Indirect Detection w/ X-rays & -rays ROSAT: 1990-1999 Fermi: - - PowerPoint PPT Presentation

GGI, 12 September 2019

Nick Rodd

ROSAT: 1990-1999 Fermi: 2008-present

& γ-rays X-rays Indirect Detection w/

2

Nick Rodd - Indirect Detection with X-rays and γ-rays

Overview

• 1. Landscape of X-ray & γ-ray indirection detection

3

Nick Rodd - Indirect Detection with X-rays and γ-rays

Overview

• 1. Landscape of X-ray & γ-ray indirection detection
• 2. Status of two anomalies: 3.5 keV line & GeV excess

H.E. γ-rays X-rays

3.5 keV Line GeV Excess

4

Nick Rodd - Indirect Detection with X-rays and γ-rays

Overview

• 1. Landscape of X-ray & γ-ray indirection detection
• 2. Status of two anomalies: 3.5 keV line & GeV excess

H.E. γ-rays X-rays

3.5 keV Line GeV Excess

• ne

Discussed last week

• 1. Landscape

6

Nick Rodd - Indirect Detection with X-rays and γ-rays

How do we detect Dark Matter?

How do we detect Dark Matter?

7

Φ(l, b) | {z }

γ/cm2/s

= hσvi 8πm2

χ

Z Emax

Emin

dNγ dE dE | {z }

“Particle Physics Factor”

⇥ Z

los

ρ2

DM(r) ds

| {z }

“J−Factor”

Nick Rodd - Indirect Detection with X-rays and γ-rays

What are the dark matter interactions? Where are they occurring?

8

Φ(l, b) | {z }

γ/cm2/s

= hσvi 8πm2

χ

Z Emax

Emin

dNγ dE dE | {z }

“Particle Physics Factor”

⇥ Z

los

ρ2

DM(r) ds

| {z }

“J−Factor”

Nick Rodd - Indirect Detection with X-rays and γ-rays

hσvi = 10−26 cm3/s mχ = 100 GeV dNγ/dE = 2δ(E mχ) (χχ ! γγ) ) PP ⇡ 10−31 cm3/s/GeV2

How do we detect Dark Matter?

9

Φ(l, b) | {z }

γ/cm2/s

= hσvi 8πm2

χ

Z Emax

Emin

dNγ dE dE | {z }

“Particle Physics Factor”

⇥ Z

los

ρ2

DM(r) ds

| {z }

“J−Factor”

Nick Rodd - Indirect Detection with X-rays and γ-rays

hσvi = 10−26 cm3/s mχ = 100 GeV dNγ/dE = 2δ(E mχ) (χχ ! γγ) ) PP ⇡ 10−31 cm3/s/GeV2

e.g. Segue 1 : J ≈ 1020 GeV2/cm5

⇒ Φ ≈ 10−11 γ/cm2/s

How do we detect Dark Matter?

10

Nick Rodd - Indirect Detection with X-rays and γ-rays

Φ(l, b) | {z }

γ/cm2/s

= hσvi 8πm2

χ

Z Emax

Emin

dNγ dE dE | {z }

“Particle Physics Factor”

⇥ Z

los

ρ2

DM(r) ds

| {z }

“J−Factor”

If we had a 1m2 space based telescope operate for 10 years:

• 10−11 γ/cm2/s
• ×
• 104 cm2

×

• 10 × π × 107 s
• ≈ 30 γ

hσvi = 10−26 cm3/s mχ = 100 GeV dNγ/dE = 2δ(E mχ) (χχ ! γγ) ) PP ⇡ 10−31 cm3/s/GeV2

e.g. Segue 1 : J ≈ 1020 GeV2/cm5

⇒ Φ ≈ 10−11 γ/cm2/s

How do we detect Dark Matter?

11

Nick Rodd - Indirect Detection with X-rays and γ-rays

Φ(l, b) | {z }

γ/cm2/s

= hσvi 8πm2

χ

Z Emax

Emin

dNγ dE dE | {z }

“Particle Physics Factor”

⇥ Z

los

ρ2

DM(r) ds

| {z }

“J−Factor”

If we had a 1m2 space based telescope operate for 10 years:

• 10−11 γ/cm2/s
• ×
• 104 cm2

×

• 10 × π × 107 s
• ≈ 30 γ

hσvi = 10−26 cm3/s mχ = 100 GeV dNγ/dE = 2δ(E mχ) (χχ ! γγ) ) PP ⇡ 10−31 cm3/s/GeV2

e.g. Segue 1 : J ≈ 1020 GeV2/cm5

⇒ Φ ≈ 10−11 γ/cm2/s

1m2 10 years

How do we detect Dark Matter?

−2 2 log10 [E/GeV] 8 10 12 log10 [E/cm2 × T/s]

12

Nick Rodd - Indirect Detection with X-rays and γ-rays

H.E. γ-rays

Landscape

Fermi

Cape Canaveral June 11, 2008

13

−2 2 4 6 log10 [E/GeV] 8 10 12 14 16 18 log10 [E/cm2 × T/s] Fermi

Nick Rodd - Indirect Detection with X-rays and γ-rays

V.H.E. γ U.H.E. γ H.E. γ-rays

Landscape

14

−2 2 4 6 log10 [E/GeV] 8 10 12 14 16 18 log10 [E/cm2 × T/s] Fermi

Nick Rodd - Indirect Detection with X-rays and γ-rays

V.H.E. γ U.H.E. γ H.E. γ-rays

Landscape

HAWC H.E.S.S.

2015-present 2002-present (H.E.S.S. II 2012-)

15

−2 2 4 6 log10 [E/GeV] 8 10 12 14 16 18 log10 [E/cm2 × T/s] Fermi

Nick Rodd - Indirect Detection with X-rays and γ-rays

V.H.E. γ U.H.E. γ H.E. γ-rays

Landscape

LHAASO

CTA

2021? ~2025

16

−2 2 4 6 log10 [E/GeV] 8 10 12 14 16 18 log10 [E/cm2 × T/s] Fermi

Nick Rodd - Indirect Detection with X-rays and γ-rays

V.H.E. γ U.H.E. γ H.E. γ-rays

Landscape

LHAASO

CTA

2021? ~2025

P r e l i m i n a r y

−4 −2 2 4 6 log10 [E/GeV] 8 10 12 14 16 18 log10 [E/cm2 × T/s] Fermi HAWC H . E . S . S . CTA LHAASO

17

Nick Rodd - Indirect Detection with X-rays and γ-rays

V.H.E. γ U.H.E. γ M.E. γ H.E. γ-rays

Landscape

18

−4 −2 2 4 6 log10 [E/GeV] 8 10 12 14 16 18 log10 [E/cm2 × T/s] Fermi HAWC H . E . S . S . CTA LHAASO

Nick Rodd - Indirect Detection with X-rays and γ-rays

V.H.E. γ U.H.E. γ M.E. γ H.E. γ-rays

Landscape

COMPTEL

1991-2000

CGRO EGRET

1991-2000

CGRO

−4 −2 2 4 6 log10 [E/GeV] 8 10 12 14 16 18 log10 [E/cm2 × T/s] Fermi HAWC H . E . S . S . CTA LHAASO

19

Nick Rodd - Indirect Detection with X-rays and γ-rays

V.H.E. γ U.H.E. γ H.E. γ-rays

Landscape

M.E. γ

e-Astrogram

2030s?

AMEGO

?

−4 −2 2 4 6 log10 [E/GeV] 8 10 12 14 16 18 log10 [E/cm2 × T/s] Fermi HAWC H . E . S . S . CTA LHAASO

20

Nick Rodd - Indirect Detection with X-rays and γ-rays

V.H.E. γ U.H.E. γ H.E. γ-rays

Landscape

M.E. γ

e-Astrogram

2030s?

AMEGO

?

−6 −4 −2 2 4 6 log10 [E/GeV] 8 10 12 14 16 18 log10 [E/cm2 × T/s] Fermi HAWC H.E.S.S. CTA LHAASO COMPTEL E G R E T AMEGO e-Astrogram

21

Nick Rodd - Indirect Detection with X-rays and γ-rays

H.E. γ-rays V.H.E. γ U.H.E. γ X-rays

Landscape

M.E. γ

−6 −4 −2 2 4 6 log10 [E/GeV] 8 10 12 14 16 18 log10 [E/cm2 × T/s] Fermi HAWC H.E.S.S. CTA LHAASO COMPTEL E G R E T AMEGO e-Astrogram

H.E. γ-rays V.H.E. γ U.H.E. γ

Chandra

22

Nick Rodd - Indirect Detection with X-rays and γ-rays

X-rays

Landscape

M.E. γ

NuSTAR INTEGRAL

2012-present 2002-present

XMM-NEWTON

1999-present 1999-present

−6 −4 −2 2 4 6 log10 [E/GeV] 8 10 12 14 16 18 log10 [E/cm2 × T/s] Fermi HAWC H.E.S.S. CTA LHAASO COMPTEL E G R E T AMEGO e-Astrogram XMM-Newton C h a n d r a N u S T A R INTEGRAL

23

Nick Rodd - Indirect Detection with X-rays and γ-rays

H.E. γ-rays V.H.E. γ U.H.E. γ X-rays

Landscape

M.E. γ

ATHENA

~2031

−6 −4 −2 2 4 6 log10 [E/GeV] 8 10 12 14 16 18 log10 [E/cm2 × T/s] Fermi HAWC H.E.S.S. CTA LHAASO COMPTEL E G R E T AMEGO e-Astrogram XMM-Newton C h a n d r a N u S T A R INTEGRAL ATHENA

24

Nick Rodd - Indirect Detection with X-rays and γ-rays

H.E. γ-rays V.H.E. γ U.H.E. γ X-rays

Landscape

M.E. γ

−6 −4 −2 2 4 6 log10 [E/GeV] 8 10 12 14 16 18 log10 [E/cm2 × T/s] Fermi HAWC H.E.S.S. CTA LHAASO COMPTEL E G R E T AMEGO e-Astrogram XMM-Newton C h a n d r a N u S T A R INTEGRAL ATHENA

25

Nick Rodd - Indirect Detection with X-rays and γ-rays

H.E. γ-rays V.H.E. γ U.H.E. γ X-rays

Landscape

M.E. γ

• 2. Anomalies

3.5 keV Line GeV Excess

• 2. Anomalies

3.5 keV Line GeV Excess

?

28

6.7 6.8 6.9 7.0 7.1 7.2 7.3 7.4

ms [keV]

10−12 10−11 10−10 10−9

sin2(2θ) 1 2 3 4 5 6 7 8 9 10

Status of the 3.5 keV line

Figure reproduced from [Abazajian 1705.01837]

Nick Rodd - Indirect Detection with X-rays and γ-rays

Legend:

• 1. [Boyarsky+ 1402.4119]

XMM-Newton & Perseus

• 2. [Boyarsky+ 1402.4119]

XMM-Newton & M31

• 3. [Bulbul+ 1402.2301]

XMM-Newton PN & stacked galaxy clusters

• 4. [Bulbul+ 1402.2301]

XMM-Newton MOS & stacked galaxy clusters

• 5. [Cappelluti+ 1701.07932]

Chandra & stacked galaxy clusters

• 6. [Aharonian+ 1607.07420]

Hitomi & Perseus

• 7. [Tamura+ 1412.1869]

Suzaku & Perseus

• 8. [Malyshev+ 1408.3531]

XMM-Newton & stacked dwarfs

• 9. [Horiuchi+ 1311.0282]

Chandra & M31 10.[Anderson+ 1408.4115] Chandra+XMM-Newton & stacked galaxy clusters

29

6.7 6.8 6.9 7.0 7.1 7.2 7.3 7.4

ms [keV]

10−12 10−11 10−10 10−9

sin2(2θ) 1 2 3 4 5 6 7 8 9 10

95% limit (this work) mean expected 1σ/2σ containment

XMM- Newton

[Dessert, NLR, Safdi 1812.06976]

Status of the 3.5 keV line

Nick Rodd - Indirect Detection with X-rays and γ-rays

MW Halo

+ =

New Strategy

30

Nick Rodd - Evidence the 3.5 keV line is not from Dark Matter Decay

dΦ dE = 1 4π ms τ δ(E − ms/2) × R

LoS ds

R

FoV dΩ ρDM(s, Ω)

R

FoV dΩ

• Expected DM flux

• Expected DM flux
• Perseus flux
• Perseus halo > XMM Field of View, reduces flux by factor of ~3

New Strategy

31

Nick Rodd - Evidence the 3.5 keV line is not from Dark Matter Decay

dΦ dE = 1 4π ms τ δ(E − ms/2) × R

LoS ds

R

FoV dΩ ρDM(s, Ω)

R

FoV dΩ

DPers ≈ 1 ΩXMM MPers d2

≈ 1 (104 sr) (1015 M) (100 Mpc)2 ∼ 1029 keV/cm2

New Strategy

32

Nick Rodd - Evidence the 3.5 keV line is not from Dark Matter Decay

• Expected DM flux
• Perseus flux
• Perseus halo > XMM Field of View, reduces flux by factor of ~3
• What about for the Milky Way?
• Number comparable! Yet more MW we can see than Perseus clusters

DPers ≈ 1 ΩXMM MPers d2

≈ 1 (104 sr) (1015 M) (100 Mpc)2 ∼ 1029 keV/cm2

DMW ≈ Z ds ρDM(s, Ω) ≈ (0.4 GeV/cm3) × (20 kpc) ≈ 2 × 1028 keV/cm2

dΦ dE = 1 4π ms τ δ(E − ms/2) × R

LoS ds

R

FoV dΩ ρDM(s, Ω)

R

FoV dΩ

33

Status of the 3.5 keV line

Nick Rodd - Indirect Detection with X-rays and γ-rays

• Key observation: Milky Way halo is bright even away from GC
• Average emission over the XMM-Newton FoV

Perseus

ψPers = 148

DPers ∼ 3 × 1028 keV/cm2

DMW(ψ = 148) ∼ 1 × 1028 keV/cm2

DMW(ψ = 45) ∼ 3 × 1028 keV/cm2

• The line is present in every
• bservation XMM has ever made
• If the line was real, we would

have detected it at over 100σ!

• back. model
• back. + signal

[Dessert, NLR, Safdi 1812.06976]

Conclusion

34

• Exciting new experiments coming online in 10-15 years
• Some of the best datasets are already on disk, need to extract all

the information we can from them

• DM interpretation of 3.5 keV excess strongly disfavoured
• GeV excess not yet as clear cut!

Nick Rodd - Indirect Detection with X-rays and γ-rays

3.5 keV Line GeV Excess

✗ ?

Backup Slides

36

Nick Rodd - Indirect Detection with X-rays and γ-rays

Landscape

41.5 42.0

Proton Decay

28 30 32 log10[τ/s]

S = 30, χ → γγ MW, χ → γγ MW+EG, χ → b¯ b

−2 2 log10[mχ/GeV] 19.0 19.5

χ → relativistic

[Steigman, Dasgupta, Beacom 1204.3622] [Super-Kamiokande 1605.03235] [Gong, Chen 0802.2296]

2 2 log10[mχ/GeV] 34 32 30 28 26 24 log10[hσvi/(cm3/s)]

S = 3 , χ χ ! γ γ M W , χ χ ! γ γ C l u s t e r s , χ χ ! b ¯ b D w a r f , χ χ ! b ¯ b M W , χ χ ! b ¯ b Thermal Relic

(Aside) Fermi Limits

• What reach should we expect? In the large count limit
• Bulbul+ detected line with and
• Blank sky observations (BSO) much lower background than Perseus,

by selection:

• As the signal is at least as bright starting at 45o, we could reach the

same significance using only

• With the full ~30 Ms dataset expect
• This analysis could detect particle dark matter at over 100σ

Estimated Sensitivity

37

ΦBSO

B

/ΦPers

B

∼ 0.02

tBSO ≈ tPers × (ΦBSO

B

/ΦPers

B

) ≈ 6 ks

TSBSO ≈ 16 × (30 Ms/6 ks) ≈ 75, 000

tPers ∼ 320 ks

TS ∼ 16

TS = 2[ln LS − ln LB] ∼ σ2 ∼ S2/B = Φ2

S/ΦB × t

Nick Rodd - Indirect Detection with X-rays and γ-rays

Original Claim

38

Nick Rodd - Evidence the 3.5 keV line is not from Dark Matter Decay

• Is the line consistent with dark matter?
• ~scale with cluster mass (see [Lovell+ 1810.05168])
• No known significant lines nearby, but cluster emission is complex -

model 31 known emission lines

• A real line we missed?
• K XVIII lines at 3.48 and 3.52 keV [Jeltema+Profumo 1408.1699]
• S XVI charge exchange at 3.5 keV [Gu+ 1511.06557]

[1402.2301]

New Strategy

39

• Strategy motivates using all ~12,000 observations
• Developed automating tools:

github.com/nickrodd/XMM-DM

• Processed all 6,350 obs with
• Apply cuts to restrict this to the best datasets
• Lowest 68% of instrumental background
• Remove
• 1,397 exposures, 752 observations, 30.6 Ms

http://nxsa.esac.esa.int Data: Exposure:

tobs < 1 ks

ψ < 90

5 < ψ < 45

I2−10 < 5 × ICXRB

2−10

Nick Rodd - Indirect Detection with X-rays and γ-rays

New Strategy

40

−40 −20 20 40 ` [degrees] −40 −20 20 40 b [degrees] 0.5 1.0 1.5 2.0 2.5 log(exposure/ks)

• Exposures well distributed over the region

Nick Rodd - Indirect Detection with X-rays and γ-rays

Profile Likelihood Analysis

41

• Analyse each exposure using profile likelihood
• Likelihoods are then joined, we do not stack the datasets
• Use narrow energy window:
• Model: astrophysical power-law, instrumental power-law, DM

ms/2 ± 0.25 keV

• In detail use Poisson likelihood for counts + Gaussian likelihood for

QPB estimates. Instrument response folded into the model prediction

(∆EXMM ≈ 0.1 keV)

Nick Rodd - Indirect Detection with X-rays and γ-rays

42

Profile Likelihood Analysis

• Nuisance parameters removed using the profile likelihood
• The background is refit for every value of the signal

Nick Rodd - Indirect Detection with X-rays and γ-rays

43

• No evidence for a DM decay line
• Left inset shows the distribution of individual exposures versus a χ2

distribution under the null, provides a good fit to the data

Results

• Calculate the TS for the DM line from the joint profiled likelihood

6.7 6.8 6.9 7.0 7.1 7.2 7.3 7.4

1 2 3 4 5

−40 −20 20 40 ` [degrees] −40 −20 20 40 b [degrees] m = 7.11 keV

1 2 3 4 5 6 7 8 TS Nick Rodd - Indirect Detection with X-rays and γ-rays

44

Nick Rodd - Evidence the 3.5 keV line is not from Dark Matter Decay

Towards a Definitive Statement

• 2. Deep observation of dark matter bright object

[1512.01239]

45

Nick Rodd - Evidence the 3.5 keV line is not from Dark Matter Decay 6.7 6.8 6.9 7.0 7.1 7.2 7.3 7.4

ms [keV]

10−12 10−11 10−10 10−9

sin2(2θ) 1 2 3 4 5 6 7 8 9 10

Towards a Definitive Statement

[1512.01239]

• 2. Deep observation of dark matter bright object

46

Nick Rodd - Evidence the 3.5 keV line is not from Dark Matter Decay

Towards a Definitive Statement

• 2. Deep observation of dark matter bright object

[1512.07217]

47

Nick Rodd - Evidence the 3.5 keV line is not from Dark Matter Decay

SYSTEMATIC CROSS CHECKS

• The result is controversial, so we have cross checked extensively
• If there was a real signal in the data, would we have excluded it?
• Check by injecting a real signal into the data

10−12 10−11 10−10 sin2(2θinj) 10−12 10−11 10−10 sin2(2θrec) mχ = 7.0 keV

48

Nick Rodd - Evidence the 3.5 keV line is not from Dark Matter Decay

SYSTEMATIC CROSS CHECKS

• How dependent are these results on the assumed halo profile?

ms [keV]

sin2(2θ) 1 2 3 4 5 6 7 8 9 10

ρNFW(r) = ρ0 r/rs (1 + r/rs)2

ρlocal = 0.4 GeV/cm3

r = 8.127 kpc

rs = 20 kpc

rc = 9 kpc

ρBurk(r) = ρ0 (1 + r/rc)(1 + (r/rc)2)

49

Nick Rodd - Evidence the 3.5 keV line is not from Dark Matter Decay

SYSTEMATIC CROSS CHECKS

• Is the result strongly dependent upon our cuts?

ms [keV]

sin2(2θ) 1 2 3 4 5 6 7 8 9 10

ms [keV]

sin2(2θ) 1 2 3 4 5 6 7 8 9 10