Dark Energy Spectroscopic Instrument

DESI: Dark Energy Spectroscopic Instrument Robert Cahn Lawrence Berkeley Na4onal Lab How We Know there is Dark Energy 2 March 28, 2014 U.C. Davis Gunion Fest GR & Cosmology in One Slide a Λ 4π GN
= −
( ρ + 3p)
a 3
3
2
# a & 8π GN ρ k Λ
− 2+
% ( =
$a'
3
a
3
Alexander Friedmann a is the size-scale of the universe
relative to size today
March 28, 2014 U.C. Davis Gunion Fest Monseigneur Georges Henri Joseph Édouard Lemaître 3 Making the Universe Collapse v = Hr
H = h ×100(km/sec)/Mpc
h ≈ 0.7
v=Hr r 1
1
KE = Δmv 2 = Δm(Hr)2
2
2
# 4π r 3 ρ & GN
PE = −Δm %
(
$ 3 ' r
0=
Δm 1 2 # 4πρ GN &
H −%
(
$ 3 '
2
ρcritical
3H 2
=
= 1.05 ×10 −5 h 2 GeV cm −3
8π GN
Zero total energy. Just enough to stop expansion. March 28, October 23, 2014 2013 Boston U.C. DCavis ollege Gunion Colloquium Fest 4 Energy Budget of the Universe •  Re-­‐write Friedmann-­‐Lemaitre equa4on: Ωm + Ωrad + ΩΛ + Ωk = 1
ρm
Ωm =
ρcrit
H (a) =
Ωrad
March 28, October 23, 2014 2013 ρΛ
ΩΛ =
ρcrit
k
Ωk = − 2
H0
a
= H 0 a −4Ωrad + a −3Ωm + a −2Ωk + a −ε ΩDE
a
1
distance ρrad
=
ρcrit
D(a) =
∫
a
da'
=
2
a' H (a')
z
∫
0
dz'
H (z')
Boston U.C. DCavis ollege Gunion Colloquium Fest 5 Dark Energy EquaHon of State w(a) = p / ρ
From Friedmann-Lemaitre Equations
1
dρ
da
= −3(1+ w(a))ρ
dt
dt
3
ρ (a) = ρ (a = 1)e
∫ daa (1+w(a))
a
MaXer: w=0 Radia4on: w=1/3 Cosmological constant: w=-­‐1 Accelera4ng Universe means w < -­‐1/3 or General Rela4vity fails. March 28, 2014 U.C. Davis Gunion Fest 6 Dark MaIer vs Dark Energy maXer radia4on dark energy Scale-­‐size of universe =a ρ ∝ a −3
March 28, 2014 ρ ∝ a −4
U.C. Davis Gunion Fest ρ ∝ a0
7 Energy Budget of Universe •  Combining three kinds of measurements we learn that –  The Universe is flat. –  32% of energy is maXer. –  68% of energy is “dark” . •  Distribu4on of elements tells us only 5% of energy is ordinary maXer. –  27% of energy is due to “dark maXer” March 28, 2014 U.C. Davis Gunion Fest 8 SNe (binned)
BOSS
MS-DESI(predicted)
Redshift
Scale of the Universe
Relative to Today’s Scale
How Hard is it to Rule out Cosmological Constant? Billions of Years from Today
March 28, 2014 U.C. Davis Gunion Fest 9 Tiny Ripples in Early Universe Cosmic Microwave Background Ripples in early universe imprint standard ruler in cosmic microwave background θ COBE, WMAP, Planck March 28, 2014 U.C. Davis Gunion Fest 10 BAO gives Ruler BAO That paXern is preserved in the distribu4on of the galaxies. By measuring the paXern looking back billions of years we can deduce the expansion history of the universe. BAO at z=0.57
Anderson et al (2012)
March 28, 2014 U.C. Davis Gunion Fest 11 CMB is 2-­‐d BAO is 3-­‐d BAO standard ruler
from Planck
θs = 0.596724 ± 0.00038 deg
BAO standard ruler
from BOSS & DESI
March 28, 2014 U.C. Davis Gunion Fest 12 How BAO Works (c/H)Δz radial
BAO
DAΔθ angular
BAO
March 28, 2014 BAO first detection plot
Eisenstein et al (2005)
U.C. Davis Gunion Fest 13 Best BAO so Far: BOSS 2.5-meter Sloan Telescope
3 deg diameter field-of-view
1000X multiplexed spectrograph
March 28, 2014 U.C. Davis Gunion Fest 14 Lyman-­‐alpha forest: First dark energy results z>2 Forest of absorp4on lines maps loca4on of neutral hydrogen along line-­‐of-­‐sight from quasar. BOSS Light absorbed when stretched to 121.6 nm. March June 11, 28, 2013 2014 DOE U.C. S&T DRavis eview Gunion -­‐ Cosmology Fest BAO at z=2.3
Busca et al (2013)
Slosar et al (2013)
15 85
BOSS Lyman-­‐alpha Sees DeceleraHon! H(z)/(1+z) (km/s/Mpc)
80
H0 75
70
SDSS 65
BOSS Ly-­‐alpha 60
BOSS galaxies 55
50
0.0
Now March 28, 2014 0.5
1.0
1.5
2.0
2.5
3.0
3.5
z
U.C. Davis Gunion Fest 16 From BOSS to DESI • 
• 
• 
• 
• 
• 
• 
• 
• 
Scale up BOSS to a massively parallel fiber-­‐fed spectrometer Broad range of target classes: LRG’s, ELG’s, QSO’s Broad redshim range: 0.5 < z < 1.6, 2.2 < z < 3.5 {region between 0.7 – 1.6 new} Sky area: 14,000 square degrees Number of redshims: 24 million Medium resolu4on spectroscopy, R ~ 4000 Spectroscopy from blue to NIR: 360 nm < z < 980 nm Automated fiber system, Nfiber ~ 5000 Up to 5 year DE survey 5000 fiber actuators
New 8 sq. deg field-ofview corrector
Mayall 4-­‐m Telescope New spectrographs
March 28, 2014 U.C. Davis Gunion Fest 17 DESI Hardware & SoUware Elements [O II] at z=1.34
Data Management Targets Prime Focus Corrector Focal Plate Spectrometer Fiber PosiHoner March 28, 2014 U.C. Davis Gunion Fest Fiber System CCD’s 18 Galaxy Targets Select photometrically, measure spectroscopically. ELG LRG March 28, 2014 U.C. Davis Gunion Fest 19 Spectroscopy 2
D 3/2 → 4 S 3/2
2
D 5/2 → 4 S 3/2
[O II] at z=1.34
OII doublet Z=1.34 LRGs QSO – Lyman –alpha forest March 28, 2014 U.C. Davis Gunion Fest 20 Measure Two-­‐Point CorrelaHon as FuncHon of z BOSS March 28, 2014 U.C. Davis Gunion Fest 21 AnHcipated Quality of DESI Expansion Measurements 85
H(z)/(1+z) (km/s/Mpc)
80
75
70
galaxies Lyman-­‐alpha 65
60
55
50
0.0
March 28, 2014 0.5
1.0
1.5
2.0
z
2.5
U.C. Davis Gunion Fest 3.0
3.5
22 DESI Achieves Space-­‐Based Precision BOSS
distance scale error
2.0%
1.0%
BOSS
WFIRST-2.4
0.5%
MS-DESI
EUCLID
0
March 28, 2014 1
redshift
2
U.C. Davis Gunion Fest 3
23 CorrelaHon FuncHon and Power Spectrum •  The Wiener (1930)-­‐Khinchin (1934) Theorem – naive version due to Einstein (1914): –  “The Fourier transform of the correla4on func4on is the power spectrum” ρ (k)ρ * (k ') =
=
March 28, 2014 ∫ dxe
∫ dxe
ikx
dx 'e−ik ' x ' ρ (x)ρ (x ') =
i(kx−k ' x )
∫ dxe
ikx
dx 'e−ik ' x 'ξ (x − x ')
dx 'e−i(k ' x '−k ' x )ξ (x − x ') = 2πδ (k − k ')ξ (k ')
U.C. Davis Gunion Fest 24 DESI: Not just BAO Power spectrum is Fourier transform of two-­‐point correla4on func4on. Power spectrum tests: General Rela4vity Infla4on Number of neutrinos Sum of the neutrino masses ns : ±0.0022
α s : ±0.0024
Σmν : ±0.024 eV
ΣNν : ±0.056
March 28, 2014 U.C. Davis Gunion Fest 25 Measuring the sum of neutrino masses 2
Δm32
= 2.32 ×10 −3 eV 2
2
Δm21
= 7.50 ×10 −5 eV 2
March 28, 2014 U.C. Davis Gunion Fest 26 RedshiU Space DistorHon •  Can’t measure distance directly. •  Mismeasure if there is “peculiar velocity” 
Assume v = Hrnˆ along line of sight

so peculiar velocity Δv leads to shift

Δr nˆ = Δv ⋅ n nˆ nˆ /H (a)
•  Gravity will amplify all density perturba4ons. δρ (t) = D(t)δρ (t = 0) [now]
March 28, 2014 U.C. Davis Gunion Fest 27 Galaxies vs MaIer •  Assume frac4onal fluctua4on in galaxy density is propor4onal to frac4onal fluctua4on in maXer: δρ galaxy
δρ matter
δgalaxy ≡
=b
= bδmatter
ρ galaxy
ρ matter
Because we observe in redshim space, there is a distor4on of the power spectrum: 
ˆ 2 f )2 P(k)matter,realspace
P(k )galaxy,RSD = (b 2 + (kˆ ⋅ n)
f=
March 28, 2014 d ln D
d ln a
U.C. Davis Gunion Fest 28 r
r (Mpc/h)
(Mpc/h)
RedshiU Space DistorHon at BOSS r
r (Mpc/h)
(Mpc/h)
Line of sight March 28, 2014 U.C. Davis Gunion Fest 29 TesHng General RelaHvity •  The growth func4on D(a) is determined by the maXer density and General Rela4vity. In prac4ce, we measure fσ8, where σ8 sets the scale for P(k). There will be 2% measurements of fσ8 at many values of z. March 28, 2014 U.C. Davis Gunion Fest 30 InflaHon •  Look at power spectrum •  Look for three-­‐point correla4ons (CMB) •  Look a “scale dependence” of bias P(k) = P(k0 )(k / k0 )
March 28, 2014 1
nS (k0 )+ α S ln(k/k0 )
2
U.C. Davis Gunion Fest Planck: nS=0.9614 ±0.0063 aS=-­‐0.015±0.017 31 March 28, 2014 k
U.C. Davis Gunion Fest rms error improvement over Planck + BOSS BAO
7
4
3
2
= 0.063
= 0.084
= 2 × 10− 3
5
ns
= 4 × 10− 3
= 3.8 × 10− 3
m
s
N
1
w’
= 2.2 × 10− 3
= 5.2 × 10− 4
= 0.28
= 0.09
wp
DESI galaxy and LyaF BAO
+galaxy broadband k < 0.2 h/Mpc
+LyaF broadband
= 3.6 × 10− 4
= 0.13
= 0.014
= 0.023
= 0.017
Improvement factors 6
DESI Improves Many Measurements 32 Price Tag [O II] at z=1.34
March 28, 2014 U.C. Davis Gunion Fest 33 Summary •  DESI: best dark energy informa4on @ 2020 •  Modest experiment using exis4ng telescope •  Based on successful BOSS experiment •  Not just dark energy, but GR, infla4on, neutrinos March 28, 2014 U.C. Davis Gunion Fest 34