASynergies ThomasD.Kitching accuracy and precision for use in weak lensing. It will image approximately 1.5 billion galaxies withalimitingR-bandmagnitude of24.5(10s extendedsource). NISP has a field of view of a similar size to VIS of 0.763×0.722 square degrees, these are matched toenable simultaneous/matched observations of the sky. Thefocal plane willcontain 16 2k×2k HgCdTe (“Mer-Ca-Tel”) detectors. There will be 3 NIR filters, that will enable imaging, and 2 grism spectroscopic elements, that will enable slitless spectroscopy (with an approximate resolution ofR=250). Theimaging isdesigned toenablephotometric measurements ofthesame galaxiesobservedusingtheVISinstrumentforthedeterminationofphotometricredshiftsforweak lensing. The spectroscopy is designed to enable precision redshifts to be determined for galaxy clustering measurements. Inorder toachieve thephotometric redshift required forthe weaklensing science Euclid will use ground-based optical imaging data from DES and KiDS, and any other available surveys; re- quiring normal broad-band imaging over the VISwavelength range. Spectroscopic redshifts from ground-based surveyssuchasBOSS,DESIandMOONSwillalsobeingestedandusedforphoto- metricredshift calibrations. Every aspect of Euclid is designed using a systems engineering approach where the science requirements aretranslated intoprogressively moredetailedrequirementsonsurvey,telescopeand instrumentdesign,aswellasrequirementsonthealgorithmsusedinthedataprocessing. Thisflow of requirements is described in a series of ESA documents (for example the Euclid SciRD) and relatedpublications (forexample(Cropperetal.2013)). Withregardtoinstrumentdesigns, Euclid and the SKAare therefore synergistic inthe sense that Euclid willobserve in the optical and NIR wavelengths and SKAwill observe at longer radio wavelengths. Aswewill discuss inthis article this enables unique scientific synergies to be exploited when combining the data between the two experiments. 1.2 SurveySynergies The Euclid primary probes, weak lensing and galaxy clustering, will be carried out over the same area of sky in a wide-field survey of 15,000 square degrees. As described in Laureijs et al. (2011)theareaisdrivenbyanoptimisationbetweenthegalaxyclustering(thatprefersawiderarea forafixedobserving time)andweaklensing(thatcanprefersashallowerdeepersurveyforintrin- sic alignment mitigation), aswellas efficiency insurvey design whereby the lowest 30degrees in ecliptic latitude are not observed to avoid zodiacal light contamination. In Figure 1 we show the Euclid reference survey from Amiaux et al. (2012) and how this builds as a function of time. It is self-evident that the areal coverage of Euclid and SKAis synergistic with both experiments ex- pectedtoobservesignificantfractionsoftheextragalactic sky. ThesynergyoftheEuclidandSKA experiments in time is also self-evident. Euclid is a nominally 6 year mission that is scheduled to beginobservationsin2020,SKAisscheduledtobeginconstructionin2018andbeginobservations onasimilartimescaletoEuclid. The depth of the Euclid and SKA surveys is also synergistic, Euclid will have a limiting magnitude in optical wavelengths that will mean it is sensitive to galaxies with redshifts over the range0<z<2,andtheNIRspectroscopy willbesensitivetogalaxiesatslightlyhigherredshifts. In Figure 2 we show as an example the normalised distribution of weak lensing galaxies that the SKAphase1andthefullSKA,over3p steradians, andtheEuclidweaklensingsurveywillcover. 4 Euclid&SKASynergies ThomasD.Kitching Figure 1: From Amiaux et al. (2012). Euclid Reference Survey constructionfrom year 1 (top) to year 6 (bottom)incylindrical(left)andorthographicprojection. Intheleftpanel,onlythewidesurveyfieldsare shown. Intherightpanel,thecalibrationfieldsarealsodisplayed. The Euclid weak lensing survey will have 30 galaxies per square arcminute with size larger than 1.5 times the PSF (R2) and magnitude RIZ ≤ 24.5. The SKA survey area and number density has several possibilities, with various area and redshift overlap scenarios (see continuum survey overview,alsoBrownetal. andJarvisetal. inthisvolume),forexampleforthefullSKA(orSKA phase2)therecouldbeeither • 3p steradian survey: RMSnoise=0.5micro-Jy, 10galsarcmin−2. • 5000sq. degsurvey: RMSnoise=0.2micro-Jy, 23galaxiesarcmin−2. • 1000sq. degsurvey: RMSnoise=0.1micro-Jy, 37galaxiesarcmin−2. In the remainder of this article we will discuss the scientific synergies that instrument and surveydesignsofEuclidandtheSKAenable. 2. CosmologicalSynergies The primary objectives of Euclid are cosmological in nature, in synergy with SKA these ob- jectivescanbesupplemented andextended. BoththeSKAandEuclidwillbeabletomeasure: • Weaklensingshearg, 5 Euclid&SKASynergies ThomasD.Kitching 1 SKA−1 SKA−1 Early 0.9 Full SKA Euclid Weak Lensing 0.8 0.7 z)0.6 p( d e s0.5 ali m or0.4 N 0.3 0.2 0.1 0 0 0.5 1 1.5 2 2.5 3 Redshift z Figure2:ThenormaliseddistributionsofgalaxiesusableforweaklensinganalysesinEuclidandtheSKA, forthedifferentphases. ThenumbersforSKAarederivedfromspecificationsquotedintheSKAImaging Science Performance memo which is in turn based on the SKA-1 Baseline Design and from the SKADS simulationsofWilmanetal.(2008). TheinstrumentofchoiceisSKA-MIDandithasbeenassumedBand 2(950-1760MHz)hasbeenusedwhich,underthecurrentversionoftheSKA-1baselinedesigngivesthe bestperformanceintermsofsensitivityattherequiredangularresolution. • Weaklensingmagnification; including sizesandflux(n(z)), • Galaxypositions, bothphotometric q ,andspectroscopic q . p s Thisresultsin102-pointobservables,powerspectra,intotalthatcanbecombinedcombinatorially to produce 55 cross and auto-correlation statistics. If, for example, a ‘tomographic’ approach is pursued and the galaxy populations are split into >10 redshift bins then this results in over 5000 powerspectra thatcould becomputed including allinterandintra-bin correlations. Someofthese correlations canbeusedtoreduced systematic effects, forexamplethecorrelation ofgalaxyshear withgalaxypositioncanbeusedininfertheintrinsicalignmentsystematicincosmicshearstudies, see for example Joachimi etal. (2011), orthe correlation between weak lensing shear inthe radio and optical data sets can be used to reduce ellipticity measurement systematics. Some of these correlations canbeused toincrease the statistical precision ofthecombined data, forexample the combination of weak lensing shear and size (Heavens et al. 2013). The SKA can also measure unresolved 21cm intensity to create maps that can then be included as an additional cosmological probetobeincluded inthesynergistic combination ofprobes. Wewillhighlight justafewofthesecombinations asexamplesinthischapter. 2.1 21cmIntensityMapping Theprincipaladvantageof21cmintensitymapping(IM;seeSantosetal.,inthisvolume)over 6 Euclid&SKASynergies ThomasD.Kitching traditional galaxy redshift surveys is that extremely large volumes can be surveyed in a relatively short time; for example, Phase I of the SKA will be capable of surveying a total area of ∼30,000 deg2 from z=0 to 3 over the course of 1-2 years. This ability stems from the modest resolution requirements of the IM method; there is no need to resolve individual galaxies, and only the in- tegrated HI emission on comparatively large angular scales matters. We refer to the Chapters on Intensity MappingandRSDwithSKAformoredetailsofthissciencesynergy. Atitsmostbasic,alargeIMsurveywithaPhaseISKAarraywouldcomplementEuclidsim- plybyincreasingthetotalvolumebeingprobed. ConsiderasituationinwhichEuclidandtheSKA targeted independent survey volumes. The sensitivity of SKA-IM to the first BAO acoustic peak wouldbecomparabletothatofEuclid,whichispracticallycosmicvariance-limited atthesescales anyway; doubling the survey volume would therefore have a significant effect in beating down cosmicvarianceandincreasing theprecision ofBAOdistance indicators andotherobservables. Preventing the surveys from overlapping would spoil a number of interesting opportunities, however. Probing the same volume with two almost cosmic variance-limited experiments is not redundant; as well as providing useful cross-checks for consistency between the two, one can also benefit from the ‘multi-tracer’ effect (see Section 2.3), whereby one can continue to gain information aboutsomeobservables inspiteofcosmicvarianceaslongasdifferent populations of tracers can be distinguished by each survey. Thiscan be used to enhance the precision ofredshift spacedistortion measurements, forexample,whichprobethegrowthofstructure. SKA IM surveys are capable of significantly extending the redshift range of Euclid, which covers0<z<2. By‘fillingin’redshiftsmissedbythegalaxysurvey,onecangainagreatdealof leverage onkeycosmological functions suchastheequation ofstateofdarkenergy andthelinear growth rate; and could also be used for cross-correlations to get better photometric redshifts for Euclid. Figure 3 shows the joint constraints that can be achieved on w and the growth index, g , 0 throughthecombinationofEuclidandanSKA-IMsurveywithmuchwiderredshiftcoverage. The additional information atlowredshift helps topin downthe evolution ofdarkenergy andpossible deviations from the standard growth history, while constraints at higher redshift act as a useful ‘anchor’, bylocating thetransition fromthematter-dominated eraandputting limitsonthespatial curvature, W . K 2.2 NeutrinoPhysics Measuringneutrinoproperties ispotentially aparticularly interesting areaofEuclid-SKAsci- ence. Massive neutrino suppress the growth of structure via free-streaming effects introducing an effective pressure thatacts inregions ofhigher density. Thiscontinues until theneutrinos become non-relativistic. Thistransitionimprintsafeatureataparticularscaleinthematterpowerspectrum. AsshowninJimenezetal.(2010)anall-skyEuclid-like surveycombinedwithagalaxyclustering survey,suchasthatcouldbeachievedbythefinalSKA,mayevenbeabletodeterminetheneutrino hierarchy. 2.3 Multi-TracerMethod TheSKAandEuclidwillobserveahugenumberofgalaxies,andtheerrorsinpowerspectrum of galaxies will be dominated by cosmic variance, rather than shot noise, at cosmological scales. 7 Euclid&SKASynergies ThomasD.Kitching Figure 3: Predicted 1s , 2 parameter, error contours for dark energy (w , w ), modified gravity (g ), and 0 a initialcondition(n )parametersforEuclidgalaxyclusteringandSKAphase1design. s 8 Euclid&SKASynergies ThomasD.Kitching This is especially serious when we try to constrain primordial non-Gaussianity whose effect is strongeratlargerscales. Cosmicvariancecouldbeavoidedwithathe‘multi-tracer’methodSeljak (2009) which uses multiple tracers of the dark matter distribution with different biases to cancel outsamplevariance. Althoughpowerspectraoftracersthemselvesarelimitedbycosmicvariance, the ratio of the power spectra of two tracers, which represents the relative bias, can evade cosmic variance and islimitedonly byshot noise. Because the massandredshift dependences ofbiasare affected bynon-Gaussianity (f ),itcanbeconstrained bythemeasurements ofrelativebiases. NL Thismulti-tracer methodiseffectivewhenthebiasdifference, hencemassdifference, islarge between tracers and it is critically important to estimate the mass of the dark matter halo hosting eachgalaxy. Adeepsurveyisalsoimportant because biasevolves rapidlywithredshift. TheSKA andEuclidsurveyswillhavedifferentredshift-distributions ofobservedgalaxiessothattheircom- bination enhances the power of multi-tracer method. Yamauchi et al. (2014) studied the potential ofcombination oftheSKAcontinuum surveyandEuclidphotometric surveyfortheconstraint on f . The SKA continuum survey reaches much further than the Euclid photometric survey, pro- NL viding a larger redshift range in combination, while the number of galaxies observed by Euclid is larger than that by the SKAat low redshifts, so they are complementary to probe the evolution of bias. Figure 4 shows expected constraints on f from Euclid, SKA phase 1, the full SKA and NL their combinations. Here, it is assumed that galaxies observed by Euclid have photometric red- shifts while SKA cannot obtain redshift information (this is a very conservative assumption, but see (Camera et al. 2014) for gains made by using optical and near IR photometric redshifts). It is seenthattheconstraint on f canreachbelowunityandapproach O(0.3). NL 2.4 MagneticFields Thetwopressingquestions regardingcosmicmagnetismare: whatistheoriginofthecosmo- logical magnetic fields and how do these magnetic fields affect (orwere affected by) the structure formation and evolution of the Universe? In a more practical perspective, we would like to know how the cosmological magnetic fields imprint observational signatures and how we can make use ofthesesignatures tobetterunderstand thestructural formationprocessinourUniverse. Charged particles inthepresence ofamagnetic fieldexperience aLorentz force. Assuch the presence ofaprimordial magnetic fieldwouldproduce andamplifydensity fluctuations inionised gas (the baryons). Through gravitational coupling between the baryons and the dark matter, the magneticfieldwouldimprintsignaturesinthedarkmatterdensitydistributions. Asmagneticfields alter the matter density power spectrum and the halo mass function, the magnification bias and the redshift distribution of galaxies are distorted (see Fedeli & Moscardini (2012); Camera et al. (2014)). The capability of Euclid weak lensing studies (and SKA weak lensing studies) together withtheFaradayrotationmeasurementsusingtheSKAwillbeabletoquantifysuchamechanism, which will not only set constraints on the strength ofthe magnetic fields inthe early Universe but also determines the role of magnetic interaction in cosmological structure formation, which has beenassumedtobesolelygravitational. Gravitational lensing, bothweakandstrong,arepowerfultoolsinthestudiesofcosmological large-scale structures. Lensing calculations rely onray tracing inageometrical framework setus- ing the theory of relativity, however absorption and scattering of photons emitted from the distant lensed sources along theray areusually ignored. Thismaywellbeasensible firstapproximation, 9 Euclid&SKASynergies ThomasD.Kitching 5-tracers 2 -- Euclid : equal shot noise -- SKA : galaxy type 1 ) L N f ( s 0.5 0.25 Euclid SKA1Euclid+SKA1SKA2Euclid+SKA2 Figure 4: Expected constraints on f from the SKA continuum survey, Euclid photometric survey and NL theircombinationsYamauchietal.(2014). but if there are ongoing activities, such as merging of clusters and/or violent large-scale galactic outflows, the situation could be more complicated. Photons on different rays originating from a lensed source could therefore have different frequency dependent attenuation, through absorption and/or scattering by the matter along the rays. Moreover, if there is a spatially inhomogeneous magnetic field permeating the lens and its surrounding, polarisation fluctuations could also be in- duced in the lensed images. This potential frequency-dependent distortion has not beinvestigated indetail,howeveritcouldbeinferredthroughmulti-wavelengthpolarisationobservationsusingthe SKA.Withtheinformationofthematterdistributionandthemagneticfieldalongthelineofslight, frequency-dependent attenuation of photons can be modelled using radiative transfer calculations andhenceprescriptions forcorrecting thecorresponding distortion andbiasescouldbederived. 3. SystematicSynergies The combination of power spectra discussed in the previous section will enable a mitigation systematic to be performed at the statistical level. However because Euclid and the SKA will observe many of the same galaxies one can directly compare the measurements to gain an ever largeradvantage. WereferheretosystematicreductionsdiscussedintheChapteronWeakLensing and Synergies, and highlight here one particular aspect that is enabled through the weak lensing 10