One-dimensional power spectrum
from first DESI Lyman-α forest

Corentin Ravoux - Researcher at LPCA, Clermont-Ferrand, France

COSMO'2024 Kyoto

 

on behalf of the DESI Lyman-α team

The Lyman-α forest

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

Ravoux et al. 2023 (MNRAS)

  • Lines in quasar spectra at                                      caused by absorbers in the intergalactic medium (IGM) at
\lambda_{\mathrm{obs}} = \left(1 + z_{\mathrm{abs}}\right) \lambda_{\mathrm{Ly\alpha}}
z_{\mathrm{abs}}

Lyman-α forest = non-linear tracer of neutral hydrogen in the intergalactic medium

DESI quasar z = 3.42

Contaminants

  • Near the quasar:
    • Intrinsic continuum
    • Broad absorption line quasars (BAL)
  • Along the line-of-sight:
    • Metal absorptions (C, Si, O, N...)
    • Damped Lyman-α systems (DLA)
  • Near the telescope:
    • Atmospheric absorption and emission
    • Instrument noise
    • Spectrograph resolution

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

One-dimensional power spectrum (P1D)

Quasar

Lyman-α forest

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

One-dimensional power spectrum (P1D)

Correlations along individual lines-of-sight in Fourier Space

Quasar

Lyman-α forest

P_{\mathrm{1D}} = \langle \left| \delta_F(k) \right|^2 \rangle_{\mathrm{forests}}
\delta_F = \frac{F}{\langle F \rangle} - 1

= Fraction of transmitted flux

F

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

Previous measurement

  • Sensitive to small-scale matter clustering and IGM thermal state

  • Probe the small-scale matter power spectrum

 

 

 

  • Intermediate resolution measurement on eBOSS DR14

Small scales

Large scales

Chabanier et al. 2019 (JCAP)

\Omega_{\mathrm{m}}, \sigma_8, n_s
A_{\mathrm{Ly\alpha}}, n_{\mathrm{Ly\alpha}}

Amplitude and slope at

z=3\ \&\ k=1\ h/ \mathrm{Mpc}

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

Neutrino masses and dark matter models

P1D range

CMB range

Baur et al. 2016 (JCAP)

  • Matter power spectrum impacted by:
    • Sum of neutrino masses
    • Dark matter model (e.g. warm dark matter)

Unique probe to constrain neutrino masses and dark matter properties

\sum m_{\nu}

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

P1D range

Baur et al. 2016 (JCAP)

Latest constraints with eBOSS

  • Neutrino mass  (P1D +CMB):
\sum m_{\nu} < 0.11~\mathrm{eV}
m_{\mathrm{X}} > 5.3~\mathrm{keV}
  • Warm dark matter model:

Palanque-Delabrouille et al. 2020 (JCAP)

  • Other constraints: Fuzzy dark matter, sterile neutrinos, running of the primordial power spectrum

at 95% C.L.

at 95% C.L.

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

P1D measurement with DESI

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

DESI (2021 - 2026)

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

  • 4 m telescope at Kitt Peak Observatory
  • 1 million Lyman-α forests expected for Y5
  • Early data release public
  • Y1 results on BAO (Cullan Howlett presentation)

DESI focal plane

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

  • 5 μm positioning in few sec. for 5000 targets
  • All fibers connected to 10 spectrographs

Spectral resolution improvement

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

R = \lambda / \Delta \lambda
  • eBOSS:

 

  • DESI:
  • Spectral resolution

Resolution power

R \sim 1500 - 2500
R \sim 2000 - 5000

DESI Lyman-α forests can probe smaller scales

P1D FFT estimator

Without correction, Fourrier transform of the Lyman-α constrast

\begin{aligned} P_{1\mathrm{D},\alpha}(k) = & \left(\left\langle\left[P_{\mathrm{raw}}(k)-P_{\mathrm {pipeline}}(k) - \alpha \right] \cdot \mathbf{R}^{-2}(k) \right\rangle - P_{\mathrm{SB1,m}}(k) \right) \\ & \cdot A_{\mathrm{line}}(z,k) \cdot A_{\mathrm{dla}}(z,k) \cdot A_{\mathrm{cont}}(z,k) \cdot A_{\mathrm{res}}(k) \end{aligned}
= \left\langle|\delta_F(k)|^2\right\rangle
\big \rangle

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

P1D FFT estimator

\begin{aligned} P_{1\mathrm{D},\alpha}(k) = & \left(\left\langle\left[P_{\mathrm{raw}}(k)-P_{\mathrm {pipeline}}(k) - \alpha \right] \cdot \mathbf{R}^{-2}(k) \right\rangle - P_{\mathrm{SB1,m}}(k) \right) \\ & \cdot A_{\mathrm{line}}(z,k) \cdot A_{\mathrm{dla}}(z,k) \cdot A_{\mathrm{cont}}(z,k) \cdot A_{\mathrm{res}}(k) \end{aligned}

Correction of the DESI pipeline noise estimation

Spectroscopic resolution correction adapted to DESI pipeline

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

P1D FFT estimator

\begin{aligned} P_{1\mathrm{D},\alpha}(k) = & \left(\left\langle\left[P_{\mathrm{raw}}(k)-P_{\mathrm {pipeline}}(k) - \alpha \right] \cdot \mathbf{R}^{-2}(k) \right\rangle - P_{\mathrm{SB1,m}}(k) \right) \\ & \cdot A_{\mathrm{line}}(z,k) \cdot A_{\mathrm{dla}}(z,k) \cdot A_{\mathrm{cont}}(z,k) \cdot A_{\mathrm{res}}(k) \end{aligned}

Measurement of  absorption from other IGM elements (metals)

Correction of the DESI pipeline noise estimation

Spectroscopic resolution correction adapted to DESI pipeline

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

P1D FFT estimator

\begin{aligned} P_{1\mathrm{D},\alpha}(k) = & \left(\left\langle\left[P_{\mathrm{raw}}(k)-P_{\mathrm {pipeline}}(k) - \alpha \right] \cdot \mathbf{R}^{-2}(k) \right\rangle - P_{\mathrm{SB1,m}}(k) \right) \\ & \cdot A_{\mathrm{line}}(z,k) \cdot A_{\mathrm{dla}}(z,k) \cdot A_{\mathrm{cont}}(z,k) \cdot A_{\mathrm{res}}(k) \end{aligned}

Masking of atmospheric lines

Masking of Damped Lyman-α systems

Correction of continuum fitting error

Correction of the resolution estimation

Measurement of  absorption from other IGM elements (metals)

Correction of the DESI pipeline noise estimation

Spectroscopic resolution correction adapted to DESI pipeline

All effects added to the systematic error budget

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

DESI P1D measurements

  • Early data release & first two months
  • 26,330 Lyman-α forests
  • Statistical & systematics error bars
  • Good agreement with alternative estimator (QMLE)

FFT

QMLE

Ravoux et al. 2023 (MNRAS)

Karaçaylı et al. 2024 (MNRAS)

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

Comparison with previous surveys

Ravoux et al. 2023 (MNRAS)

Ravoux et al. 2023 (MNRAS)

eBOSS DR14 measurement

Chabanier et al. 2019 (JCAP)

High resolution measurement

Karaçaylı et al. 2022 (MNRAS)

  • Good agreement for intermediate wavenumbers
  • Discrepancy at large-scale (possible cause: DLA contamination)

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

Cosmological interpretation

Walther et al. 2021 (JCAP)

  • Interpretation of P1D with high resolution hydrodynamical simulations:
    • Nyx grid
    • MP-Gadget grid
  • Gaussian processes
    emulators with adapted
    cosmological/IGM parameters
    • Interpolation between simulations
    • Covers cosmological parameter space

Walther et al. 2021 (JCAP)

Pedersen et al. 2021 (JCAP)

Cabayol et al. 2023 (MNRAS)

Pedersen et al. 2023 (ApJ)

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

Forecasts for DESI

\sigma\left(\sum m_{\nu}\right) = 0.03~\mathrm{eV}
  • Constraints on WDM improved by a factor 1.6
  • IGM thermal parameters by a factor 2.
  • Neutrino mass, in association with BAO and CMB:

 Valluri et al. 2022

Credits to A. Font-Ribera

Preliminary

Corentin Ravoux                                                                                                                                             P1D from first DESI Data - COSMO2024 Kyoto

  • P1D is probing small-scale matter power spectrum
  • First P1D measurement with DESI data
  • Active work on DESI first year data set and cosmological interpretations

Conclusion

Thanks to our sponsors and 72
participating institutions

DESI comparison to eBOSS

  • Improved resolution compared to eBOSS
    • eBOSS: R ~ 1500 - 2300
    • DESI: R ~ 3000 - 5000

 

  • End of survey:
    • eBOSS: ~250,000 Lyman-α forests
    • DESI: ~1 million Lyman-α forests

Noise estimation

  • Noise computed with DESI spectroscopic pipeline
  • Alternative noise estimation:
    • Resolution suppresses physical signal at small scales
    • Use asymptotic value of raw power spectrum
  • Comparison with pipeline:
    • Done with successive data reductions
    • Helped to improve noise model in pipeline

Noise correction

  • Use of pipeline noise with additive correction.

 

  • Different corrections for three DESI data sets with specific observation strategies (SV1, SV3, M2)

Metal power spectrum

  • Contribution from metal absorptions statistically computed using side bands
  • Oscillations in side band power spectrum: CIV and SiIV doublets
  • Physically motivated model to closely reproduce side band power spectrum
  • Model subtracted to P1D measurement

Atmospheric emission lines

  • Lines from atmosphere removed by pipeline but increase noise and impact P1D
  • Build a new atmospheric line mask using DESI sky fibers
  • Automated algorithm based on sky flux level
  • Reduced masked length compared to eBOSS

Corrections with mocks

  • Characterization of effects from:
    • Masking atmospheric lines
    • Masking DLAs
    • Continuum fitting method
  • Mocks with uncorrelated lines-of-sight and well-defined P1D (Ohio mocks)
    • Compare "truth" to measured P1D
  • Correction applied on the P1D data measurement

Uncertainties

  • Statistical uncertainties from FFT mode counting

 

  • Estimation of systematic uncertainties similarly to eBOSS measurement:
    • Lower statistics with respect to eBOSS
    • Relatively good systematics/statistics ratio

Fuzzy dark matter

m_{\mathrm{a}}>2 \times 10^{-21} \mathrm{eV}
  • Fuzzy Dark Matter (Armengaud et al. 2017, Irsic et al. 2017):
    • De Broglie length close to structure formation and DM halo dynamics
    • Smooth the density fluctuation by quantum wave effects
  • Constraint by P1D:

Nyx simulation sub-grid physics

  • Nyx = Hydrodynamical code on grid + Dark matter particles on PM scheme
    • Lyman-α forest not very sensitive to very dense IGM regions
    • AMR is not adapted
  • Other physical processes modeled in Nyx:
    • Gas chemistry = fixed composition with H and He abundance
    • Inverse Compton + atomic collisional processes
  • Effects not included:
    • Thermal feedback from AGN or supernovae
    • Inhomogeneous radiative background (UV)
  • High redshifts: full reionization history (assumed homogeneous)
  • Choice: No explicit simulation of these effects but taken into account as a nuisance at the fitting stage
  • Example: AGN effect on P1D accounted for (Chabanier et al. 2020, Horizon-AGN simulation)
  • More modeling effort needed to take into account other effects.

AGN feedback

  • Physical effect = baryons and temperature redistribution in the IGM
  • P1D correction, using different feedback parameters with HorizonAGN simulations

DESI Resolution