AMS-02 ANTIPROTONS ARE CONSISTENT WITH A SECONDARY ASTROPHYSICAL - - PowerPoint PPT Presentation

▶
ams 02 antiprotons are consistent with a secondary
SMART_READER_LITE
LIVE PREVIEW

AMS-02 ANTIPROTONS ARE CONSISTENT WITH A SECONDARY ASTROPHYSICAL - - PowerPoint PPT Presentation

Mar 22, 2023 181 likes •349 views

AMS-02 ANTIPROTONS ARE CONSISTENT WITH A SECONDARY ASTROPHYSICAL ORIGIN arXiv:1906.07119 M.B, Y. Gnolini, L. Derome, J. Lavalle, D. Maurin, P. Salati and P. D. Serpico Table of contents 1. Secondary antiprotons: a new prediction 2.

slide-1
SLIDE 1

1

AMS-02 ANTIPROTONS ARE CONSISTENT WITH A SECONDARY ASTROPHYSICAL ORIGIN

arXiv:1906.07119 M.B, Y. Génolini, L. Derome, J. Lavalle, D. Maurin, P. Salati and P. D. Serpico Table of contents

  • 1. Secondary antiprotons: a new prediction
  • 2. Propagation of uncertainties
  • 3. Prediction vs AMS-02 data
  • 4. Conclusion
slide-2
SLIDE 2

2

Secondary antiprotons: a new prediction

2015: AMS-02 antiprotons, at last! Secondary astrophysical component and immediate implications for DM

Giesen+(2015) arXiv:1504.04276

slide-3
SLIDE 3

3

2015: AMS-02 antiprotons, at last! Secondary astrophysical component and immediate implications for DM New inputs since 2015 DATA MODEL AMS-02

  • pbar flux
  • H, He, C, N, O fluxes

(most abundant CRs ⟹ main pbar parents)

  • B/C ratio (CR tranport)
  • Systematic errors

AMS-02 collaboration does not provide the covariance matrix of errors ⟹ homemade covariance matrix based on the description of systematics in AMS-02 papers CR transport New models from AMS-02 B/C (Génolini et al.)

  • QUAINT: historical model diff/conv/reac
  • SLIM: pure diffusion w/ 2 breaks in the diff. coef.
  • BIG: diffusion/convection/reacceleration + 2

breaks in the diffusion coef. (QUAINT, SLIM) ⊂ BIG Production XS

  • NA61: p+p —> pbar + X

√s = 7.7, 8.8, 12.3 and 17.3 GeV Tp = 31, 40, 80 and 158 GeV

  • LHCb: p+He —> pbar + X

Tp = 6.5 TeV Production XS

  • Prompt pbar in pp reactions

New parametrisations of the Lorentz invariant production XS σinv = Ed3σ/dp3

Winkler+(2016), Korsmeier+(2018)

  • Antihyperons (Δ𝛭) and isospin asymmetry (ΔIS)

New energy dependant (√s) parametrisations

Giesen+(2015) arXiv:1504.04276

Secondary antiprotons: a new prediction

slide-4
SLIDE 4

101 102 103 104

√s [GeV]

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

¯ Λ/¯ p

BHM NAL MIRABELLE NA49 30-in ISR STAR ALICE CMS

median parameters median 68% median parameters median 68%

4

Production XS

  • Prompt pbar in pp reactions (p + p —> pbar + X)

Parametrisation II from Korsmeier+(2018)

  • functional form of σinv(√s, xR, pT) from Winkler+(2016)
  • updated parameters using NA49, NA61, BRAHMS, Dekkers+(1965) Korsmeier+(2018)
  • Prompt pbar in AA reactions (A1 + A2—> pbar + X)

Parametrisation B from Korsmeier+(2018)

  • functional form of the nucleon scaling fA1A2(√s, xF) from Winkler+(2016)
  • updated parameters using LHCb data (p + He —> pbar + X) Korsmeier+(2018)
  • Antihyperons correction (p + p —> X + [(𝛭bar, Σbar) —> pbar])

Parametrisation of Δ𝛭(√s) from Winkler+(2016)

∆Λ(√s) = (0.81 ± 0.04)(¯ Λ/¯ p)

  • Isospin asymmetry correction (p + p —> X + [nbar—> pbar])

We introduce the following parametrisation to reproduce the results of

Winkler+(2016)

∆IS(√s) = c0(x + c2)c3 exp(−x/c1), x = log(√s)

101 102 103 104

√s [GeV]

−0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

∆IS

starting parameters median 68% NA49 pC NA49 np Fermilab STAR ALICE NA49 pC NA49 np Fermilab STAR ALICE

slide-5
SLIDE 5

101 102 103 104

√s [GeV]

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

¯ Λ/¯ p

BHM NAL MIRABELLE NA49 30-in ISR STAR ALICE CMS

median parameters median 68% median parameters median 68%

5

Production XS

  • Prompt pbar in pp reactions (p + p —> pbar + X)

Parametrisation II from Korsmeier+(2018)

  • functional form of σinv(√s, xR, pT) from Winkler+(2016)
  • updated parameters using NA49, NA61, BRAHMS, Dekkers+(1965) Korsmeier+(2018)
  • Prompt pbar in AA reactions (A1 + A2—> pbar + X)

Parametrisation B from Korsmeier+(2018)

  • functional form of the nucleon scaling fA1A2(√s, xF) from Winkler+(2016)
  • updated parameters using LHCb data (p + He —> pbar + X) Korsmeier+(2018)
  • Antihyperons correction (p + p —> X + [(𝛭bar, Σbar) —> pbar])

Parametrisation of Δ𝛭(√s) from Winkler+(2016)

∆Λ(√s) = (0.81 ± 0.04)(¯ Λ/¯ p)

  • Isospin asymmetry correction (p + p —> X + [nbar—> pbar])

We introduce the following parametrisation to reproduce the results of

Winkler+(2016)

∆IS(√s) = c0(x + c2)c3 exp(−x/c1), x = log(√s)

101 102 103 104

√s [GeV]

−0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

∆IS

starting parameters median 68% NA49 pC NA49 np Fermilab STAR ALICE NA49 pC NA49 np Fermilab STAR ALICE

σtot

inv = σinv(2 + ∆IS + 2∆Λ)

slide-6
SLIDE 6

6

A model for the covariance matrix of AMS-02 errors

  • Covariance matrix (see David’s talk)

Coefficient corresponding to the uncertainty 𝛽 (Acc., Unf., Trig., etc.)

(Cα)ij = σα

i σα j exp

✓ −1 2 (log10(Ri/Rj)2 (lα

ρ )2

◆

  • 𝝍2 calculation

Quadratic distance between model and data:

χ2 = X

i,j

xi

  • C−1

ij xj ≡ xTC−1x

xi = datai − modeli

  • Visual inspection
  • Standard z-score

Misleading when correlations between data points

zi = xi/ p Cii

  • Rotated z-score

˜ xi = Uij xj, ˜

C = U C U T

˜ C ˜ Cii = ˜ σ2

i

is diagonal with elements

˜ zi = ˜ xi/˜ σi

χ2 = X

i

˜ z2

i

Rotated rigidity

˜ Ri = X

j

U 2

ij Rj,

˜ Ri ' Ri

slide-7
SLIDE 7

7

Model of CR transport

New models from AMS-02 B/C (see Yoann’s talk)

  • QUAINT: historical model diff/conv/reac
  • SLIM: pure diffusion w/ 2 breaks in the diff. coef.
  • BIG: diffusion/convection/reacceleration + 2 breaks in the diffusion coef. (BASELINE)

0.05 0.10 0.15 0.20 0.25 0.30 0.35 B/C

BIG SLIM QUAINT AMS-02 Data

100 101 102 103 Rigidity [GV] −2 2 Z-score [σtot]

100 101 102 103 R [GV] 10−1 100 K(R) for A/Z = 2 [kpc2.Myr−1]

BIG SLIM QUAINT

100 101 0.2 0.3 0.5

4He

p e−

slide-8
SLIDE 8

8

Antiproton parents

Combined fit of AMS-02 H, He, C and O Most abundant CRs ⟹ main pbar parents

BIG

Excluded from fit

101 102 103 R [GV] −3 −1 1 3 z-score [σtot] H He C O BIG

Excluded from fit

101 102 103 ˜ R [GV] −3 −1 1 3 ˜ z-score [˜ σeigen

tot ]

BIG −4 −2 2 4 ˜ z-score [˜ σeigenv

tot

] 0.0 0.2 0.4 0.6 0.8 Distribution H He C O SLIM

Excluded from fit

101 102 103 R [GV] −3 −1 1 3 z-score [σtot] H He C O SLIM

Excluded from fit

101 102 103 ˜ R [GV] −3 −1 1 3 ˜ z-score [˜ σeigen

tot ]

SLIM −4 −2 2 4 ˜ z-score [˜ σeigenv

tot

] 0.0 0.2 0.4 0.6 0.8 Distribution H He C O QUAINT

Excluded from fit

101 102 103 R [GV] −3 −1 1 3 z-score [σtot] H He C O QUAINT

Excluded from fit

101 102 103 ˜ R [GV] −3 −1 1 3 ˜ z-score [˜ σeigen

tot ]

QUAINT −4 −2 2 4 ˜ z-score [˜ σeigenv

tot

] 0.0 0.2 0.4 0.6 0.8 Distribution H He C O

slide-9
SLIDE 9

9

Antiproton parents

Combined fit of AMS-02 H, He, C and O Most abundant CRs ⟹ main pbar parents

BIG

Excluded from fit

101 102 103 R [GV] −3 −1 1 3 z-score [σtot] H He C O BIG

Excluded from fit

101 102 103 ˜ R [GV] −3 −1 1 3 ˜ z-score [˜ σeigen

tot ]

BIG −4 −2 2 4 ˜ z-score [˜ σeigenv

tot

] 0.0 0.2 0.4 0.6 0.8 Distribution H He C O SLIM

Excluded from fit

101 102 103 R [GV] −3 −1 1 3 z-score [σtot] H He C O SLIM

Excluded from fit

101 102 103 ˜ R [GV] −3 −1 1 3 ˜ z-score [˜ σeigen

tot ]

SLIM −4 −2 2 4 ˜ z-score [˜ σeigenv

tot

] 0.0 0.2 0.4 0.6 0.8 Distribution H He C O QUAINT

Excluded from fit

101 102 103 R [GV] −3 −1 1 3 z-score [σtot] H He C O QUAINT

Excluded from fit

101 102 103 ˜ R [GV] −3 −1 1 3 ˜ z-score [˜ σeigen

tot ]

QUAINT −4 −2 2 4 ˜ z-score [˜ σeigenv

tot

] 0.0 0.2 0.4 0.6 0.8 Distribution H He C O

scores distribution close to Gaussian (𝜈=0, 𝜏=1) as expected from statistical fluctuations

slide-10
SLIDE 10

10

Ranking parents contributions

slide-11
SLIDE 11

11

Heavy parents, heavy species in the ISM

10−1 100 101 102 103 104 R [GV] 0.96 0.97 0.98 0.99 1.00 1.01 1.02 1.03 1.04 ( ¯ p / ¯ pref )TOA

BIG

58Fe

H,He

(BIG & ¯ p)

58Fe

H,He

BIG

58Fe

H...Fe

(BIG & ¯ p)

58Fe

H...Fe

We rescale our pbar prediction by the black solid line to account for heavy species in the ISM Fast calculation (ref)

  • pbar and B/C parents: H … 30Si
  • ISM species: H, He

Full (slow) calculation

  • pbar and B/C parents: H … 58Fe
  • ISM species: H…Fe
slide-12
SLIDE 12

12

Propagation of uncertainties

Uncertainties on parameters entering:

  • Production XS (fit collider data)
  • Transport (fit B/C)
  • Parents (fit H, He, C and O)

Assume parameter distribution is Gaussian ⟹ use the covariance matrix of errors to propagate uncertainties and their correlations In practice

  • 1. Draw randomly 10000 pbar predictions from the covariance matrix for each source of

uncertainties (XS, Transport, Parents)

  • 2. Determine the 1𝜏 confidence intervals

R [GV] R [GV]

1 10 100 1000 1 10 100 1000

slide-13
SLIDE 13

13

Covariance matrix of model uncertainties

  • For the source of uncertainty a ∈ (XS, Transport, Parents)

Ca

ij = 1

N

N

X

n=1

  • Φa

i,n − µa i

Φa

j,n − µa j

  • µa

i

mean prediction at the energy bin i

1 3 10 30 100 R [GV] XS Transport 1 3 10 30 100 R [GV] 1 3 10 30 100 R [GV] Parents 1 3 10 30 100 R [GV] Total 0.5 0.6 0.7 0.8 0.9 1.0

  • Associated correlation matrix:

cα

ij =

Cα

ij

p Cα

ii

pCα

jj

slide-14
SLIDE 14

14

Prediction vs AMS-02: visual inspection

R [GV] 10−2 10−1 100 101 ΦTOA

¯ p

[GV−1 m−2 s−1 sr−1] × R3

AMS-02 (σtot) Baseline prediction Total uncertainties AMS-02 (σtot) Baseline prediction Total uncertainties

R [GV] −40 −20 20 40 Residuals [%]

Parents XS Transport Total Parents XS Transport Total

1 10 100 103 R [GV] −3 −1 1 3 ˜ z-score [˜ σtot] ˜ z-score

C = Cdata + Cmodel

1 3 10 30 100 R [GV] XS Transport 1 3 10 30 100 R [GV] 1 3 10 30 100 R [GV] Parents 1 3 10 30 100 R [GV] Total 0.5 0.6 0.7 0.8 0.9 1.0

1 3 10 30 100 R [GV] Stat. Cut-off Sel. 1 3 10 30 100 R [GV] Temp. XS Unf. 1 3 10 30 100 R [GV] 1 3 10 30 100 R [GV] Scale 1 3 10 30 100 R [GV] Acc. 1 3 10 30 100 R [GV] Total 0.0 0.2 0.4 0.6 0.8 1.0

data model

scores distribution close to Gaussian (𝜈=0, 𝜏=1) as expected from statistical fluctuations

slide-15
SLIDE 15

15

Statistical test: 𝝍2 vs Kolmogorov-Smirnov (KS)

xi = datai − modeli

χ2 = X

i,j

xi

  • C−1

ij xj,

C = Cdata + Cmodel

q σ2

stat + σ2 syst

🤕

𝝍2-test

  • Does not assess a possible overestimate of errors
  • Rely on the notion of degrees of freedom which is not well defined when correlations

KS-test

  • Assess a possible overestimate of errors
  • Does not rely on the notion of degrees of freedom 🤔

(i) No model uncertainty + data uncertainty: good agreement for both KS and 𝝍2 (ii) Model uncertainty + data statistics only: KS: good agreement, 𝝍2 marginally consistent (2𝜏) (iii) Model uncertainty + data uncertainty: good agreement for both KS and 𝝍2

  • AMS-02 data are consistent w/ a pure secondary origin
  • Conclusion robust wrt error mismodelling of model or data
slide-16
SLIDE 16

16

Conclusion

AMS-02 data are consistent w/ a pure secondary origin

  • Conclusion robust wrt the statistical test used (𝝍2-test or KS-test)
  • Conclusion robust wrt error mismodelling of model or data

Prediction of the pbar flux computed from external data (not a fit)

  • Use the latest constraints on transport parameters from AMS-02 B/C
  • Propagate all the uncertainties (and their correlation) to the pbar flux prediction
  • Account for correlated errors in pbar data

Thank you for your attention! Questions?