Probing the nature of Dark Matter with the SKA 5 1 0 2 SergioColafrancesco∗1, MarcoRegis2, PaoloMarchegiani1,GeoffBeck1,Rainer n Beck3, HannesZechlin2, AndreiLobanov3, DieterHorns4 a J 1SchoolofPhysics,UniversityoftheWitwatersrand,Johannesburg,SouthAfrica 2 2DipartimentodiFisica,UniversitàdegliStudidiTorinoandINFN-SezionediTorino, viaP.Giuria,1,10125Torino,Italy ] E 3Max-Planck-InstitutfürRadioastronomie,AufdemHügel69,53121Bonn,Germany H 4InstitutfürExperimentalphysik,UniversitätHamburg,LuruperChaussee149,22761Hamburg, h. Germany p - o E-mail: [email protected] r t s Dark Matter (DM) is a fundamentalingredientof our Universe and of structure formation, and a [ yetitsnatureiselusivetoastrophysicalprobes.Informationonthenatureandphysicalproperties 1 oftheWIMP(neutralino)DM(theleadingcandidateforacosmologicallyrelevantDM)canbe v obtainedbystudyingtheastrophysicalsignalsoftheirannihilation/decay.Amongthevariouse.m. 8 3 signals, secondaryelectronsproducedbyneutralinoannihilationgeneratesynchrotronemission 7 inthemagnetizedatmosphereofgalaxyclustersandgalaxieswhichcouldbeobservedasadiffuse 3 0 radio emission (halo or haze) centered on the DM halo. A deep search for DM radio emission . 2 withSKAinlocaldwarfgalaxies,galaxyregionswithlowstarformationandgalaxyclusters(with 0 offsetDM-baryonicdistribution,likee.g. theBulletcluster)canbeveryeffectiveinconstraining 5 1 theneutralinomass,compositionandannihilationcross-section. Forthecaseofadwarfgalaxy, : v likee.g.Draco,theconstraintsontheDMannihilationcross-sectionobtainablewithSKA1-MID Xi will be at least a factor ∼103 more stringent than the limits obtained by Fermi-LAT in the g - r rays. These limits scale with the value of the B field, and the SKA will have the capability to a determine simultaneously both the magnetic field in the DM-dominatedstructures and the DM particleproperties. TheoptimalfrequencybandfordetectingtheDM-inducedradioemissionis around∼1GHz,withtheSKA1-MIDBand1and4importanttoprobethesynchrotronspectral curvatureatlow-n (sensitivetoDMcomposition)andathigh-n (sensitivetoDMmass). AdvancingAstrophysicswiththeSquareKilometreArray, June8-13,2014 GiardiniNaxos,Sicily,Italy ∗Speaker. (cid:13)c Copyrightownedbytheauthor(s)underthetermsoftheCreativeCommonsAttribution-NonCommercial-ShareAlikeLicence. http://pos.sissa.it/ UnveilingDarkMatterwithSKA SergioColafrancesco 1. Unveiling thenature ofDarkMatter The Square Kilometre Array (SKA) is the most ambitious radio telescope ever planned, and it is a unique multi-disciplinary experiment. Even though the SKA, in its original conception, has been dedicated to constrain the fundamental physics aspects on dark energy, gravitation and magnetism,muchmorescientificinvestigationcouldbedonewithitsconfiguration: theexploration ofthenatureofDarkMatterisoneofthemostimportantadditional scientificthemes. Among the viable competitors for having a cosmologically relevant DM species, the leading candidate is the lightest particle of the minimal supersymmetric extension of the Standard Model (MSSM, Jungman et al. 1996), plausibly the neutralino c , with a mass Mc in the range between afewGeVtoseveralTeV.Information onthenatureandphysical properties oftheneutralino DM can be obtained by studying the astrophysical signals of their interaction/annihilation in the halos ofcosmicstructures. Thesesignals(seeColafrancesco etal. 2006,2007fordetails)involve,inthe case of a c DM, emission along a wide range of frequencies, from radio to g -rays (see Fig. 1 for aDMspectral energydistribution (SED)inatypicaldwarfgalaxy). Neutralpionsproducedin cc annihilation decay promptly in p 0 →gg and generate most of the continuum photon spectrum at energies E &1GeV.Secondaryelectronsareproducedthrough variouspromptgeneration mecha- nisms and by the decay of charged pions, p ± →m ±+n m (n¯m ), with m ± →e±+n¯m (n m )+n e(n¯e). The different composition of the cc annihilation final state will in general affect the form of the electron spectrum. The time evolution of the secondary electron spectrum is described by the transport equation: ¶ n ¶ e =(cid:209) [D(cid:209) n ]+ [b (E)n ]+Q (E,r), (1.1) ¶ t e ¶ E e e e where Q (E,r) is the e± source spectrum, n (E,r) is the e± equilibrium spectrum (at each fixed e e time) and b ≡−dE/dt isthe e± energy loss per unit time, b =b +b +b +b (see e e ICS synch brem Coul Colafrancesco et al. 2006 for details). The diffusion coefficient D in eq.(1.1) sets the amount of spatial diffusion for the secondary electrons: it turns out that diffusion can be neglected in galaxy clusters while it is relevant on galactic and sub-galactic scales (see discussion inColafrancesco et al. 2006,2007). Undertheassumptionthatthepopulationofhigh-energy e± canbedescribedbya quasi-stationary (¶ n /¶ t≈0)transportequation,thesecondaryelectronspectrumn (E,r)reaches e e its equilibrium configuration mainly due to synchrotron and ICS losses at energies E &150 MeV and due to Coulomb losses at lower energies. Secondary electrons eventually produce radiation by synchrotron emission inthe magnetized atmosphere of cosmic structures, bremsstrahlung with ambient protons and ions, and ICS of CMB (and other background) photons (and hence an SZ effect, Colafrancesco 2004). These secondary particles also heat the ambient gas by Coulomb collisions. A large amount of efforts have been put in the search for DM indirect signals at g -ray en- ergies looking predominantly for two key spectral features: the p 0 → gg decay spectral bump , and the direct cc →gg annihilation line emission, with results that are not conclusive yet (e.g., Daylan et al. 2014, Weniger 2012, Doro etal. 2014). The non-detection of signals related to DM annihilation/decay from various astrophysical targets (including observations of dwarf spheroidal galaxies, the Galactic Center, galaxy clusters, the diffuse gamma-ray background emission) is in- terpreted intermsofconstraints onthe(self-)annihilation cross-section (ordecaytime)oftheDM 2 UnveilingDarkMatterwithSKA SergioColafrancesco -10 -10 10 10 ] ] 2-1 s 10-11 Mc = 100 GeV, s v at EGRET level 2-1 s 10-11 Mc = 100 GeV, s v at EGRET level -m B = 1 m G p 0 -m rg c10-12 m rg c10-12 B = 4 m G e e m ) [ 10-13 ) [ 10-13 2 r IC on r -30 s10-14 synch. CMB -30 s10-14 1 = 1 -15 = 1 -15 0.5 10 10 nDWI(, 10-16 IC on nDWI(, 10-16 0.1 -17 -17 n 10 starlight n 10 -18 -18 10 10 7.5 10 12.5 15 17.5 20 22.5 25 27.5 7.5 10 12.5 15 17.5 20 22.5 25 27.5 log ( n [ Hz ] ) log ( n [ Hz ] ) Figure 1: Left. The Draco dSph multi-wavelength spectrum for a 100 GeV WIMP annihilating into bb¯. Right. TheeffectofvaryingthemagneticfieldstrengthontheDracomulti-wavelengthspectrumfora100 GeVWIMPannihilatingintobb¯. TheWIMPpairannihilationratehasbeentunedastogiveag -raysignal attheleveloftheEGRETmeasuredfluxupperlimit(fromColafrancescoetal. 2007). particle candidate. Assuming, for instance, a canonical WIMP of Mc =100GeV annihilating to b-quarks, stacked observations of dwarf spheroidal galaxies with Fermi-LAT put a constraint of hs Vi<2×10−25cm3s−1 (95% c.l.) on the thermally-averaged self-annihilation cross-section of DM particles (Ackerman et al. 2014). Hopes of discovering annihilating WIMPs in g -rays are relegated toFermi-LATsuccessors andtheforthcoming CTAexperiment (Doroetal. 2013). Thereare,however,alsogoodhopestoobtainrelevantinformationonthenatureofDMfromradio observations ofDMhalosonlargescales, i.e. fromdwarfgalaxiestoclustersofgalaxies. 2. Radio emissionsignalsfrom DarkMatter annihilation Observations of radio halos produced by DM annihilations are, in principle, very effective in constraining the neutralino mass and composition (see, e.g., Colafrancesco and Mele 2001, Colafrancescoetal. 2006,2007),underthehypothesisthatDMannihilationprovidesanobservable contribution totheradio-halo flux. Thewiderangeoffrequencies probedbytheSKAandthevarietyofachievable observational targets (and in turn of magnetic fields) will allow testing the non-thermal electron spectrum from about 1 GeV to few hundreds of GeV, that is the most relevant range in the WIMP search. Fig- ure2displays thepredicted spectral differences between variousannihilation channels andWIMP masses: notethatthesedifferences manifestmainlyinlow-n slopevariationfordifferingannihila- tion channels and inthehigh-n spectral flattening/steepening forlarger/smaller masses. TheSKA sensitivity curvesforSKA1-LOWandSKA1-MIDaretakenfromDewdneyetal. (2012). The surface brigthness produced by DM-induced synchrotron emission is heavily affected by dif- 3 UnveilingDarkMatterwithSKA SergioColafrancesco 102 102 101 bb 101 bb60 100 τ+τ− 100 bb500 1100−−21 bb 1015M⊙ 1100−−21 bb60 1015M⊙ 10−3 τ+τ− 10−3 bb500 Jy() 1100−−54 1012M⊙ Jy() 1100−−54 1012M⊙ ) ) ν( 10−6 ν( 10−6 F 10−7 bb F 10−7 bb60 SKA SKA 1100−−98 τ+τ− 107M⊙ 1100−−98 bb500 107M⊙ 10−10 10−10 10−11 10−11 10−12 10−12 101 102 103 104 105 101 102 103 104 105 ν(MHz) ν(MHz) Figure 2: Flux densitiesfor dwarfgalaxies(M=107 M ), galaxies(M=1012 M ), andgalaxyclusters ⊙ ⊙ (M =1015 M ). The halo profile is NFW, hBi=5 m G, and the annihilation cross section is fixed to the ⊙ valuehs Vi≈3×10−27 cm3 s−1. Black linesarefordwarfgalaxies, redlinesforgalaxiesandgreenfor clusters. Left: WIMPmassis60GeV,solidlinesareforcompositionbbanddashedlinesfort +t −. Right: Compositionisbb,solidlinesareforWIMPmass60GeVanddashedlinesfor500GeV.Dash-dottedlines areSKAsensitivitylimitsforintegrationtimesof30,240and1000hours(Dewdneyetal. 2012). Theflux iscalculatedwithinthevirialradius(fromColafrancescoetal. 2014). fusioninsmallscalestructures, e.g.,dwarfandstandardgalaxies,whileitislessimportantinlarge structures, e.g.,galaxyclusters(seeColafrancesco etal. 2006-2007 foradetailed discussion). Polarization from DM-induced radio emission isexpected atverylowfractional levelsdueto thefactthatDMspatial andvelocity distribution isnearlyhomogeneous andthatDMannihilation is mediated by secondary particle production. Therefore, a low polarization level of detectable radio signals in the directions of DM halos would be consistent with the DM origin of such radio emission. Residual high-polarization signals could be hence attributed to astrophysical sources in the direction or within the DM halos, and one could use these signals to infer properties of the magneticfieldinthesestructures (seeR.Becketal. 2015,andF.Govonietal. 2015). 2.1 Cosmological evolutionofDarkMatterradioemission Figure 3 displays the evolution of radio emission from DMhalos of mass 107 M , 1012 M , ⊙ ⊙ and 1015 M for a constant magnetic field of 5 m G, in accordance with arguments made in Co- ⊙ lafrancesco etal. (2014). Theannihilation channel isbb,themassoftheneutralino is60GeVand a DM annihilation cross-section hs Vi=3×10−27 cm 3 s−1 was adopted. Emission from dwarf galaxy halos (M ≈107 M ) is just below the SKA detection threshold for this value of hs Vi but ⊙ would be visible at redshift z≤0.01 for the assumed DM annihilation cross-section. This justify the search of DM-induced radio signals mainly indwarf galaxies of the local environment, at dis- tances ∼<3 Mpc (z ∼<0.0007). Emission from galactic DM halos (M ≈1012 M⊙) are detectable by SKA out to z≈0.8 even with the reference value of hs Vi, and can provide a non-detection upper-bound onhs Vianorderofmagnitude belowtheassumedvalueevenatsuchhighredshifts. Emissionformgalaxyclusterhaloscanprovidesimilarconstraintsbutouttohigherredshiftsz∼<3. These objects thus offer theoption ofdeep-field observations thatcanscan alarger fraction ofthe DMparameterspacethanthebestcurrentdata. 4 UnveilingDarkMatterwithSKA SergioColafrancesco Figure 3 also shows that the effect of diffusion are far less significant when observing higher-z objects, againsimplifying themodelling andanalysis oftheSKAobservations. 10−3 102 10−1 FνJy()()111111000000−−−−−−111531975 zzz===310...0001 SKA FνJy()()1111100000−−−−−119753 zzz===310...0001 SKA FνJy()() 111100001−−−−086420 zzz===310...0001 SKA 10−17 10−13 10−10 10−19 z=5.0 10−15 z=5.0 10−12 z=5.0 10−21 10−17 10−14 10−21301 102 103 104 105 10−11901 102 103 104 105 10−11601 102 103 104 105 ν(MHz) ν(MHz) ν(MHz) Figure 3: Flux densities for various DM halos within the virial radius. Left: dwarf spheroidal galaxies (M =107 M ); mid panel: galactic halos (M =1012 M ); right panel: galaxy clusters (M =1015 M ). ⊙ ⊙ ⊙ We assumeaNFWhaloprofile,hs Vi=3×10−27 cm3 s−1 andhBi=5 m G.WIMPmassis60GeVand thecompositionisbb. Solidlinesarewithoutdiffusion,anddottedarewithdiffusion. TheSKAsensitivity limits(dash-dottedlines,Dewdneyetal. 2012)areshownforintegrationtimesof30,240and1000hours, respectively(fromColafrancescoetal. 2014). 2.2 OptimalDMlaboratories In order to identify the optimal DM laboratories for radio observations we scan a parame- ter space extending from dwarf galaxies to galaxy clusters over a wide redshifts range z≈0−5. The choice to examine the halos of both large and small structures is crucial, as dwarf spheroidal galaxiesarewellknowntobehighlyDMdominatedbutproducefaintemissions,whilelargerstruc- tures, but not immaculate test-beds for DM emissions, provide substantially stronger fluxes. This indicates that a survey of DM halos with different mass is essential to identify the best detection prospects forfutureradiotelescopes liketheSKA. Figure4showstheredshift-mass exclusion plotobtained byusing theSKAsensitivity bound forSKA1LOWandSKA1MID(at1GHzinBand1). ForeachDMhalo weobtain theDMhalo mass and redshift combination that produce the minimal SKA-detectable fluxes. DM-dominated objectslyingabovetheblackandgreencurvescannotbedetectedwiththeSKA1atthegivenconfi- dencethresholdforhs Vi=3×10−27 cm3s−1. ObjectsbelowacurvearevisibletoSKA1,andthe further below thecurvetheyliethegreater theregion ofthecross-section parameter space wecan explore through the observation of the object. For reference, the yellow dash-dotted line displays thecurvegivenforhs Vi=3×10−30 cm3s−1and1s confidencelevel. Afewrepresentativeknow objects(irrespectiveoftheirlocationinthesky)withgoodestimatesoftheDMmassareplottedin theM −zplaneforthesakeofillustration oftheDMsearchpotential withtheSKA. DM Dwarfgalaxies, giventheirextremeproximity,provideanexcellenttest-bedforDMradioprobes, granting access to a parameter space that extends even below the value hs Vi = 3×10−30 cm3 s−1. Additionally their large mass-to-light ratios and absence of strong star formation and diffuse non-thermal emissionmakethemverycleansources forradioDMsearches. Galaxies can be probed to significantly larger redshifts than the dwarf galaxies due to their larger DMmass,andthoselocatedintheredshiftrange0.5∼<z∼<1.0providestrongerconstraints. How- ever, an optimized DM search should be confined to galaxies with little background radio noise, 5 UnveilingDarkMatterwithSKA SergioColafrancesco 101 100 Bullet 10−1 A2163 A2255 Coma z10−2 NGC3877 M87 NGC3198 10−3 M81 NGC7793 10−4 dSph 10−5 107 108 109 1010 1011 1012 1013 1014 1015 1016 M(M⊙) Figure4: ExclusionplotinredshiftversushalomassbasedonprojectedSKA(at1GHzinBand1)sensi- tivitydataforthereferencevalueofhs Vi=3×10−27cm3 s−1 with30hourintegrationtime(blacklines) and1000hourintegrationtime(greenlines).hBi=5m Gwasadopted.Solidlinesarethe1s sensitivityex- clusion,dashedlinesthatof2s anddottedlinescorrespondto3s .Theyellowdash-dottedlinecorresponds to30hoursofintegrationand1s confidencewithhs Vi=3×10−30 cm3 s−1. Anannihilationchannelbb isassumedwithaneutralinomassof60GeV. RepresentativeobjectswithknownDMmassareshownfor illustrativepurposesofDMradiosignaldetection. ThedSphgroupcontainsthegalaxies: Draco,Sculptor, Fornax,CarinaandSextans. UnlabelledGalaxiesare: NGC3917,NGC3949andNGC4010. Forverylocal objectstheredshiftisestimatedfromtheaveragedistancedata. FromColafrancescoetal. 2014. making low star-formation-rate galaxies good candidates. High-z galaxies come also with the ad- vantage ofobserving moreprimitivestructures withfewersources ofbaryonic radioemission. Clustersofgalaxiesprovideextremelygoodcandidates incases,suchastheBulletcluster, where thedarkandbaryonic matterarespatially separated. Ourrecent analysis oftheATCAobservation of the Bullet cluster (Colafrancesco and Marchegiani 2014) indicates that deeper radio observa- tions (possible with the SKA) will be able indeed to separate the DM-induced signal from the CR-induced one and hence have the possibly to investigate the nature of DM particles using the technique hereproposed. Moreingeneral, thelarge predicted radiofluxesduetoDMannihilation inclusters indicate thatDM-induced radioemissioncanbeobserved inradioouttolargeredshifts z≈2,againwiththeadvantage offewersourcesofbaryonic radioemission. 2.3 DisentanglingmagneticfieldsandDarkMatter Studyingthemagneticproperties ofDMhalosarecrucialtodisentangle theDMparticle den- sityfromthemagneticfieldenergydensitycontributingtotheexpectedsynchrotronradioemission fromDMannihilation. TheSKAisthemostpromisingexperimenttodeterminethemagneticfield structure in extragalactic sources (see Johnston-Hollitt et al. 2015), and will have the potential of measuring RMs toward a large number of sources allowing a detailed description of the strength, structure, and spatial distribution of magnetic fields in dSph galaxies, galaxies (see Beck et al. 2015)andgalaxyclusters(seeGovonietal. 2015). Westressthatthesemeasurementsofthemag- netic field can be obtained by the SKA simultaneously, for the first time, with the constraints on DMnaturefromtheexpectedradioemission. 6 UnveilingDarkMatterwithSKA SergioColafrancesco 3. The impactofSKA onthe search forthe DMnature Deep observations of radio emission in DM halos are not yet available, and this limits the capabilities of the current radio experiments to set relevant constraints on DM models. We have already explored a project (Regis et al. 2014a,b,c) dedicated to the WIMP search making use of radio interferometers, that could be considered as apilot experiment for the next generation high- sensitivity andhigh-resolution radiotelescope arraysliketheSKA.Fortheparticle DMsearchwe are interested in, the use of multiple array detectors having synthesized beams of ∼ arcmin size hasanumberofadvantageswithrespecttosingle-dishobservations. First,thelargecollectingarea allows for an increase in the sensitivity over that a single-dish telescope. The best beam choice forthedetection ofadiffuseemissionrequiresalargesynthesized beam(inordertomaximizethe integratedflux),butstillsmallerthanthesourceitselftobeabletoresolveit. Agoodangularreso- lutionisalsocrucialinordertodistinguish betweenapossiblenon-thermal astrophysical emission andtheDM-induced signal, whichclearly becomes veryhardiftheDMhaloisnotwell-resolved. Thepossibilityofsimultaneouslydetectingsmallscalesourceswiththelong-baselines ofthearray allows one to overcome the confusion limit. In the case of arcmin beams, the confusion level can be easily reached withobservations lasting for fewtens of minutes, even bycurrent telescopes. A source subtraction is thus a mandatory and crucial step of the analysis. Finally, single dish tele- scopes face the additional complication related to Galactic foreground contaminations, which are instead subdominant fortheangularscalestypically probedbytelescope arraysatGHzfrequency. The limits derived from ATCA observations of 6 dSphs (Regis et al. 2014c) on the WIMP annihilation/decay rate as a function of the mass for different final states of annihilation/decay are already comparable to the best limits obtained with g -ray observations and are much more constraining than what obtained in the X-ray band or with previous radio observations (Spekkens etal. 2013,Natarayanetal. 2013). Inthiscontext,theSKAwillhavethepossibilitytoexploreDM modelswithcross-section valueswellbelowtheDMrelicabundance one(seeFig.6inRegisetal. 2014c). TheSKA1-MIDBand1(350-1050MHz)willprobablybethemostpromisingfrequency range for the majority of WIMP models. The full SKA-2 phase will bring another factor ∼10× increaseinsensitivityandanextendedfrequencyrangeuptoatleast25GHz. Typicalvaluesofthe SKAsensitivity (A /T =2×104 m2/K)and bandwidth (300 MHzatGHzfrequency) provide eff sys rms flux values of ≈30 nJy for 10 hours of integration time. This is about a 103 factor of gain in sensitivity with respect to the most recent ATCA observations (Regis et al. 2014a). A further improvement byafactor of2-3canbeconfidently foreseen duetothelarger numberofaccessible dSphsatellites fromthesouthern hemisphere. TheSKAwillalsohavetheunique advantage tobe able to determine the dSph magnetic field (via FR measurements and possibly also polarization), provideditsstrengthisaroundthem Glevel(asexpectedfromstarformationratearguments,Regis et al. 2014c). This will make the predictions for the expected DM signal much more robust and obtainable with a single experimental configuration. The prospects of detection/constraints of the WIMP particle properties with the SKA will therefore progressively close in on the full parameterspace,eveninapessimisticsensitivitycase,andupto∼TeVWIMPmasses,irrespective ofastrophysical assumptions. TheSKAwillalso allowtoinvestigate thepossibility thatpoint-sources detected intheprox- imity of the dSph optical center might be associated to the emission from a DM cuspy profile. 7 UnveilingDarkMatterwithSKA SergioColafrancesco This possibility is likely only in the "loss at injection" scenario, while spatial diffusion should in any case flatten the e± distribution, making the source extended rather than point-like. The in- vestigation ofthese sources with the SKAwilldeserve particular attention, since wehave already foundthattheWIMPscenariocanfitthepoint-likeemissionwithannihilationratesconsistentwith existing bounds(Regisetal. 2014c). 105 104 104 103 103 b¯b 102 b¯b 107M⊙ 31−ms) 110012 ττ¯ 107M⊙ 31−ms) 110001 ττ¯ c c 27− 100 27− 10−1 1010−1 1010−2 (3×10−2 b¯b (3×10−3 b¯b 1012M⊙ σVi10−3 ττ¯ 1012M⊙ σVi10−4 ττ¯ h10−4 b¯b h10−5 b¯b 1015M⊙ 10−5 ττ¯ 1015M⊙ 10−6 ττ¯ 10−6 10−7 101 102 103 101 102 103 Mχ(GeV) Mχ(GeV) Figure5: Thehs Viupperlimitsfrom30hourofSKAintegrationtimeforz=0.01at300MHz(top)and 1 GHz (bottom)as the neutralinomass Mc is varied with annihilationchannelbb in solid lines and t t¯ in dashedlines. A valuehBi=5 m G was adopted. Black linescorrespondto haloswith mass1015 M , red ⊙ linesto1012M andgreenlinesto107M (fromColafrancescoetal. 2014). ⊙ ⊙ The SKA1-MID Band 1 (350-1050 MHz) to Band 4 (2.8-5.18 GHz) are important to probe the DM-induced synchrotron spectral curvature at low-n (sensitive to DM composition) and at high-n (sensitive to DM particle mass), and the implementation of Band 5 (4.6-13.8 GHz) will bring further potential to assess the DM-induced radio spectrum. As we can see from Fig.3, the best frequency range to detect these radio emissions is around 1 GHz. So, the upper frequency regions of SKA1-MID Band 1 provides the strongest spectral candidate for probing the cross- sectionparameterspaceduetoanoptimalcombinationoftheSKAsensitivitywithinthisbandand therelativelystrongfluxesatthesefrequencies. Thisfrequencybandwillalsoallowforanoptimal description ofthemagneticfield(seeJohnston-Hollitt etal. 2014). TherearetwomaincaveatsintheforecastsforDMdetectionintheradiofrequencyband. The firststemsfromthefactthat,foranextended radioemission,the confusion issuebecomesstronger and stronger as one tries to probe fainter and fainter fluxes. Thus, the source subtraction proce- dure becomes crucial and this can affect the estimated sensitivities. The impact of this effect on the actual sensitivity ishardly predictable atthepresent time, especially fortheSKA,since itwill depend on the properties of the detected sources, the efficiency of deconvolution algorithms, and theaccuracy ofthetelescope beamshape. The second caveat is that by bringing down the observational threshold, one can possibly start to probetheverylowlevelsofpossiblenon-thermalemissionassociatedtothetinyrateofstarforma- tionindSph, oringalaxies andgalaxy clusters. TheDMcontribution shouldbethendisentangled fromsuchastrophysicalbackground. ThesuperiorangularresolutionoftheSKAwillallowforthe precise mappingofemissions, putatively eitherDMorbaryonically induced, andwillenable their correlation withthestellarorDMprofiles(obtained viaopticaland/orkinematic measurements). 8 UnveilingDarkMatterwithSKA SergioColafrancesco Early DM science can be done with a small sample of local dSphs, a small sample of nearby galaxies with good DM density profile reconstruction, and the Bullet cluster, observing with a somewhatlargerbeam(≈7′′−10′′). Theseobjectshavebeenalreadystudiedinradiowithsimilar objectives (e.g., DM limits) and therefore provide the best science cases to prove the capabilities of the SKA1 for the study of the nature of DM with radio observations. The implementation of the SKA1-MID Band 5 (4.6-13.8 GHz) will increase the ability to detect the expected high- frequency spectral curvature of DM-induced radio emissions, and the ability to place constraints on the DM cross-section and to differentiate between different annihilation channel spectra. The 10× increased sensitivity of SKA2-MID compared to SKA1-MID will allow us to increase the angular resolution by a factor ≈ 20 or to increase the sensitivity of SKA2-MID to DM-induced radio signals by afactor ≈10. Thepossibility toextend thefrequency coverage of theSKAinits Phase-2 realization up to ∼25 GHz, will allow to detect the expected high-n spectral cut-off of DM-induced radioemissions, andthensetaccurate constraints totheDMparticlemass. 4. Conclusion The SKA has the potential to unveil the elusive nature of Dark Matter. Its ability to resolve the intrinsic degeneracy (between magnetic field properties and particle distribution) of the syn- chrotron emission expected from secondary particles produced in DM annihilation (decay) will allow such a discovery tobe unbiased and limited only by the sensitivity to the DM particle mass and annihilation cross-section (decay rate). The unprecedented sensitivity of the SKA to the DM fundamental properties will bring this instrument in a leading position for unveiling the nature of thedarksectoroftheuniverse. Theinformation provided bytheSKAcanbecomplemented withanalogous studiesinotherspec- tralbands, whichwillbeabletoprovetheICSsignal ofDM-produced secondary electrons (span- ningfromm wavestohardX-raysandg -rays)andthedistinctivepresenceofthep 0→gg emission bumpintheg -rays(seeFig. 1). Thenextdecadewillofferexcellentmulti-frequency opportunities inthis respect withtheadvent ofMillimetron, thelargest space-borne single-dish mm. astronomy satellite operating in the 102−103 GHz range (optimal to prove the DM-induced SZ effect), the Astro-Hmissionoperating inthehardX-raysfrequency range(withthehighest expectedsensitiv- itytoprobethehigh-energy tailoftheDM-induced ICSemission), andtheCTAwithunprecedent sensitivity intheenergyrangebetweenafewtensGeVtohundreds TeV. References [1] Ackermann,M.,etal.,2014,Phys.Rev.D,89,042001 [2] Beck,R.,Bomans,D.,Colafrancesco,S.,etal.2015,inAdvancingAstrophysicswiththeSquare KilometreArray,PoS(AASKA14)94 [3] Borriello,E.etal.2010,ApJ,709,L32 [4] Burns,J.O.etal..,1995,ApJ,446,583 [5] Colafrancesco,S.,2004,A&A,422,L23 [6] Colafrancesco,S.2010,invitedlectureatthe4thGamowInternationalConferenceonAstrophysics& CosmologyAfterGamow(9thGamowSummerSchool),AIPC,1206,5C 9 UnveilingDarkMatterwithSKA SergioColafrancesco [7] Colafrancesco,S.&Mele,B.2001,ApJ,562,24 [8] Colafrancesco,S.,Profumo,S.andUllio,P.2006,A&A,455,21 [9] Colafrancesco,S.,Profumo,S.andUllio,P.2007,PhRvD,75,3513 [10] Colafrancesco,S.etal.2011,A&A,527,80 [11] Colafrancesco,S.,andMarchegiani,P,2014,A&Ainpress [12] Colafrancesco,S.,Marchegiani,P.andBeck,G.,2014,[2014arXiv1409.4691C] [13] Daylan,T.etal.,2014,[arXiv:1402.6703] [14] Deiss,B.M.,Reich,W.,Lesch,H.&Wielebinski,R.1997,A&A,321,55 [15] Dewdney,P.,Turner,W.,Millenaar,R.,McCool,R.,Lazio,J.&Cornwell,T.,2012,SKAbaseline designdocument,http://www.skatelescope.org/wp-content/uploads/2012/07/SKA-TEL-SKO-DD- 001-1_BaselineDesign1.pdf [16] Doro,M.,2014,NuclearInstrumentsandMethodsinPhysicsResearchA,742,99 [17] Doro,M.,etal.,2013,AstroparticlePhysics,43,189 [18] Govoni,F.,Murgia,M.,Feretti,L.,etal.,2006,A&A,460,425 [19] Govoni,F.etal.2015,inAdvancingAstrophysicswiththeSquareKilometreArray, PoS(AASKA14)105 [20] Johnston-Hollitt,M.etal.2015,inAdvancingAstrophysicswiththeSquareKilometreArray, PoS(AASKA14)092 [21] Jungman,G.,Kamionkowki,M.&Griest,K.1996,Phys.rep,267,195 [22] Moran,E.C.etal.2014,[arXiv:1408.4451] [23] Natarajan,A.etal.,2013,Phys.Rev.D88083535[arXiv:1308.4979[astro-ph.CO]]. [24] PlanckCollaboration,2011,A&A,536,A7 [25] Regis,M.,Richter,L.,Colafrancesco,S.,Massardi,M.,deBlok,W.J.G.,Profumo,S.,Orford,N. 2014a,[arXiv:1407.5479] [26] Regis,M.,Richter,L.,Colafrancesco,S.,Profumo,S.,deBlok,W.J.G.,Massardi,M.2014b, [arXiv:1407.5482] [27] Regis,M.,Colafrancesco,S.,Profumo,S.,deBlok,W.J.G.,Massardi,M.,Richter,L.2014c,JCAP, 10,016R[arXiv:1407.4948] [28] Spekkens,K.,Mason,B.S.,Aguirre,J.E.andNhan,B.,2013,ApJ,773,61[arXiv:1301.5306 [astro-ph.CO]]. [29] Weniger,C.,2012,JCAP,1208,007 10