1. First stars and Near Infrared Extragalactic Background Light Sapporo, March 1, 2005 T. Matsumoto (ISAS/JAXA) 1. Impact of WMAP 2. First stars (pop.III) ? 3. Near Infrared Extragalactic Light(NIR EBL) 4. Future observations

2. Recent topics on Cosmology WMAP(Wilkinson Microwave Anisotropy Probe ） Launched on June, 2001 Orbit S-E L2 Frequency 23,33,41,61,94 GHz Polarization can be observed Angular resolution 0.2 degree （ cf. COBE:7 degree ） First data was opened on Feb.2003

3. Fluctuation of CMB observed by WMAP Consistent with COBE Finer structure is detected Power spectrum of CMB fluctuation Positions, heights of peaks provide cosmological parameters Geometry of the Universe Life of the Universe Baryon, dark matter, dark energy Hubble constant

4. Douspis et al.  ~0.17 +/- 0.04 z_rei ~ 20 +/- 8 -----------

5. Summary of WMAP results suggesting inflation universe ● Flat universe Ω=1.04 ± 0.04 ● Life of the Universe134±3 10 8 yr ● Baryon densityΩ m =0.046 ± 0..02 ● Dark matter densityΩ dm =0.23 ± 0.04 ● Hubble constant h=0.72 ± 0.05 ● Dark energy Ω Λ =0.71 ± 0.07 ● Optical depth for CMB τ =0.17 ± 0.04 Constraint on the re-ionization epoch: z 〜 17

6. Gunn-Peterson troughs -> Absorption by neutral Hydrogen Reionization of the Universe

7. From Gunn-Peterson troughs in Sloan quasars: 1. Small neutral fractions at z ~ 6 (1% neutral) 2. Sharp transition at z~6 (end of reionization?) Fan et al. 2001

8. Reionization epoch is earlier than previously thought What caused reionization? Super novae AGN mini black holes First stars (Pop.III stars) Integrated light of first stars can be observed as near infrared background!

9. First stars (pop.III stars)? After the recombination era, universe was neutralized No metal, H and He only Cooling through hydrogen molecules ⇨ massive star formation Luminosity (Eddington limit): 4  G mc 2 M/  T ~1.3x10 38 M/M ○ erg/sec Temperature T~10 4.8-5 K Life time t L ~  Mc 2 /L~3x10 6 yr (  ~0.007) Final stage of the evolution M< 40M ○ type II super nova 40M ○ <M< 130M ○ black hole 130M ○ <M<260M ○ pair instability super nova M>260M ○ black hole

10. Can we detect the signature of first stars directly? A 300 M ○ first star at z~15, K-band mag 33 (unlikey to be detectable) First proto-galaxies can contain as many as 10 5 stars. (Still not detectable) (Scherrer 2002; Bromm et al. 2002; Santos et al. 2001) Interesting wavelength range is 1 to 3 microns!!

11. Infrared Extragalactic Background Light (IREBL) Cosmic Infrared Background (CIB) integrated light of distant galaxies and stars UV and optical radiation can be observed at near Infrared wavelengths due to redshift A key observation to delineate the dark age of the Universe Complementary to galaxy deep survey Space observation is inevitable! Several rocket flights COBE/DIRBE IRTS/NIRS

12. COBE(COsmic Background Explorer) FIRAS DMR DIRBE(Diffuse Infrared Background Experiment) Absolute photometry of the sky brightness at 1.25, 2.2, 3.5, 4.9, 12, 25, 60, 100, 140, 240  m beam size ~0.7 degree COBE was launched on 1989 and attained all sky survey. As for the CIB,COBR team reported detections at far infrared bands upper limits for other bands Several authors obtained significant detections at J, K, L bands using COBE data

13. IRTS(Infrared Telescope in Space) NIRS(Near Infrared Spectrometer) One of 4 focal plane instruments of IRTS wavelength coverage 1.4-4.0  m spectral resolution 0.13  m beam size 8 arcmin. x 8 arcmin. Compared with COBE/DIRBE smaller beam capability of the spectroscopy smaller spatial coverage ~7% of the sky One of mission instruments of small space platform, SFU launched on March 15, 1995 15cm cold telescope Optimized for diffuse Extended sources Mission life ~ 1 month

14. Near Infrared Sky Foreground emission sources zodiacal light scattered sunlight by interplanetary dust (IPD) zodiacal emission thermal emission from IPD >3.5  m Milky Way, integrated star light ・ It is important to resolve and remove as faint stars as possible. ・ Smaller beam is better to avoid confusion IRTS/NIRS: 8 arcmin COBE/DIRBE: 0.7 degree Subtraction of foreground emission Is a critical issue to detect EBL

15. IRTS observations 7% of the sky was surveyed during IRTS observation period (4 weeks) The data for 5 days before liq. He ran out were used to avoid contamination The data at high galactic latitudes are sampled 40<b<58 degree, 10<  <70 degree spectra of 1010 blank skies where no stars are detected effective beam size is 8 ’ x20 ’ due to scanning effect

16. Integrated light of faint stars Constructed logN/logS model based on the NIRS observation (M.Cohen) Obtained magnitudes of stars that correspond to the noises → cut off magnitudes for all wavelength bands cf. 10.4 mag. at 2.24  m Calculated integrated light of stars fainter than cut off magnitudes for b=42, b=45, b=48 and applied cosec(b) law The result is consistent with 2MASS For the H and K bands

17. Zodiacal light and emission Apply physical model by Kelsall et al. (ApJ, 508, 44 1998) to NIRS bands. Model is based on the annual variation of the zodiacal light observed by DIRBE/COBE. Calculate the brightness of zodiacal light/emission for all NIRS bands and observed points.

18. After subtracting the star light and zodiacal light/emission Significant isotropic emission was detected for all bands !

19. Residual emission shows no dependence on the galactic plane

20. Breakdown to emission components Observed sky brightness at high ecliptic latitude Zodiacal light/emission Isotropic emission ~20 % of dark sky Integrated light of faint stars

21. COBE/DIRBE and star counts Comparison with other observation J-band K-bandL-band Dwek & Arendt (1998) 9.9 ±2.9 Gorjian et al. (2000) 22.4±611.0 ±3.3 30.7±6 15.4 ±3.3 Wright and Reese (2000) 23.1±5.916.8 ± 3.2 31.4±5.9 Wright (2001) 28.9 ±16.3 20.2±6.3 61.9 ±16.3 28.5 ±6.3 Cambresy 54 ±16 27 ±6.7 IRTS/NIRS 27±5 （ 2.24  m) In unit of nW.m -2.sr -1 Red numbers are based on "very strong no-zodi principle" (VSNZP) All observations are consistent if same zodi model is used!

22. Spectrum of the observed isotropic emission Stellar like spectrum was found. Main error is uncertainty of the zodiacal light model Consistent with COBE/DIRBE Significantly brighter than the integrated light of galaxies ! Spectral gap around 1  m In-band energy flux is ~ 35 nW.m -2.sr -1

23. Spectrum of excess emission over ILG can be explained well by integrated light of first stars! Sarvaterra and Ferrara (MN 339, 973 (2003) z end ~8.8, redshifted Ly  J band f ★ =10 〜 50 ％ massive star formation -> produced metals were confined in black holes z=17 2.2x10 8 yr z=8.8 5.5x10 8 yr

24. Another evidence of NIR EBL Inverse process of pair anihiration  ~TeV) +  ~eV  -> e + + e - when E  >(mc 2 ) 1/2 Cross section is maximized when the soft phton energy is e~2(mc 2 ) 2 /E=0.5(1 TeV/E) eV ~2  m Absorption of TeV-  blazer!

25. BL Lac object H1426+428 z=0.129 ■ CAT(1998-2000) ▲ Whipple(2001) ● HEGRA(2002) ○ HEGRA(2002) lines: model by Mapelli and Ferrara

26. Fluctuation of the sky -1 rms fluctuation Stellar fluctuation is estimated by using the model, but consistent with 2MASS Fluctuation of zodiacal emission at 12  m is less than 1% (IRAS, COBE, ISO)! ⇨ Zodiacal light can not explain observed sky fluctuation!

27. Fluctuation of the sky -2 Correlation between wavelength bands Clear correlation between wavelength bands was detected. Spectrum (color) of fluctuation component is similar to that of isotropic emission ⇨ Isotropic emission is spatially fluctuating

28. Spectrum of fluctuation ⇨ Excess emission is fluctuating keeping the similar spectrum! Observed rms fluctuation: ~5% of the sky brightness, ~6% of the zodiacal light, ~20% of the isotropic emission Nearest pop.III stars (z~8.8) are responsible for the fluctuation !

29. What causes NIREBL fluctuation? 1.Stellar fluctuation? Model is fairly consistent with 2MASS data! 2.Zodiacal light and/or emission? IRAS, COBE, ISO 3.Faint galaxies? 4. Pop.III stars?

30. Zodiacal emission is very isotropic! IRAS:0.5 degree beam at 15 and 25  m COBE:0.7 degree beam at 12, 25 and 60  m Residual from smooth distribution is less than 1% of peak brightness! ISO: 3’x3’ pixel, 45’x45’ frame 5 fields at different  were observed at 25  m rms fluctuation in one field is ± 0.2% ! Observed fluctuation is 6% of ZL It is unlikely there exists big difference between scattering and emission

31. Observations (DIRBE, IRTS/NIRS and NITE) and theory (Cooray et al. 2004)

32. z-dependence of the model brightness (Salvaterra and Ferrara 2004) Can pop.III explain observed Fluctuation?

33. 2-point correlation function Observed sky Analysis is made for wide band brightness (integrated brightness for 1.43-2.14  m) Read out noise is negligible Fluctuation is celestial origin Data points lie along the belt ↓ One dimensional analysis

34. Power spectrum Specific feature at 1 〜 2 deg. This scale is, 20 Mpc at z=8.8 200 Mps at present First peak of CMB (l~220, 0.8 deg) corresponds to 1.45 deg. at z~8.8

35. Power spectrum for subsections

36. Expected fluctuation and detection capability of IRC/ASTRO-F (Cooray et al. 2004, Ap.J, 606, 611) ⌒ Theory: Based on the fluctuation of dark matter. Observation: Much larger fluctuation Sharp peak at 2 deg. Radiation of pop.III stars do not follow dark matter? Underlying fluctuation may exists. Future observations: Subtraction of foreground galaxies is essentail. ASTRO-F is powerful

37. Theoretical estimation of fluctuation Kashlinsky et al. 2004

38. Future observations Issues to be observed Spectral shape Confirmation of the spectral gap at ~1  m real? Other spectral features? Fluctuation Spatial correlation over the wide range of angular scale Confirmation of 2 degree feature in 2 dimensional image Observe underlying large scale structure Absolute measurements Observation free from ambiguity of the model ZL ASTRO-F: image at K and L CIBER (Rocket experiment): spectral observation, image at I and H Out of zodiacal cloud mission: zodi free observation

39. ASTRO-F Formation and evolution of galaxies, stars, and planets First dedicated infrared mission of ISAS 70cm cooled infrared telescope Advanced Infrared Survey 50 times higher sensitivity, 10 times better spatial resolution, has longer wavelength band, than IRAS Instruments IRC(Infrared Camera) 512x412 InSb array camera, 1.5”/pixel band imaging: K, L, and M bands low resolution spectroscopy: R-30 slit 2x50 pixel, R-15 4x50 pixel 256x256 SiAs array FIS(Far Infrared Surveyor) Launch target : January, 2006 Orbit : sun synchronous orbit, 750km altitude Mission life: ~1.5 year (liq. He holding time) + 2 years (dedicated to NIR Observations)

40. Observation of NIREBL with ASTRO-F Advantages of IRC/ASTRO-F observation Point-source rejection by high-resolution imaging observation Limiting magnitude at the K band is ~20 mag. for one pointing observation (~10 min.) This corresponds to ~30 nW m -2 sr -1 for 1 pixel (5  ) Almost all galactic stars and faint galaxies can be identified Discrimination of the fluctuation of the zodiacal light Observation of the same field at the different time epoch Spitzer does not have K band Observation plans 1. Detection of the NIREBLfluctuation over the wide range of angular scale Wide area survey towards north ecliptic pole (NEP) is being proposed. Coordination with galaxy deep survey group 2. Detailed study of the spectrum of IREBL Low resolution spectroscopy at different ecliptic latitudes (2~5  m) Spectrum without contamination of stars and galaxies can be obtained

41. “NEP-Deep & Wide” : Summary NEP-Deep Field, 50 pointing/FOV 0.5 deg 2 2.8 deg NEP-Wide Field, 4 pointing/ FOV Area: 2.8 deg 2 N2N3 or N4 22 S7S11 22 L15L24 22 Revised on 28 th Oct. 2004

42. ASTRO-F detection limit Wide(1pixel) 5  Deep(1pixel) 5  Wide(100pixels) 5  IRC imaging observations at NEP are enough sensitive to detect the CNIRB fluctuation seen by IRTS Spectroscopic measurement of the CNIRB mean level avoiding the contamination by normal galaxies Spectroscopy (100pixels x 10sky) 5  K >20mag Integrated flux of galaxies

43. Expected fluctuation and detection capability of IRC/ASTRO-F (Cooray et al., submitted to Ap.J.)

44. CIBER: Cosmic Background Explorer Objectives Sounding rocket observations at the wavelengths below the K band! NASA’s Black Brandt rocket 1. Spectrometer: Confirmation of spectral gap at ~1  m low resolution spectroscopy, 0.8  m< <2.0  m 2. Imager: Observation of sky fluctuations at the I and H bands 2-dimension analysis

45. Instrumentation of CIBER Spectrometer 7.3 cm dia. Telescope 1 arcmin./pixel, 4 degree frame low resolution spectroscopy 0.8-2.0  m, R~10 Imager 15 cm dia. Telescope x2 10 arcsec/pixel, 2.8 degree frame I and H

46. Optical design of the spectrometer

47. Imager Optics

48. Detection limit of the spectrometer

49. Simulated spectrum of the sky

50. Expected performance of the imager Spatial power spectrum of Pop III fluctuations (red curves), local galaxy fluctuations (correlations term light blue curves, shot term dashed curves) for 3 different cutoff magnitudes, and the total signal (solid blue curves). The 18.5 mag cutoff is for the rejection level from the NAME images alone; the faintest cutoff (I = 25.5 and H = 21) comes from ground-based measurements overlapping our images. The data points show the errors from NAME in a 100 s observation, including both instrument noise and sample variance. We assume there are no Pop III fluctuations detectable at I-band, following the IRB star spectrum in Fig. 3. NAME can easily detect the optimistic Pop III signal (this model produces a cumulative background of 25 nW m -2 sr -1, consistent with the missing amount in Figs. 1 and 3), clearly distinguished by its different power spectrum from local galaxies at H-band. NAME has sufficient sensitivity to detect the pessimistic Pop III signal (this model produces a background of 3 nW m -2 sr -1 ), although it is obscured by local galaxy fluctuations at a limiting magnitude of H = 21. Reducing the cut-off magnitude further is possible, and would allow us to positively extract even the signal of the pessimistic model.

51. Configuration of the telescope system

52. Payload configuration

53. Observation plan Fig. 17. Proposed sequence of observations superposed on the trajectory of the NITE payload. Separation from the rocket engine occurs at 85 s, followed by despin, opening of the vacuum shutter door at 95 s, and slewing the payload to the first science target. The instrument observes 5 science fields before closing the shutter door, reentry into the atmosphere, and recovery operations

54. Organization and schedule Organization Japan: ISAS(Matsumoto, Matsuura, Wada, Matsuhara) Nagoya U. (Kawada, Watabe) US: Caltech/JPL(J.Bock) UCSD (B.Keating) Korea: KAO (S. Pak, D-H. Lee) Schedule June, 2004Proposal to NASA Now approved! No-funded launch! Spring, 2007First launch at White Sands Spring, 2008Second launch Funding?

55. The life of an IR rocket (Jamie’s previous experiment)

56. Solar sail mission Out of zodiacal cloud mission ● Free from ZL and IPD emission ● Accurate absolute measurement of EBL without IPD model ambiguity is possible ● Observation of the mid-infrared background is possible

57. Free from zodiacal light/emission provides decisive result for the NIREBL!

58. Possible mission concept of out of zodiacal cloud mission! Scientific objectives Accurate measurement of spectrum and fluctuation of IREBL Instrumentation Telescope5cm dia. lens system Wavelength range0.8-2.2  m Pixel FOV~10 ’ DetectorHgCdTe Cooling systemradiation cooling Weight3 kg

59. Summary 1.CMB polarization observed by WMAP indicates that the Universe was reionized at z~17 by the first massive stars (pop.III stars). 2.Independent observations by COBE and IRTS provide detections of significant near infrared extragalactic background light. Recent observations of Tev-  Blazers support its cosmological origin. 3.The near infrared extragalactic background observed by IRTS and COBE could be consistent with pop.III star scenario. 4.Spectrum observed by IRTS suggests the redshift at the end of pop.III era is ~9. 5.Fluctuation of the sky was detected (~20% of EBL) by IRTS and COBE which is too large to be explained with the standard model. 6. Near infrared background is a unique tool to investigate the pop.III stars. ASTRO-F, CIBER(Rocket experiment) and Solar-Sail missions will provide valuable information on the pop.III era.

60. NIREBL is a unique tool to investigate the first stars! CMB z=1,0003x10 5 year ? Near infrared background z~105x10 8 year

62. Cosmic Microwave Background (CMB) Most distant observable object The Universe ~4x10 5 years after big bang Fossil photons COBE(COsmic Background Explorer) CMB Map (launched on 1989 by NASA) CMB is very uniform But Fluctuation of ~10 -5 is detected ⇨

63. Present Universe Extremely non uniform! Large scale structure, Cluster of galaxies, galaxies, stars planets, ------- Evolution from uniform and isotropic Universe to extremely non uniform Universe? How first stars and galaxies formed?

64. Evolution of the Universe Dark age of the Universe

67. Prepared by N. Fujishiro Proposed Survey Field

68. IRC background measurements around NEP Spectral resolution /  Survey area [sq.degree] Exposure time per frame [# of pointings] Single pixel detection limit (5  ) * [nW/m 2 /sr] Number of galaxies per camera frame** Number of dark pixels per frame *** Ultra-Wide (Phase-3) 3100TBD~ 10 (pixel binning) -- Wide32.82 (500 s)10-- Deep30.525 (1.4 hrs)33000>10^5 Ultra-Deep3----- * in unit of surface brightness ( I  ** FOV of the IRC camera frame is 10’x10’ *** number of pixels available for the background analysis = total number of pixels – (confusion factor) x (number of galaxies) = 1.7x10^5 – (3pics x 3pics) x (number of galaxies) Spectral resolution /  Survey areaExposure time Detection limit (5  ) [nW/m 2 /sr] Number of galaxies Number of dark pixels Spectroscopy30 (  2 – 5  m) 3 ” x 73 ” x 100 directions (various b and  ) 1 pointing (500 s) x 100 directions 30 (pixel binning) ~10 (10 sky average) 290 1. Wide-band deep imaging in K, L and / or M bands 2. Spectroscopy

70. Lensed galaxy at z~10? Pello et al. 2004, A&A 416, L35-L40.

71. SED and spectrum

72. IRC SURVEY STRATEGIES 0.02 1 10 100 1000 ? Area (sq. deg.) Number of Pointings 100 10 1 Depth and Area

73. Theory to Reality: Near-IR wide-field surveys 5-sigma point source detection (planned) (experiments at this end are preferred)

75. IRC background measurements around NEP 1. Wide-band deep imaging in K, L and M bands * in unit of surface brightness ( I  ** FOV of the IRC camera frame is 10’x10’ *** number of pixels available for the background analysis = total number of pixels – (confusion factor) x (number of galaxies) = 1.7x10^5 – (3pics x 3pics) x (number of galaxies) Spectral resolution /  Survey areaExposure time Detection limit (5  ) [nW/cm 2 /sr] Number of galaxies Number of dark pixels Spectroscopy30 (  2.0 – 5  m) 3” x 73” x 100 directions (various b and  ) 1 pointing (500 s) x 100 directions 30 (pixel binning) ~10 (10 sky average) 290 2. Spectroscopy Spectral resolution /  Survey area [sq.degree] Exposure time per frame [# of pointings] Single pixel detection limit (5  ) * [nW/cm 2 /sr] Number of galaxies per camera frame** Number of dark pixels per frame *** Wide-field Shallow （ phase-3) 31001 (500 s)302x10^3>1.5x10^5 Shallow (Phase-1,2)3101 (500 s)302x10^3>1.5x10^5 Deep (Phase1,-2)3110 (1.4 hrs)10(3-4)x10^3>1.3x10^5 Ultra Deep (Phase-1, 2) 30.02100 (14 hrs) 3(0.5-1)x10^4>8x10^4

76. Fluctuation of the sky-4 Detection of fluctuation with 2MASS data Kashlinsky et al. ApJ 279, L53 (2002), Odenwald et al. ApJ 283, 535 (2003)

77. Interpretation of the 2MASS fluctuation with pop.III stars

78. Theoretical estimation of fluctuation I. Magliocchetti, Salvaterra and Ferrara MN, 342, L25 (2003) Sharp drop at ~200 arcsec 8.6 Mpc at z end =8.8 Fluctuation is dominant at the J band