The Diffuse Galactic and Extragalactic Radio Emission Nic icol - PowerPoint PPT Presentation

The Diffuse Galactic and Extragalactic Radio Emission Nic icol olao F o For ornengo Department of Physics, University of Torino and Istituto Nazionale di Fisica Nucleare (INFN) – Torino Italy fornengo @ to.infn.it nicolao.fornengo@unito.it www.to.infn.it/~fornengo www.astroparticle.to.infn.it Radio Synchrotron Background Workshop University of Richmond, 20 July 2017

The isotropic radio background revisited NF, Lineros, Regis, Taoso JCAP 04 (2014) 008 Assessment of the size of the ARCADE excess Reanalysis to include detailed galactic foreground modeling and treatment of point-like and extended sources Galactic synchrotron emission from WIMPs at radio frequencies NF, Lineros, Regis, Taoso JCAP 01 (2012) 005 Bounds on particle DM from diffuse galactic radio emission

Tot Total l brightness of of ext xtra gala lact ctic c sky ky Collect all radio emission arriving from the sky Requires subtraction of galactic foregound Add up single contributions from all EG sources Individually resolved sources Statistical determination from fluctuations below detection threhold The two do not match Faint emitters are required New population(s) of astrophysicsl sources Dark matter

Fixsen et al, ApJ 734 (2011) 5 22 MHz – 10 GHz T E = 1 . 15 ( ν / GHz) − 2 . 6 K Vernstrom et al, MNRAS 440 (2014) 2791

Ra Radio Surveys Requirements Good coverage of high latitudes necessary to determine the EG emission Large fraction of the sky observed useful to anchor galactic foreground models Frequency Angular rms Noise Calibration Zero-level Fraction Survey [MHz] resolution [K] error [K] of Sky reference 1 . 1 � × 1 . 7 � / cos Z 22 3000 5% 5000 73% Roger et al. [24] 45 5 � 2300/300 10% 544 96% Guzman et al. [25] 408 0 . 85 � 1.2 10% 3 100% Haslam et al. [26] 820 1 . 2 � 0.5 6% 0.6 51% Berkhuijsen [27] 1420 0 . 6 � 0.017 5% 0.2 (0.5) 100% Reich et al. [29–31] 2326 0 . 33 � 0.03 5% 0.08 67% Jonas et al. [33]

Ma Maps ps 22 MHz 45 MHz 408 MHz 820 MHz 1420 MHz 2326 MHz

Mod Models ls T ( l, b ) = T E + T S ( l, b ) + T G ( l, b ) Isotropic EG Galactic diffuse Discrete sources emission (constant) emission (mask or templated) (model)

Ga Galactic diffuse emission Free-free Finkbeiner et al, ApJS 146 (2003) 407 Traced through H α line template with free norm Not that crucial, since we mask the galactic plane Synchrotron Primary electrons Secondary electrons and positrons Most relevant energy range (1 - 30) GeV

Sy Sync nchro hrotro ron s n sourc rce t term rm Primary electrons [a,b] Radial profile from SNR distribution Vertical profile: with z s = 0.2 Kpc exp( − z/z s ) Secondary electrons and positrons Interactions of primary p and He on ISM Sources injection spectra: broken power laws Spectral indeces and β inj , nuc β inj , e Breaks at 9 GeV/4 GeV for nuclei/electrons [a] Strong et al, ApJ 772 (2010) L58 [b] Lorimer et al, MNRAS 372 (2006) 777

Pr Propa opagation on setup GALPROP v. 54.1.984 Cylindrical box: Radial size: 20 Kpc Vertical half-height: L = 1 ÷ 40 Kpc Pure diffusion (no reacceleration) Reacceleration: increases secondary e ± at low energies as compared to pure diffusion: some tension with low frequency radio data Energy losses

Ma Magnetic c field lds ApJ 761 (2012) L11 Reference model: Jansson & Farrar ApJ 757 (2012) 14 Large-scale regular field disk component toroidal halo out-of-plane component Small-scale random field Striated random field Constrained on extragalactic Faraday rotation measures and on 22-GHz WMAP7 polarized and total intensity

Ma Magnetic c field lds To allow flexibility in the mid-high latitude emission (relevant for the determination of the extragalactic background), we let the random component to be more general: B ( R, z ) = B 0 exp[ − ( R − R T ) /R B ] exp( − | z | /z B ) R T = 8.5 kpc R B = 30 Kpc B 0 : determined by the fit Model a: z B = L Model b: z B = 2 kpc < L (only for L = 4, 8, 16 kpc) The z-scaling represents the main source of uncertainty related to the magnetic field modeling

Be Bench chmark k pr propa opagation on mod odels ls code name L D 0 β inj , nuc β inj ,e B 0 color coding [10 28 cm 2 s − 1 ] [kpc] [ µ G] L1 1 0.75 1.80/2.3 1.20/2.3 12 red L2 2 1.7 1.80/2.3 1.20/2.35 8.0 blue L4 4 3.4 1.80/2.3 1.20/2.35 6.0/7.0 green L8 8 5.8 1.80/2.3 1.20/2.35 4.6/4.7 orange L16 16 8.0 1.80/2.3 0.5/2.35 4.0/4.7 cyan L25 25 8.1 1.80/2.3 0.5/2.35 3.9 maroon L40 40 8.3 1.80/2.3 0.5/2.35 3.8 brown [1] [2] [3] [4] [1] D ( ρ ) = D 0 ( ρ / 4 GV) 0 . 5 [2] Index below/above break at 9 GeV [3] Index below/above break at 4 GeV [4] Model a or Model a/Model b T uned on CR data

Co Comparis ison n with ith CR CR da data ta Boron/Carbon

Co Comparis ison n with ith CR CR da data ta Antiprotons

Co Comparis ison n with ith CR CR da data ta Electrons

Fi Fitting pro rocedure re of rad radio map maps CMB monopole is subtracted: T = (2.72548 ± 0.00057 ) K Radio maps averaged over 15 deg scale (N side = 4) The GMF components have 2 different scales Regular: O(kpc) Random: O(100 pc) Stocasticity due to the random component introduces variance on the scale of its coherence length Better to compare emission averaged on this scale Best angular scale not obvious, due to LOS effect 15 deg as a conservative assumption

Fitting pro Fi rocedure re of rad radio map maps ( T data − T model ) 2 χ 2 = X i i σ 2 i i =pixels = T E + c gal T gal , synch + c brem T gal , brem T model i i i i ) 2 + ( σ exp σ 2 i = ( σ B ) 2 i σ i B : Variance induced by turbulence (data variance in pixel i) σ i exp : Experimental uncertainty

Ex Exten ended ded so sources es Galactic disk mask: |b| < 10 deg Intercepts galactic points sources and low lat sources Extended local sources (like radio loops) High-lat sources Masks Modeling

Masks Ma ks – Ta Take ke 1 Iterative method: 1. Fit of the map (out of the |b|<10 deg mask) with model = T E + c gal T gal , synch + c brem T gal , brem T model i i i 2. Compute residuals R i − T data − T BFmodel i i 3. Compute mean T R,i and σ R,i in 50 deg regions around the pixel i 4.Mask defined as R i > T R,i + 5 σ R,i 5. Repeat, with masked pixel excluded Iteration stops when masks stabilises The model fit is then performed on N side = 4 downgraded maps

Iter Iterativ tive e masks

Ma Masks ks – Ta Take ke 2 and 3 In order to cross-check and/or improve the impact of masks, we perform two additional trials: WMAP mask at 22 GHz SExtractor to determine masks at different frequencies Analyze original maps on 50 deg scale: mean, std deviation, detection threshold at 5 σ Similar to the iterative method: difference stays in flat local backgroud, while with iterative method galactic foregroud variations are taken into account Slightely larger masks

Wmap mask SExtractor masks

Te Templa plates Polarization template to intercept the most intense synchro sources T > 5 σ σ = 45 mK Template: DRAO + Villa Elisa (noise: 15mK zero level accuracy: 30mK( Polarization map at 1420 MHz T emplate = T E + c gal T gal , synch + c brem T gal , brem + c pol T gal , pol T model i i i i

Re Results T emperature Norm coefficients Chi squared 3 1.8 6 NUMBER COUNTS L1a 22 MHz L1a L2a 45 MHz 1.6 2.5 5 L2a L4a 408 MHz T E ν 2.5 [10 7 K MHz 2.5 ] L8a 820 MHz L4a L8a L16a 1.4 1420 MHz 2 4 L16a L25a 2326 MHz L25a L40a 1.2 χ 2 /ndf L40a c gal 1.5 3 1 1 2 0.8 1 0.5 0.6 0.4 0 0 10 1 10 2 10 3 10 4 10 1 10 2 10 3 1 10 ν [MHz] ν [MHz] L [Kpc] 1.8 6 4.5 L4b NUMBER COUNTS L4b L8b L4b 4 L8b 1.6 L16b 5 L8b L16b T E ν 2.5 [10 7 K MHz 2.5 ] L4F 3.5 1.4 L16b L4F L4L L4F L4L 4 3 1.2 L4L 2.5 χ 2 /ndf 1 c gal 3 2 0.8 2 1.5 0.6 1 0.4 1 0.5 0.2 0 0 0 10 1 10 2 10 3 10 4 10 1 10 2 10 3 10 1 10 2 10 3 10 4 ν [MHz] ν [MHz] ν [MHz]

Re Results 6 NUMBER COUNTS L1a 5 L2a T E ν 2.5 [10 7 K MHz 2.5 ] L4a L8a 4 L16a L25a L40a 3 2 1 From mumber counts 0 10 1 10 2 10 3 10 4 ν [MHz]

Results Re 3 L1a L2a 2.5 L4a L8a L16a 2 L25a L40a c gal 1.5 1 0.5 0 10 1 10 2 10 3 ν [MHz]

Results Re T emperature Norm coefficients 3 6 NUMBER COUNTS L1a L1a L2a 2.5 5 L2a L4a T E ν 2.5 [10 7 K MHz 2.5 ] L8a L4a L8a L16a 2 4 L16a L25a L25a L40a L40a c gal 1.5 3 1 2 1 0.5 0 0 10 1 10 2 10 3 10 4 10 1 10 2 10 3 ν [MHz] ν [MHz] Models with large L e ± softer: larger radio at low ν No large impact on T 820 MHz Calibration issues? Limited fraction of the sky? (smaller map: gal and EG more degenerate)

Re Results T emperature 6 Some dependence on the galactic model NUMBER COUNTS L1a 5 L2a T E ν 2.5 [10 7 K MHz 2.5 ] Not large (within a factor of 2) L4a L8a 4 L16a Smaller scatter for large-coverage maps L25a L40a 3 L1a 2 L2a L4a 1 ∆χ 2 (T E = number counts) L4b L8a L8b 0 100 L16a 10 1 10 2 10 3 10 4 L16b ν [MHz] L25a L40a L4F L4L 6 NUMBER COUNTS L4b 10 5 L8b T E ν 2.5 [10 7 K MHz 2.5 ] L16b 5 σ L4F 4 L4L 3 1 2 10 1 10 2 10 3 10 4 1 ν [MHz] Increase in chi2 assuming T E 0 10 1 10 2 10 3 10 4 from number counts ν [MHz]