Research

What I Study

Most generally, I study theoretical astrophysics. More specifically, I study the endstates of stars, primarily in the form of accreting white dwafs and supernovae, both of which you can learn about in the tabs above.

Who I Work With

My current supervisor is professor Frank Timmes of the School of Earth and Space Exploration at Arizona State University. I maintain ties to my doctoral adviser, Lars Bildsten of the Kavli Institute for Theoretical Physics (KITP), where he is the director, and the Physics Department at UCSB , where he is a professor. There are other collaborators that I need to add here...

Tools I Use

I am primarily a theorist, so my main tools are computational. I use the open-source stellar evolution code, MESA which was largely developed here at UCSB, specifically by Bill Paxton of the KITP, though the community is continually growing as more and more astronomers help to develop the code. I use MESA to model various stellar configurations to try to explain and predict observations in the night sky.

In the process of working with simulation data from MESA, I use many other computational tools to aid me, including several I have written myself that you can learn about here. Other useful tools I use are the python tools Numpy , Scipy , and matplotlib for numerical and plotting tasks. For more astro-specific needs, I use the wonderful astropy module. For other projects I have used the radiative transfer code Cloudy, which is another open source project for astronomers, and I also still use Bill Paxton and Vincent Fourmond's lovely plotting tool Tioga.

Accreting White Dwarfs

My primary research while at UCSB has been on the various ways white dwarf stars react when they accrete matter from a companion. A large fraction (roughly one third) of stars form in binary systems. As a result, a star's binary companion can affect its evolution.

When a star that's not too massive (perhaps up to several times the mass of the sun) runs out of nuclear fuel, it sheds its outer layers, leaving behind a very hot, earth-sized object called a white dwarf that is of such tremendous density that the primary pressure preventing it from further collapse is that from electron's wave functions not wanting to overlap.

Now if a white dwarf has a binary companion, it is possible that due to the winds or just a spillover of material from expansion as it evolves, the compaion can donate matter onto the surface of the white dwarf, either by direct accretion or more likely through an accretion disk. What happens to that matter once it makes its way to the surface of the white dwarf depends on the mass of the white dwarf, the composition of the material, and the temperature of the white dwarf. These outcomes are the primary focus of my research.

Outcome 1: Steady Burner

For a given mass white dwarf accreting solar-like material from a compaion, there is a narrow range of accretion rates that allow for the material to be burned at the same rate it is being accreted. In such a configuration, The matter can be thought to arrive at the surface of the white dwarf, pass through a thin (~10-7 M) hydrogen-rich layer before reaching the burning layer, where it is processed into nearly pure helium at a temperature of around 108 K. These steady-state burners are still compact objects, but are prodigiously luminous, burning at nearly 10,000 times the luminosity of the sun. As such, they are remarkably hot, with effective temperatures ranging from 100,000 K to 1,000,000 K. Being so hot, they emit strongly in the ultraviolet and soft X-ray. This, along with their original mode of discovery gives them their more common name, supersoft sources .

In my work, I have used the computational tool MESA to find the accretion rates that allow for this outcome as well as the resulting structure of thse objects. I continue to study them to understand what the fate of the growing helium layer under the burning layer is as well as how that may affect the long-term mass growth or decay of the objects, as they are speculated to be one channel towards producing a Type Ia supernova .

Outcome 2: Nova

If the accreting matter is too deficient in hydrogen or it is accreting too slowly, it arrives on the white dwarf cold and instead accumulates with little to no hydrogen burning. After a sufficiently large envelope has accumulated, the pressure at the base of the freshly accreted layer becomes high enough to initiate unstable burning that is much more vigorous than that in the steady burning case. This rapid injection of energy into the thin outer envelope drives a convective wave through the accumulated layer, which in turn drives the radius of the white dwarf to expand dramatically, cooling the outer layers.

In this now engorged and highly luminous state, the white dwarf becomes about as luminous as the steady state burner, but since it is so much cooler, we can observe it easily with optical telescopes both in the Milky Way and in nearby galaxies like M31. This state is a nova outburst, where the expanded white dwarf is bright in the optical and is typically observed to drive a wind, blowing off much of the material it had built up before it can be burned. After the envelope mass has been reduced by mass loss to a sufficiently small mass, it shrinks back down to a more supersoft source-like state, where it burns its remnant envelope quasi-stably for days to years before finally extinguishing and returning to its accreting state.

My research concerning the nova outcome has focused on finding what recurrence times (times between nova outbursts) arise from various combinations of white dwarf mass and accretion rate of solar-like material. This is important since measuring a recurrence time of tens of years is typically assumed to indicate a massive white dwarf and a possible progenitor system for a type Ia supernova or accretion-induced collapse.

Additionally I've done extensive work characterizing the supersoft phase of the nova after outburst, attempting to use measurements of this phase to constrain the progenitor system. Most recently, collaborators working with the Intermediate Palomar Transient Factory discovered and analyzed the quickest-recurring nova known, which goes off in M31 about once every year.

Most recently I've been working with collaborators to better understand how the supersoft phase after outburst could affect the radio observations of novae long after their optical peak. This work is ongoing.

Relevant Publications

Superluminous Supernovae

Supernovae are among the most luminous energy sources in the known universe, often outshining their host galaxies for days when they go off. They are also believed to be the primary generators of metals heavier than carbon in the universe, and type Ia supernovae can even be used as cosmic distance indicators, leading us to the discovery of dark energy.

Recently, though new classes of superluminous supernovae have emerged, challenging our current understanding of how stars die. One of my current projects is to search for these superluminous supernovae in the dataset from the Supernova Legacy Survey (SNLS).

To determine if a supernova can be deemed "superluminous", we need to have measured how bright we observe it to be as well as how far away it is. After all, two flashlights of the sam intrinsic brightness will appear to differ if viewed from different distances. To do that we use cosmic redshift, as first made famous by Edwin Hubble. The supernovae the SNLS observed are typically far enough away that cosmic expansion is moving them away so fast that their emergent spectrum is reddened significantly, like how a receding train whistle sounds lower in pitch than it "actually" is.

Usually finding a redshift (and thus getting a distance) is done by measuring a spectrum of an event. The SNLS was primarily interested in finding type Ia supernova for their cosmological purposes, so most events that would later be found to be superlumious super novae do not have spectra on record. In this case, we rely on the spectrum of their host galaxy to infer a redshift, and thus a distance.

If we have a useful redshift, we then model the supernova as an expanding fireball whose radius and temperature evolve with time according to $$R(t) = R_0 + v_{\mathrm{exp}}t$$ $$T(t) = T_0 + \frac{dT}{dt}t$$ so that the radius is expanding linearly with time and the temperature is assumed to decline linearly with time (dT/dt < 0). While this isn't a very robust model, it is often good enough to model the event at its peak brightness.

With a full model light curve for the event, we can determine if it is superluminous or otherwise interesting. This work is ongoing.

Relevant Publications