Rise and Fall of HI

Rise and Fall of HI

Ue-Li Pen 彭威礼

Jeff Peterson, CMU

Radiative Xfer simulation by Iliev, Mellema & Pen

Search for Cosmic Hydrogen

• About half of horizon volume filled with HI, dominated by z>6

• hi-z: lots out there, structures with up to 1016 Mּס HI. Difficult to observe window, foregrounds.

• Mid-z: neutral HI in galaxies, 0.1<z<2. Billions of galaxies, HI mass function evolution poorly known. Intrinsically faint (μJy), requires large collecting area (>105 m2). GMRT/SKA.

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

Traditional Radio Telescope Cost Drivers

• High Frequency: cryogenic receivers, surface, pointing accuracy

• Correlation/bandwidth: N2 cost• General purpose – steerable, reconfigurable• HSHS target: $10/m2, <1.4 GHz, transit• Molonglo actual: $12/m2, steerable cyl• GMRT actual: $100/m2, <1.4 GHz• SKA target: $1000/m2

• VLA actual: $10000/m2

HSHS: Large Area, Low Cost

• Surface: low frequency mesh (GMRT)

• Structure: passive transit cylinder

• Correlations: FFT is N log N instead of N2

• Signal processing: PC’s are $3/Gflop

• Astro-ph/0606104

Existing operational cylinders

MOLONGLO1600x12mCost:$12/m2

(current dollars)

OOTY530x30mBoth rotate in one dimension

Molonglo

AUSTRALIABrisbane

Darwin

PerthCanberra

Hobart

Adelaide

Melbourne

Sydney

+

Cylinder History

• Popular 1960-1980• Lost favor with advent of cryogenically cooled pre-amplifiers.• Room temp amplifiers with 20K noise temp now available.

Illinois 400 ftTelescope ca. 1960

Local HI Luminosity Function

Zwann et al

at z=1.5 this is 30 microjanskyWe detect all these across 1/2 of sky in 6 months

Cosmic Magnification

• Cosmic shear has evolved as a direct way to map dark matter

• Several major surveys under way or planned – CFHTLS, LSST, SNAP

• Anticipated limitations: redshift distribution, PSF

• With redshifts, these limits can be overcome, and magnification is measured directly

• Measured through cross correlation in SDSS (Scranton et al 2005)

• Forecasts and models by Zhang and Pen (2005, 2006): overcomes intrinsic clustering.

Gravitational Lensing: increase flux, decrease density

Magnification: increases number of bright galaxies, decreases faint ones.

Four Cylinders each 2 km long, 50m wide

• Line feeds at foci used to create 4000 beams N

2 km

CMU Prototype cylinders under construction. Funded by Seljak/Packard

Sept 17 status

Line Feed

LNA, $1.81

Filters,$1.99 ea.

HI Evolution

• Major cost uncertainty is luminosity function evolution.

• Popular models differ by factor of several

• Effort under way to measure z=1.4 mass function using DEEP cross correlation (T. Chang, M. Davis, U. Seljak)

τ=0.09+/-0.03, zr=11

WMAP 3yr: rise of HI

Cosmic precombination

White et al 2003: universe fully ionized at z<6

Barkana & Loeb 2001

Reionization

• First objects: • 21cm @ z=6-15• 90-200 Mhz• ΔT = 23 mK, • ~μJy- mJy (up to 1016

Mּס of HI)• Angular scale 5’<Θ<3

0’, freq res 500 khz• Challenging theory

z=19-12 simulation, Iliev, Mellema, Pen 2005. 1o FOV

Cosmic Reioniation

Largest radiative transfer cosmological reionization simulations: 1 degree FOV.

Detection in 21cm hyperfine transition with radio telescopes. Structure on large scales (>20’).

Iliev, Mellema, Pen 2006

Foreground: Galactic Synchrotron

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

Detectability

• Luminosity proportional to object volume: bigger structures easier to find

• Noise dominated by galaxy: T=300(ν/150 Mhz)-2.5, higher frequency (lower redshift) much easier

• Mean emission challenging to discern (Gnedin and Shaver 2004).

• First targets: Stromgren spheres around bright quasars (Wyithe and Loeb 2004).

First Light Experiments

• Existing w/prelim data: PAST/21CMA (China), GMRT (India)

• Under construction: LOFAR (Netherlands), MWA (Australia), T-rex (Canada), CorE (Australia), VLA-VHF (USA)

• Future: SKA, JWST

LOFAR

Mileura Wide-angle

Array

Photos: Brian Corey and Eric Kratzenberg, MIT Haystack Observatory

Indian Giant Meterwave Radio Telescope

30 dishes @45m ea.

Operates in 2m band

Collaborators: Y. Gupta (chief scientist), Rajaram Nityananda (director), R. Subramanian, S. Sethi, A. Roshi (Raman), C. Hirata (IAS), T. Chang (UCB)

GMRT Search

• Operating telescope, 50000m2, up to 32 Mhz BW, fully polarized, lowest band 100-200 Mhz. Biggest collecting area in this band.

• Half in central 1km core, rest in 50 km Y.• Currently hits RFI limit after a few minutes of

integration: power lines, TV stations, HAM amateurs, faulty home electronics.

• Exploration of new RFI mitigation schemes, software correlator, nearfield clean.

• Search for SDSS QSO Stromgren spheres in TV band: 6 antenna filters already replaced.

GMRT 150 Mhz image

By Ishwara Chandra (NCRA)

Image noise 2 mJy

Thermal limit 0.5 mJy

Software correlator: $300/node

Power Line Noise

Time-lag: Folded 70ms data on short baseline

CMB

Comoving distances

reionization

Low l anomaly: model primary CMB to l~20

EoR

ISW: lensing map predictions

T-E correlations: can be predicted!

Pen 2003

Potential Theoretical Benefits

• Precision measurement of power spectrum at 10-8 accuracy (beyond PAST)

• Dark energy dynamics: q0, a(t), ISW, dark matter dynamics/clustering (through lensing), gravity waves.

• Initial conditions: 2nd order inflation effects, backreaction, curvature, etc. (through hydrogen matter P(k) and 3pcf).

Outlook

• Existing constraints: optical depth from WMAP, SDSS QSO’s.

• Theoretical progress: direct simulations indicate power on larger scales (>20’), making detections more tractable.

• current: initial data from Past/21CMA and GMRT, data analysis, ionospheric solution (Hirata).

• Several other experiments developing: LOFAR, VLA/VHF• Bright outlook: several experiments underway or planned

to tap the next cosmic horizon• Exciting new window on universe for precision cosmology.

Open field for theory and experiment.