Research

 

Interests: Reionization, Early Universe, Cosmology, Meaning of Life, etc.

  1. First Light
    The dawn of the first astrophysical structures: What were they? In what numbers? When? How did they interact with their surroundings: feedback processes (radiative, chemical, mechanical)? How can we observe this epoch?  How can we learn about cosmology and astrophysics from the redshifted 21cm line?

  2. Reionization
    How can the reionization history be probed? How did it proceed? Who was responsible? What we learn about the underlying structures?

  3. The Nature of Dark Matter
    Is Dark Matter «Warm»? What are its properties? How can we constrain the properties of Dark Matter using high-redshift observations?

  4. High-Redshift Sources
    What were the properties and abundances of quasars and galaxies?

  5. Modeling Techniques
    How do we develop analytic, numeric and semi-numeric modeling tools and match the most efficient and relevant of these to each scientific inquiry?

collaborating with Luka (2005)

collaborating with Nikola (2008)

collaborating with La Chouffe

Here are some research snippets chosen mainly on the basis of having nice pictures.
For more on the cosmic 21-cm signal, click on the other tabs above.  The latest 21-cm media will be on the
EOS page.


Heating during the Cosmic Dawn:  The light emerging from the first galaxies signaled the end of the Dark Ages, and the beginning of the Cosmic Dawn and Reionization, the last major phase change of our Universe.  Observations of the CMB suggest that the mid-point of reionization occurred around z ~ 10.  Empirical scaling relations based on local, star-forming galaxies imply that before this, the intergalactic medium (IGM) was heated to temperatures above the CMB by X-rays from the first galaxies (z~15).  These X-rays are thought to originate from High Mass X-ray Binaries, though our recent work also highlighted the importance of the hot interstellar medium as a source of X-ray photons.

    These exciting epochs are about to be observed in 21cm: the spin-flip transition of neutral hydrogen.  The 21cm signal is sensitive to both the ionization and the thermal history of the cosmic gas, making it an ideal probe of the Cosmic Dawn and Reionization.  It encodes the multi-wavelength (X-rays, soft UV, ionizing) imprint of the first galaxies; the challenge becomes how to robustly interpret these exciting upcoming observations.

    Astrophysical parameter studies are becoming invaluable in this effort.  Although most of the focus is on reionization, the X-ray heating during the Cosmic Dawn can be detectable at comparable signal-to-noise, and can unambiguously tell us about the X-ray spectra and luminosities of the first galaxies.


    The figure above shows 21CMFAST simulated maps of the 21cm brightness temperature offset from the CMB. The horizontal axis shows evolution along the comoving line-of-sight coordinate, from z ~ 62 to z ~ 7. From right to left we see the expected major milestones in the signal: (i) collisional decoupling (red-->black); (ii) Wouthuysen-Field coupling (black-->yellow); (iii) IGM heating (yellow-->blue); (iv) reionization (blue-->black). The top panel corresponds to a “fiducial” model, while the lower panel corresponds to an “extreme X-ray” model in which primordial galaxies are much more luminous, with harder X-ray spectra, compared to local star-busts. The astrophysical milestones are very different in the two models. The white rectangle in the top panel approximately corresponds to the typical box size for state-of-the-art radiative transfer reionization simulations resolving the relevant source population at z <10; earlier epochs require even higher resolution as well as additional radiation fields.  For movies of these simulations, see http://homepage.sns.it/mesinger/21cm_Movie.html.
http://arxiv.org/abs/1403.6125
http://arxiv.org/abs/1310.0465
http://arxiv.org/abs/1310.7936
http://arxiv.org/abs/1210.7319


Sub-grid modeling of Reionization:  Sub-grid recipes are a necessity for modeling unresolved physics (<~ pc scales) in reionization simulations (having scales of ∼0.1-1 Gpc). We have been developing a more physics-rich version of our popular semi-numerical code which self-consistently tracks the sub-grid evolution of both sources (galaxies) and sinks (Lyman limit systems) of ionizing photons.  Both can have a substantial impact on the progress and morphology of reionization. Recent improvements include prescriptions for: (i) the depletion of the gas reservoir available for star-formation by an ionizing UV background (UVB); and (ii) inhomogeneous recombinations inside HII regions. Both effects are calibrated to numerical simulations, and depend on the properties and ionization history of each (~ 1 Mpc) simulation cell.

     The plot below shows a light-cone slice (with redshift on the horizontal axis) through the cumulative recombination field in a 21CMFAST realization which includes both of the feedback mechanisms mentioned above.  Recombinations (and to a lesser degree UVB feedback on star-formation) act to slow down the growth of large, early forming HII regions, allowing newly-formed HII regions to “catch-up” in size.  This results in a delay as well as a dramatic deficit of large HII regions and large-scale power (factors of >~2-3), which is not captured by state-of-the-art radiative transfer simulations of reionization, due to their dynamic range limitations.


http://arxiv.org/abs/1402.2298

http://arxiv.org/abs/1301.6781

http://arxiv.org/abs/1301.6776

http://arxiv.org/abs/1008.0003



The nature of Dark Matter from high-redshift observations:  Since structures form hierarchically, the early Universe was a much emptier and simpler place than our current one.  It serves as a powerful test bed for non-standard cosmologies.  For example, the mere presence of collapsed objects such as galaxies at high redshifts can provide robust constraints on cosmological models which include a suppression of small-scale power [such as popular Warm Dark Matter (WDM) models].  In this respect, Gamma-Ray Bursts (GRBs), having unambiguous detections out to very high redshifts, are particularly useful.   GRBs provide competitive and robust constraints on the WDM particle mass (e.g. m_x > 2 keV).  Moreover, the faint z~10 galaxies being detected by current and upcoming cluster lensing surveys can provide the only WDM limits without any astrophysical degeneracies.

    Upcoming 21cm observations serve as another probe of physical cosmology.  The 21cm signal during the Dark Ages (see the top figure above) is a clean probe of the matter distribution, on scales much smaller than the CMB.  More accessible with upcoming observations is the epoch of X-ray heating during the Cosmic Dawn.  Some popular dark matter models have annihilation signatures which can dominate over galactic X-rays during this period.

http://arxiv.org/abs/1310.0029
http://arxiv.org/abs/1209.2120
http://arxiv.org/abs/1306.0009
http://arxiv.org/abs/1303.5060
http://arxiv.org/abs/astro-ph/0501233


Radiative feedback in relic HII regions:  In the absence of efficient metal cooling, the first stars are thought to have been relatively massive and thus short-lived.  Moreover, star-formation is likely a very bursty process inside the tiny potential wells hosting these molecularly-cooled galaxies. Following their period of activity, the first galaxies could have left “fossil” or “relic” HII regions, impacting future generations of star formation in nearby, clustered halos.  Feedback in relic HII regions is complicated: positive feedback (promoting future star formation) can result from the gas’ shorter cooling time due to the enhanced H2 fraction, while negative feedback (impeding future star formation) can result from nearby halos being photo-evaporated.  We ran suites of simulation runs with the AMR hydrodynamical code ENZO, varying the strength of the transient ionizing background and a persistent H2 disassociating background, in an effort to gain physical insight into this early feedback mechanism. 

    The movie shows a slice through the temperature field, with the ionizing background being turned on from z=25 to z=24.6.  After the radiation turns off, one can see the resulting pressure shock moving outward and expelling gas from the filament.  However, at the halo center, gas can still cool through the enhanced H2 cooling channel.

http://lanl.arxiv.org/abs/astro-ph/0604148

http://lanl.arxiv.org/abs/0812.2479



Radiative feedback in the advanced stages of reionization:  The importance of radiative feedback is poorly known even in the later, “simpler” regime, when a persistent ionizing background (UVB) starts establishing itself.  As mentioned above, the UVB during reionization can suppress the gas content of low-mass galaxies, even those capable of efficient atomic cooling.  The main difficulty in modeling this negative radiative feedback is the large parameter space to consider.  Redshift, UVB strength, evolution, turn-on redshift, halo mass are all factors which impact the results.  Most estimates have been based on studies which looked at photo-evaporation at lower redshifts, in halos which have been exposed to a UVB their entire life (z_init~100).  More recent studies find that during reionization, more compact halo profiles, increased cooling efficiencies, and shorter exposure times to the UVB could lessen the importance of negative radiative feedback.  We quantify this using a tiered approach: using hydrodynamical simulations to calibrate very large scale, high resolution seminumerical simulations (DexM).  The hydrodynamical simulations of gas and dark matter allow us to model the complicated dynamical and cooling processes of individual halos, over large swaths of parameter space.  The cosmological simulations then serve to model the ionizing flux seen by halos in representative volumes of the universe, including such effects as source clustering and ionization topology. By combining these results, we find that under all reasonably conservative scenarios, UV feedback on atomically-cooled halos is not, by itself, strong enough to notably delay the bulk of reionization.  Our recent work has implemented these results in a sub-grid fashion in large-scale simulations of reionization (see above).

    The plot below shows slices through three of our ionization fields generated by DexM, which allows us to treat the free parameters independently and systematically. 



http://lanl.arxiv.org/abs/0806.3090

http://arxiv.org/abs/1301.6776

http://arxiv.org/abs/1301.6781



Was reionization complete by z ~ 5-6?  It is generally taken for granted that reionization has completed by z=6, due to the detection of flux in the Lyman alpha forest at redshifts z<6.  However, since reionization is expected to be highly inhomogeneous, much of the spectra pass through just the ionized component of the intergalactic medium (IGM) even for non-negligible values of the volume-weighted mean neutral hydrogen fraction, x_HI.  We studied the expected signature of an incomplete reionization at z ~ 5-6, using very large-scale (2 Gpc) seminumeric simulations.  We found that ruling out an incomplete reionization is difficult at these redshifts since: (1) quasars reside in biased regions of the ionization field, with fewer surrounding HI patches than implied by the global mean, x_HI; this bias extends tens of comoving megaparsecs for x_HI < 0.1; (2) absorption from the residual neutral hydrogen inside the ionized IGM generally dominates over the a
bsorption from the remaining HI regions, allowing them to effectively "hide" among the many dark spectral patches; (3) modeling the Lyman alpha forest and its redshift evolution even in just the ionized IGM is very difficult, and nearly impossible to do a priori.  There is currently no direct evidence that reionization was complete by z ~ 5--6.  We proposed using the fraction of pixels which are dark as a simple, model-independent upper limit on x_HI.  Alternately, the size distribution of regions with no detectable flux (dark gaps) can be used to place a more model dependent constraint. 
   The plot on the left is a slice through an ionization field at z=5 and x_HI=0.1. Ionized regions are shown in white. 
The slice is 2 Gpc on a side and 3.3 Mpc deep.

http://arxiv.org/abs/1101.3314

http://lanl.arxiv.org/abs/0910.4161



Lyman alpha emitters during reionization:  Clustering of Lyman alpha emitters (LAEs) can be a powerful probe of reionization.  This is because their visibility is modulated by the pattern of ionized bubbles. The neutral IGM from outside the ionized bubbles can imprint strong Lyman alpha damping wing absorption, potentially wiping out their Lyman alpha emission line. Large bubbles (which surround overdensities with many galaxies) have a small damping-wing optical depth since the neutral IGM is on average distant from the sources, so that most sources inside them will be visible. On the other hand, small bubbles (where galaxies are rare to begin with) will appear to be entirely empty.  Thus, during reionization the observed LAEs should appear more clustered than the underlying population.  We investigated the use of the “counts-in-cell” statistic for detecting this reionization-induced excess clustering, and found it to be a simple, powerful probe of reionization.  With this statistic, a fully ionized universe can be robustly distinguished from one with x_HI > 0.5 using a survey containing only ~ 20--100 galaxies.

    The figures below show the observable LAEs at various global neutral fractions.  The increased modulation by the ionization structure is clearly seen as the universe becomes more neutral, going from left to right.


increasing global neutral fraction 

http://arxiv.org/abs/0708.0006

http://arxiv.org/abs/0710.0371