Ue-Li Pen - 21cm Cosmology

21cm  cosmology  Ue-­‐Li  Pen  

K.  Masui,  E.  Switzer,  R.  Shaw,  L.  Cailin,  G.  Paciga,  K.  Bandura,  J.  Peterson,  T.  

Chang,  Y.  Liao,  X.  Chen,  Y.  Li,  T.  Voytek,  A.  Natarajan,  M.  Dobbs,  M.  Halpern,  

J.R.  Bond,  G.  Hinshaw,  J.  Roy,  Y.  Gupta,  R.  Nityananda,  and  many  more  

Overview  

21cm  physics  Theory:  dark  energy,  gravity  waves,  neutrinos,  enhanced  21cm  pre-­‐reionizaTon  structures.  

ObservaTons:  first  detecTons  and  upper  bounds  (GBT,  GMRT)  

Future:  dedicated  surveys  (CRT,  CHIME,  etc  )  

21cm:  HI  hyperfine  transiTon  

•  Hyperfine  transiTon  when  electron  and  proton  spins  flip:  change  in  magneTc  moment  

•  h  ν  =α4  me  c2  (me/mp)    •  ν=1420.40575  MHz  •  Highly  forbidden,  lifeTme  t~107  years  •  Astronomical  Tme  scales  T>>t  

ApplicaTons  

Hydrogen  Maser  Clock  Milky  way  mapping  Nearby  redshif  surveys  (HIPASS,  ALFALFA)  Cosmological  redshif  surveys  (GBT,  CHIME/CRT/Tianlai)  

High  redshif  universe  (GMRT/LOFAR/MWA/PAPER)  

Kalberla et al 2005, from WMAP LAMBDA

21cm sky at z=0

Image courtesy of NRAO/AUI and Chung et al., Columbia University

Atomic  physics  

-­‐-­‐  energy  splihng=0.068K<<TCMB.  -­‐-­‐  Emission/absorpTon  depends  on  spin  temperature  TS  -­‐-­‐  Typically  collisionally/uv  excited.  -­‐-­‐  for  TS  >>  TCMB,  emissivity  independent  of  TS:  measure  hydrogen  (HI)  mass  robustly.  

21cm:  emission  

δ Tb=τ TS = const

Spin  evoluTon  

Furlanetto Oh & Briggs, 2006

21cm:  observaTons  

•  Hydrogen  is  the  most  abundant  element  in  the  universe  

•  21cm  line  opTcally  thin  from  the  ground  over  most  lines  of  sight  to  very  high  z  (~100)  

21cm:  observaTonal  challenges  •  Inverse  square  law:  required  collecTng  area  proporTonal  to  distance2.  

•  Highest  redshif  direct  galaxy  detecTons  at  z~0.2  using  2000  hours  of  WSRT  (5000m2)  

•  For  cosmological  interests  (1<z<100)    and  wide  sky  maps  needs  km2  array:  SKA  

Technical  challenges  

•  SensiTvity:  bigger  telescopes?  •  Foreground  contaminaTon:  design/analysis?  •  Target:  Galaxies,  Intensity  Mapping/BAO,  EoR?  

The  21  cm  universe  •  Cosmological  LSS  

treasure  trove  (UP04,  Loeb&Zaldarriaga  04,  Lewis&Challinor  07,  etc)  

•  Up  to  1018  modes:    (Jeans/Hubble)3  

•  Physics:  Lensing,  gravity  waves,  primordial  NG,  BAO,  AP  

•  GW  to  r  ~  10-­‐8  

•   fNL~  10-­‐4  

•  Astrophysics:  EoR,  galaxy  evoluTon  

•  Experiments  NOW   Tegmark & Zaldarriaga 08

EoR: GMRT/ LOFAR MWA Paper

CHIME/GBT

SDSS

Fundamental  Physics  

•  0<z<2:  dark  energy,  BAO,  w-­‐w'  •  z>2:  Large  angle  lensing:  modified  gravity  (Lu  &  UP  2009,  Masui  et  al  2010)  

•  z>10:  gravitaTonal  waves  

Intensity  Mapping  •  Stars  get  fainter  with  distance:  hard  to  see  individually  at  cosmological  distance.    Galaxies  sTll  visible.  

•  Galaxies  get  fainter  with  distance:  hard  to  see  in  HI.    Large  scale  structure  sTll  visible?  

•  Large  scale  structure  is  LARGE:  degree  scale.    High  resoluTon  not  needed.  

•  Modest  size,  monolithic  radio  telescopes  needed.    (CPPM  2008,  Wyithe&Loeb  2008)  

From: talk by O. Lahav

Foreground:  GalacTc  Synchrotron  

Haslam 408 MHz Much brighter than signal, but no spectral structure

Robert  C  Byrd  Telescope:  100m  

HI  content  at  z=0.8  Cross-­‐correlaTng  GBT  HI  &  DEEP2  opTcal  galaxies  at      

z  ~  0.7-­‐1.1  (Chang  et  al,  Nature  2010)  GBT radio continuum

sources + HI

GBT HI

DEEP2 density

Chang, Pen, Bandura, Peterson, Nature, 2010

•  Measure HI & optical cross-correlation on 9 Mpc (spatial) x 2 Mpc (redshift) comoving scales

•  HI brightness temperature on these scales at z=0.8:

• 

•  Highest-redshift detection of HI in emission at 4-sigma statistical significance.

GBT 15h auto power

IniTal  Intensity  Mapping  

•  Detected  collecTve  large  scale  structure  with  100h  of  GBT  Tme:  first  demonstraTon  of  distant  IM.    No  individual  galaxies  detected,  many  galaxies  per  resoluTon  element  

•  Proposal  submited  for  z=1  BAO  survey  on  4  pixel  array  at  GBT  

Baryon  AcousTc  OscillaTons  –  Dark  Energy  Probe  

•  CMB  acousTc  oscillaTons:  imprinted  standard  ruler,  100  Mpc.  

•  Present  in  current  mater  distribuTon  

•  KinemaTc  metric  of  universe  

WMAP5  and  other,  Nolta  et  al  (2008)  

Present  LSS  BAO  DetecTons  

•  Percival  et  al  2007  

Percival  et  al  2007  

Eisenstein  et  al  2005  

Blake et al 2011: WiggleZ, z~0.6 Masui et al: GBT-BAO proposal 2012, z~1

Dedicated  Survey  Experiment  

•  Low  frequency  technology  cheap,  modest  size:  (100  m)2  to  z<2  

•  Large  field  of  view:  receiver  arrays  •  High  surface  brightness  sensiTvity:  compact  arrays  

•  Stable,  reliable:  no  moving  parts  •  Technologies:  aperture  arrays  (Wyithe,  Loeb,  Geil  2008),  cylinders  (Peterson  et  al)  

Molonglo  

Molonglo  

AUSTRALIA  Brisbane  

Darwin  

Perth  Canberra  

Hobart  

Adelaide  Melbourne  

Sydney  

+  Northern  

Cross  

Cambridge  

Pushchino

Ooty

CMU  cylinder  in  operaTon:    J.  Peterson,  U.  Pen,  U.  Seljak,  K.  Bandura,  K.  Sigurdson  

CHIME, Tianlai, CRT, etc

Fast Fourier Transform Telescope

New  Cylinder  Radio  Telescopes  

Canada,  US,  China,  France,  Morocco  Map  the  Hubble  volume  to  0.8<z<2.5  Most  sensiTve  BAO,  RSD,  lensing  survey  Ancillary  science:  21cm  absorbers,  radio  transients,  pulsar  search/monitoring,  magneTc  fields  

Scalable  to  z~20:  tens  of  square  kilometers    

Gravity  Waves  

•  Masui  &  Pen  2010    (PRL  105,  161302)  •  Analogous  to  lensing:  shearing  of  cosmic  structure.  •  Fossil  memory  effect:  h~10-­‐6  (inflaTon)  •  Measure  r,nT  at  z~15  •  Requires  1/h2  modes  •  Separates  from  lensing:  transverse  traceless  •  Structures  available  to  k~10-­‐3-­‐103:horizon  to  Jeans  scale,  ~1018  modes,  peaks  at  z~15  

Linear  gravity  wave  memory  •  GW  in  iniTal  condiTon,  then  redshifs  away  

Detectability  

Conclusions  •  21cm  cosmology:  probes  of  dark  energy  (BAO),  modified  gravity  (lensing),  InflaTon  (tensor  modes),  etc.    Beyond  intrinsic  power  spectrum.  PotenTal  measure  of  gravity  waves.  

•  Intensity  Mapping:  21cm  unresolved  galaxies,  accessible  in  redshif  desert  z=1-­‐3,  iniTal  HI  detecTon  and  surveys  with  GBT  at  z~1.    Upper  bounds  at  z=9.  

•  Prototypes  and  observaTons  under  way.    Cylinder  telescopes  a  promising  technology  for  fast,  large,  economic  surveys.    OpTmal  calibraTon  and  dynamic  range.