How Very Massive Metal Free Stars Start Cosmological Reionization

 

John H. Wise – NASA / GSFC; Stanford University / KIPAC

Tom Abel – Stanford University / KIPAC

Abstract

The initial conditions and relevant physics for the formation of the earliest galaxies are well specified in the concordance cosmology. Using ab initio cosmological Eulerian adaptive mesh refinement radiation hydrodynamical calculations, we discuss how very massive stars start the process of cosmological reionization. The models include non-equilibrium primordial gas chemistry and cooling processes and accurate radiation transport in the Case B approximation using adaptively ray traced photon packages, retaining the time derivative in the transport equation. Supernova feedback is modeled by thermal explosions triggered at parsec scales. All calculations resolve the local Jeans length by at least 16 grid cells at all times and as such cover a spatial dynamic range of ~106. These first sources of reionization are highly intermittent and anisotropic and first photoionize the small scales voids surrounding the halos they form in, rather than the dense filaments they are embedded in. As the merging objects form larger, dwarf sized galaxies, the escape fraction of UV radiation decreases and the HII regions only break out on some sides of the galaxies making them even more anisotropic.  In three cases, SN blast waves induce star formation in overdense regions that were formed earlier from ionization front instabilities.  These stars form tens of parsecs away from the center of their parent DM halo. Approximately 5 ionizing photons are needed per sustained ionization when star formation in 106 Msun halos are dominant in the calculation.  As the halos become larger than ~107 Msun, the ionizing photon escape fraction decreases, which in turn increases the number of photons per ionization to 15-50, in calculations with stellar feedback only.  Supernova feedback in these more massive halos creates a more diffuse medium, allowing the stellar radiation to escape more easily and maintaining the ratio of 5 ionizing photons per sustained ionization.

Movies

High Resolution Images

Projections:

These movies follow the evolution of the refined region where star formation occurs.  The field of view is 125 and 150 comoving kpc (1/8 and 1/10 of the simulation volume) for SimA and SimB, respectively.

Density:

    Star Formation Only (SimA, SimB)

    +Supernova Feedback (SimB)

Temperature:

    Star Formation Only (SimA, SimB)

    +Supernova Feedback (SimB)

Contact: John Wise (john {dot} h {dot} wise {at} nasa {dot} gov

Figure 4: Radial profiles of gas density and temperature around 12 selected star formation regions.  The solid lines depict the dark matter density in units of hydrogen mass per cm-3.  Halos 1 and 2 host the first star in the simulation.  Halos 3, 4, and 5 form a star after matter is reincorporated into the halo after the first episode of star formation.  Halos 6, 7, and 8 have had their star formation delayed by H2 dissociating radiation.  Halo 9 is a case where star formation was induced by a supernova blastwave.  Halos 10, 11, and 12 have a virial temperature of 104 K.  (view Halos 1-6, Halos 7-12)

Figures 5 and 6: Density-squared weighted projections of gas density and temperature of the inner 150 comoving kpc at redshift 17, comparing cases with an adiabatic equation of state, atomic hydrogen and helium cooling, star formation and feedback, and plus supernova explosions.

(view SimA, SimB)

Figure 15: Two-dimensional slice of gas density and temperature, illustrating the anisotropic nature of HII regions in more massive (>107 Msun) halos.  The field of view is 900 physical parsecs.

(view)