The Galactic Center GeV Excess: Have We Started to See Dark Matter? - - PowerPoint PPT Presentation

▶
the galactic center gev excess have we started to see
SMART_READER_LITE
LIVE PREVIEW

The Galactic Center GeV Excess: Have We Started to See Dark Matter? - - PowerPoint PPT Presentation

Jun 26, 2023 145 likes •852 views

The Galactic Center GeV Excess: Have We Started to See Dark Matter? Sam McDermott Based on: various observational works (Daylan et al 1402.6703, Calore et al 1409.0042, ) SDM, I. Cholis, P. Fox, S. K. Lee (preliminary / in progress) GGI,

slide-1
SLIDE 1

The Galactic Center GeV Excess: Have We Started to See Dark Matter?

Sam McDermott

various observational works (Daylan et al 1402.6703, Calore et al 1409.0042, …)

Based on:

GGI, 9/30/15

SDM, I. Cholis, P. Fox, S. K. Lee (preliminary / in progress)

slide-2
SLIDE 2

2

slide-3
SLIDE 3

Outline

  • 1. Observational facts (“introduction”)
  • how many photons? from where?
  • what is it?
  • 2. A new observational idea

3

slide-4
SLIDE 4

Basics

  • Two kinds of analyses
  • Galactic center — below the Bubbles
  • Inner Galaxy — excludes disk, goes out > O(10˚)
  • “Excess” found both near and far from SgrA*
  • Appears to be spherical and smooth; radial fall-off

compatible with (gNFW profile)2: ρgNFW(x)~ρ0/[xγ(1+x)3-γ]

4

slide-5
SLIDE 5

Galactic Center

point sources; isotropic; diffuse emission; map of 20 cm synchrotron excess with normalization ~ 30% of raw! ∫los(gNFW profile)2 fits excess well

1402.6703

5

slide-6
SLIDE 6

Galactic Center

point sources; isotropic; diffuse emission; map of 20 cm synchrotron excess with normalization ~ 30% of raw! ∫los(gNFW profile)2 fits excess well

1402.6703

“π0’s” = hadronic CRs interacting with dust “bremsstrahlung” = leptonic CRs interacting with dust “ICS” = leptonic CRs interacting with background light

6

slide-7
SLIDE 7

Galactic Center

point sources; isotropic; diffuse emission; map of 20 cm synchrotron excess with normalization ~ 30% of raw! ∫los(gNFW profile)2 fits excess well

1402.6703

“π0’s” = hadronic CRs interacting with dust “bremsstrahlung” = leptonic CRs interacting with dust “ICS” = leptonic CRs interacting with background light cosmic rays interacting with previously mapped stuff

7

slide-8
SLIDE 8

Inner galaxy

diffuse map Fermi bubbles NFW

180 90

  • 90
  • 180
  • 90
  • 45

45 90

  • 5
  • 4
  • 3
  • 2
  • 1
  • 5
  • 4
  • 3
  • 2
  • 1

180 90

  • 90
  • 180
  • 90
  • 45

45 90

  • 5
  • 4
  • 3
  • 2
  • 1
  • 5
  • 4
  • 3
  • 2
  • 1
  • 90
  • 45

45 90

  • 5
  • 4
  • 3
  • 2
  • 1
  • 5
  • 4
  • 3
  • 2
  • 1

1402.6703

slide-9
SLIDE 9

Inner galaxy

0.5-1 GeV residual

  • 20
  • 10

10 20

  • 20
  • 10

10 20 5 10 15 20 5 10 15 20 10-6 counts/cm2/s/sr

1-3 GeV residual

  • 20
  • 10

10 20

  • 20
  • 10

10 20

  • 2

2 4 6 8 10 12

  • 2

2 4 6 8 10 12 10-6 counts/cm2/s/sr

3-10 GeV residual

  • 20
  • 10

10 20

  • 20
  • 10

10 20 1 2 3 4 1 2 3 4 10-6 counts/cm2/s/sr

10-50 GeV residual

  • 20
  • 10

10 20

  • 20
  • 10

10 20 0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5 10-6 counts/cm2/s/sr

1402.6703

9

slide-10
SLIDE 10

Total Normalization

at energies of interest, much brighter than Bubbles (~ O(30%) of total!)

−20 −15 −10 −5 5 10 15 20 b [deg] 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Count density [sr−1] ×106

2.12-3.32 GeV −2.0◦ < ℓ < 2.0◦ GCE PSCs π0+Bremss ICS Isotropic Bubbles Sum

1409.0042

10

slide-11
SLIDE 11

Seen out to > 10˚

1402.6703

γ=1.3

11

slide-12
SLIDE 12

Highly spherical…

1402.6703

12

slide-13
SLIDE 13

… robust to diffuse map

Presence of a signal with energy peak ~ 2 GeV is robust to changes in diffuse template

−20 −15 −10 −5 5 10 15 20 ℓ [deg] −20 −15 −10 −5 5 10 15 20 b [deg] I II III IV V VI VII VIII XI X

−0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 ×10−5

I

×10−5

II

−1 1 2 3 4 E2 dN dE [GeV/(cm2s sr)] ×10−6

III

×10−6

IV

−0.5 0.0 0.5 1.0 1.5 ×10−6

V

×10−6

VI

−1.0 −0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 E2 dN dE [GeV/(cm2s sr)] ×10−6

VII

×10−6

VIII

100 101 102 E [GeV] −1.0 −0.5 0.0 0.5 1.0 1.5 ×10−6

IX

100 101 102 E [GeV] ×10−6

X

1409.0042

13

slide-14
SLIDE 14

Qualitative thing we are not yet sure of:

The existence of an excess is pretty well agreed upon (independent methods by independent groups* agree something is there)

*also see work by: Abazajian and collaborators (1207.6047, 1402.4090, 1410.6168); Gordon, Macias, and collaborators (1306.5725, 1312.6671, 1410.1678, 1410.7840); Murgia’s Fermi symposium slides

Is the excess from astrophysics or dark matter?

14

slide-15
SLIDE 15

Qualitative thing we are not yet sure of:

If DM, we need to confront other issues:

  • are there other indirect detection signals? bounds?
  • what are its interactions with the SM?
  • what is the UV theory?

Is the excess from astrophysics or dark matter?

15

slide-16
SLIDE 16

“Secondaries”

basic statements: no positron “bump” found, understanding of anti-baryons is murky

1306.3983 1410.1527

101 102 mχ [GeV] 10−29 10−28 10−27 10−26 10−25 10−24 10−23 ⟨σv⟩ [cm3s−1] dashed: Fermi LAT WMAP7 solid: AMS-02 (this work)

τ +τ − µ+µ− e+e−γ e+e−

Bergstr¨

  • m et al. (2013)

16

slide-17
SLIDE 17

Cosmic Ray Constraints

20 40 60 80 100 10-27 10-26 10-25 10-24

MDM @GeVD Xsv\ @cm3 s-1D

THN CON KOL THK KRA Benchmark propagation models 20 40 60 80 100 10-27 10-26 10-25 10-24

MDM @GeVD Xsv\ @cm3 s-1D

THN CON KOL THK KRA Benchmark propagation models 20 40 60 80 100 10-27 10-26 10-25 10-24

MDM @GeVD Xsv\ @cm3 s-1D

THN CON KOL THK KRA Benchmark propagation models 20 40 60 80 100 10-27 10-26 10-25 10-24

MDM @GeVD Xsv\ @cm3 s-1D

THN THN2 THN3 Thin propagation models 20 40 60 80 100 10-27 10-26 10-25 10-24

MDM @GeVD Xsv\ @cm3 s-1D

THN THN2 THN3 Thin propagation models 20 40 60 80 100 10-27 10-26 10-25 10-24

MDM @GeVD Xsv\ @cm3 s-1D

THN THN2 THN3 Thin propagation models

different colors: different choices

  • f diffusion zone

parameters different rows: different choices

  • f relation b/w

φFp and φFp local and Galactic

1407.2173

17

slide-18
SLIDE 18

Dwarf Galaxies

Geringer-Sameth et al, 1503.02320

100 101 102 Energy [GeV] 10−7 10−6 10−5 E2 dF/dE [GeV cm−2 s−1 sr−1] 13390 51 33 22 18 11 10 6 1 1 101 102 103 Mass [GeV] −3σ −2σ −1σ 0σ 1σ 2σ 3σ Significance

τ +τ − Ret2 Seg1

18

slide-19
SLIDE 19

Dwarf Galaxies

Geringer-Sameth et al, 1503.02320

100 101 102 Energy [GeV] 10−7 10−6 10−5 E2 dF/dE [GeV cm−2 s−1 sr−1] 13390 51 33 22 18 11 10 6 1 1 101 102 103 Mass [GeV] −3σ −2σ −1σ 0σ 1σ 2σ 3σ Significance

τ +τ − Ret2 Seg1

19

101 102 103 Mass [GeV] −3σ −2σ −1σ 0σ 1σ 2σ 3σ Significance

τ +τ − Ret2 Seg1

slide-20
SLIDE 20

Dwarf Galaxies

Drlica-Wagner et al, 1503.02632

101 102 103 104

DM Mass (GeV/c2)

10−27 10−26 10−25 10−24 10−23 10−22 10−21

hσvi (cm3 s1)

b¯ b

DES J0222.7-5217 DES J0255.4-5406 DES J0335.6-5403 DES J0344.3-4331 DES J0443.8-5017 DES J2108.8-5109 DES J2339.9-5424 DES J2251.2-5836 Combined DES Candidate dSphs Combined Known dSphs

Thermal Relic Cross Section (Steigman et al. 2012)

20

slide-21
SLIDE 21

Dwarf Galaxies

101 102 103 104

DM Mass (GeV/c2)

10−27 10−26 10−25 10−24 10−23 10−22

hσvi (cm3 s1)

b¯ b

Pass 8 Combined dSphs Fermi-LAT MW Halo H.E.S.S. GC Halo MAGIC Segue 1 Abazajian et al. 2014 (1σ) Gordon & Macias 2013 (2σ) Daylan et al. 2014 (2σ) Calore et al. 2014 (2σ)

Thermal Relic Cross Section (Steigman et al. 2012)

  • B. Anderson et al,

1503.02632

21

slide-22
SLIDE 22

How Bright?

Essig, Massari, et al 1503.07169

1 101 102 103 104 10-27 10-26 10-25 10-24 10-23 10-22 10-21 mDM [GeV] 〈σ v〉 [cm3s-1] DM DM → b b

Isothermal NFW Einasto NFWc solid: data dashed: average MC limit shading: population st. dev.

  • f the 10 MC limits

thermal WIMP

22

slide-23
SLIDE 23

If SM, we need a consistent explanation:

  • existence of the Fermi bubbles is suggestive; but hard

to get smooth structure from this kind of burst

  • millisecond pulsars show up over the correct scales

range with plausibly correct morphology; but…

Qualitative thing we are not yet sure of:

Is the excess from astrophysics or dark matter?

23

slide-24
SLIDE 24

Point Sources

1412.6099

n~r-δ

δ~2.5 observed in Andromeda

(cf. ρ2~r-2γ)

24

slide-25
SLIDE 25

Point Source Fits

based on non-Poissonian template fit, point sources can account for excess

25

1506.05124

3 5 18 9 2 2

slide-26
SLIDE 26

Millisecond Pulsars

  • Spectra are “significantly”

different

  • Should have resolved many

more MSPs in inner 1.8 kpc (~few˚) given “reasonable” luminosity function: N(L>1034 erg/s) ~ 200, N(L>1035 erg/s) ~ 60 1407.5625

26

slide-27
SLIDE 27

Globular Clusters

  • globular clusters are dense

stellar environments (=> star- star encounters are common)

  • Some star-star encounters

create X-ray binaries, some create MSPs

  • X-ray binaries fizzle out sooner

than MSPs

27

slide-28
SLIDE 28

DISRUPTED GLOBULAR CLUSTERS CAN EXPLAIN THE GALACTIC CENTER GAMMA RAY EXCESS

Timothy D. Brandt1,3 and Bence Kocsis1,2

1507.05616

(with zero free parameters)

28

slide-29
SLIDE 29

DISRUPTED GLOBULAR CLUSTERS CAN EXPLAIN THE GALACTIC CENTER GAMMA RAY EXCESS

Timothy D. Brandt1,3 and Bence Kocsis1,2

1507.05616

(with zero free parameters)

29

some concerns, still

slide-30
SLIDE 30

Point Sources, II

still some missing point sources?

30

1506.05124

3 5 18 9 2 2

slide-31
SLIDE 31

existence of a signal is pretty robust, but…

  • … diffuse templates house large, energy-

dependent uncertainties

  • … serious caution and healthy skepticism are

required when interpreting as BSM physics

  • … a few opportunities so far that “could” have

been convincing (either way) have not panned out

Lessons

31

slide-32
SLIDE 32

How else can we convince ourselves this is or isn’t dark matter?

32

slide-33
SLIDE 33

Particle physics ideas How else can we convince ourselves this is or isn’t dark matter?

33

slide-34
SLIDE 34

Particle physics ideas New observational ideas How else can we convince ourselves this is or isn’t dark matter?

34

slide-35
SLIDE 35

Particle physics ideas New observational ideas How else can we convince ourselves this is or isn’t dark matter?

35

work in progress with Ilias Cholis, Paddy Fox, and Samuel K. Lee 1510.$#%@&!

slide-36
SLIDE 36

Current Technique

Test assumption of dark matter annihilation:

  • statistical discrimination (χ2 test) between

fits with and without signal template

  • fits with template do better

36

slide-37
SLIDE 37

Test assumption of dark matter annihilation:

  • statistical discrimination (χ2 test) between

fits with and without signal template

  • fits with template do better

…but what if there is a totally different shape on the sky that was not adequately tested?

Current Technique

37

slide-38
SLIDE 38

Test assumption of dark matter annihilation:

  • statistical discrimination (χ2 test) between

fits with and without dark matter template

  • fits with template do better

…but what if there is a totally different shape on the sky that was not adequately tested?

It would be nice to find evidence without making this assumption!

Current Technique

38

slide-39
SLIDE 39

Wavelets

Allow analysis sensitive to both location and scale Used for a wide variety of industrial and academic applications:

  • image compression (JPEG-2000)
  • fast astrophysical signal identification
  • cochlear transforms (mimic hearing)
  • image denoising
  • jets (this is still in its infancy…)
  • etc.**

39

slide-40
SLIDE 40

What are wavelets?

W(a, b) = 1 √a Z f(x)ψ∗ ✓x − b a ◆ dx

scale position wavelet coefficients

  • riginal signal

mother wavelet (different choices)

Z ψ(x)dx = 0 Z |ψ(x)|2dx = 1 ψ(x) ∈ L2(R) and 1 √aψ ✓x − b a ◆ ∈ L2(R) for a, b ∈ Z

40

slide-41
SLIDE 41

What are wavelets?

W(a, b) = 1 √a Z f(x)ψ∗ ✓x − b a ◆ dx

scale position

  • riginal signal

mother wavelet (different choices)

Z ψ(x)dx = 0 Z |ψ(x)|2dx = 1 ψ(x) ∈ L2(R) and 1 √aψ ✓x − b a ◆ ∈ L2(R) for a, b ∈ Z

How (and why) do they work?

wavelet coefficients

41

slide-42
SLIDE 42

sine wave

  • 4
  • 2

2 4

  • 1.0
  • 0.5

0.5 1.0

0.0 0.5 1.0 1.5 2.0 1 2 3 4 Freq. power

F

  • u

r i e r w a v e l e t

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.2 0.4 0.6 0.8 1.0 scale

M e x . h a t

42

slide-43
SLIDE 43

two sine waves

F

  • u

r i e r w a v e l e t

0.0 0.5 1.0 1.5 2.0 2.5 1 2 3 4 5 Freq. power

  • 4
  • 2

2 4

  • 2
  • 1

1 2

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.2 0.4 0.6 0.8 1.0 scale

M e x . h a t

43

slide-44
SLIDE 44

sine waves with transition

F

  • u

r i e r w a v e l e t

0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Freq. power

  • 4
  • 2

2 4

  • 1.0
  • 0.5

0.5 1.0

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.2 0.4 0.6 0.8 1.0 scale

M e x . h a t

44

slide-45
SLIDE 45

How might this approach improve upon templates?

GeV sky can be thought of as a high resolution picture; wavelets can find structures in it Poisson noise and SM uncertainty dominate at scales that are small relative to bubbles or NFW, and the wavelets can identify those scales

45

slide-46
SLIDE 46

How might this approach improve upon templates?

GeV sky can be thought of as a high resolution picture; wavelets can find structures in it Poisson noise and SM uncertainty dominate at scales that are small relative to bubbles or NFW, and the wavelets can identify those scales

46

slide-47
SLIDE 47

How might this approach improve upon templates?

GeV sky can be thought of as a high resolution picture; wavelets can find structures in it Poisson noise and SM uncertainty dominate at scales that are small relative to bubbles or NFW, and the wavelets can identify those scales

47

slide-48
SLIDE 48

How might this approach improve upon templates?

GeV sky can be thought of as a high resolution picture; wavelets can find structures in it Poisson noise and SM uncertainty dominate at scales that are small relative to bubbles or NFW, and the wavelets can identify those scales by identifying and removing such structures, wavelets provide a background expectation that is (relatively) robust against systematic astrophysics uncertainties

slide-49
SLIDE 49

Example (mock data)

ℓmax=512 256<ℓ<512 0.7˚<θ<1.4˚ ⇒

mock data only diffuse templates subtracted

49

slide-50
SLIDE 50

Example (mock data)

ℓmax=512 128<ℓ<256

1.4˚<θ<3˚

⇒

mock data only diffuse templates subtracted

50

slide-51
SLIDE 51

Example (mock data)

ℓmax=512 64<ℓ<128

3˚<θ<6˚

mock data only diffuse templates subtracted

⇒

51

slide-52
SLIDE 52

Example (mock data)

ℓmax=512 32<ℓ<64

6˚<θ<10˚

mock data only diffuse templates subtracted

⇒

52

slide-53
SLIDE 53

Example (mock data)

ℓmax=512 4<ℓ<256

1.4˚<θ<90˚

mock data only diffuse templates subtracted

⇒

53

slide-54
SLIDE 54

Example (mock data)

ℓmax=512 4<ℓ<128

3˚<θ<90˚

mock data only diffuse templates subtracted

⇒

54

slide-55
SLIDE 55

Example (mock data)

ℓmax=512 4<ℓ<64

6˚<θ<90˚

mock data only diffuse templates subtracted

⇒

55

slide-56
SLIDE 56

Example (mock data)

ℓmax=512 4<ℓ<32

10˚<θ<90˚

mock data only diffuse templates subtracted

⇒

56

slide-57
SLIDE 57

Example (mock data)

ℓmax=512 4<ℓ<16

22˚<θ<90˚

mock data only diffuse templates subtracted

⇒

57

slide-58
SLIDE 58

Lesson:

Getting rid of some wavelet levels can provide a much clearer picture of a signal How can we do this in a data-driven (model- independent) (unbiased) (etc….) way?

Question:

58

slide-59
SLIDE 59

Kolmogorov-Smirnov Test

"KS2 Example" by Bscan - Own work. Licensed under CC0 via Commons - https://commons.wikimedia.org/wiki/ File:KS2_Example.png#/media/ File:KS2_Example.png

maximum distance between two CDFs

59

slide-60
SLIDE 60

Kolmogorov-Smirnov Test

"KS2 Example" by Bscan - Own work. Licensed under CC0 via Commons - https://commons.wikimedia.org/wiki/ File:KS2_Example.png#/media/ File:KS2_Example.png

maximum distance between two CDFs

KS test offers a selection criterion for each wavelet level

60

slide-61
SLIDE 61

“Thresholded” wavelets

wj>={ wj if KS(S | Asimov) > 40% KS(Bi | Asimov) 0 otherwise signal = S set of backgrounds = {Bi} define “cleaned maps:” C>=Σ8j=2 wj>(S) Bi>=Σ8j=2 wj>(Bi)Θ[wj>(S)] B>=avg({Bi>})

61

and “cleaned residual:” ΔC>=C> - B>

slide-62
SLIDE 62

Cleaned Map Method

C> B> ΔC> C> B> ΔC> ΔM ΔM

62

wavelets provide clearer residual than maps

slide-63
SLIDE 63

Cleaned Map Threshold

ΔC> ΔC> 30% as bright is much harder to see

63

slide-64
SLIDE 64

DM vs. Point Sources?

64

slide-65
SLIDE 65

Cleaned Map, Bubbles

ΔC> ΔM

65

slide-66
SLIDE 66

What are wavelets?

Allow analysis sensitive to both position and size wavelets find structures, and the GCE is a qualitatively new structure that we ought to learn more about different structures have “power” at different levels of the decomposition (edges = sharp variation, important first; larger scale objects = broader variation, important later)

66

slide-67
SLIDE 67

Conclusions

Galactic center gamma ray excess is exciting to follow, but still so much more to learn about it Need some less-model-dependent information Wavelets are a promising tool for learning about this data

67

slide-68
SLIDE 68

Conclusions

Much more to do!

Galactic center gamma ray excess is exciting to follow, but still so much more to learn about it Need some less-model-dependent information Wavelets are a promising tool for learning about this data

68

slide-69
SLIDE 69

Thanks!

69