Chelsea Harris

University of California, Berkeley

A good joke intrigues Chelsea Harris, and to her there’s none more engrossing than type Ia supernovae, the exploding stars that have become indispensable tools for astrophysicists.

Supernovae are hugely energetic – so bright observers can spot them even in distant galaxies. Type Ia supernovae (or SNe Ia, in astrophysics parlance) are especially interesting because they’re standardizable candles: Their intrinsic brightness is deduced from how long they shine and their color. By comparing how bright an observed type Ia supernova is to how bright they know it should be, scientists can calculate its distance from Earth.

Evidence indicates SNe Ia are the explosion of a carbon-oxygen white dwarf in orbit with another star. The white dwarf gains mass from the companion, becoming unstable and exploding. Little is known, however, about the companion star or about how mass moves between it and the dwarf.

“Type Ia supernovae are probably nature’s best joke, because we use them as extremely precise cosmological tools without knowing what they are, and then nature makes it almost impossible to find that out,” says Harris, a Department of Energy Computational Science Graduate Fellowship recipient at the University of California, Berkeley.

Harris, an astrophysics doctoral candidate, seeks answers with high-performance computing. She builds simulations of how SNe Ia interact with the circumstellar medium (CSM) – gas and dust surrounding a supernova. The calculations follow the shock that the collision creates and predict the resulting radiation signatures, from X-ray to radio wavelengths in the electromagnetic spectrum, that astronomers collect with instruments when observing the exploding stars.

Electrons moving at near-light speed produce the radiation as they move through powerful magnetic fields the supernova shock generates. The radiation’s electromagnetic signature is altered as it passes through the CSM.

Harris’s predicted signatures help narrow the CSM’s possible characteristics and, thus, may help understand what comprises the companion star. “I try to see what the radiation signatures are to then say, OK, when we see this at this strength, this is what it tells us about the gas, or when we don’t see this, this is what we’ve learned.” That helps researchers focus more complex SNe Ia simulations on the most likely scenarios.

In a recent paper published in The Astrophysical Journal, Harris and her doctoral advisors, Peter Nugent and Daniel Kasen, calculated predicted radiation signatures in the radio wavelength for a type Ia supernova interacting with a narrow, low-density circumstellar shell. The calculations showed that radiation signatures evolved similarly under varying CSM conditions, peaking at a characteristic time. That means astronomers can use observed radio light curves to understand specific CSM properties.

“The nice thing about those light curves is they peak at the time that the shock front runs over the edge of the CSM, and the shock just dies,” Harris says. The relation “between the time of that peak and the width of the shell also tells me something about the shock that’s independent of that specific radiation signature.” Simulations that produced no radiation signature were perhaps even more intriguing. “What does a non-detection really mean? If the interaction signature in the radio (wavelength) was really short and it just fell between the observations, then it’s not like it didn’t have (radiation) – just that you didn’t see it.”

Harris is using her algorithms to analyze surprising optical-wavelength observations another Berkeley researcher has gathered. “Based on even the gaps in their observations, I’m able to make some sort of constraint on when the (supernova) shock overtook” the CSM shell edge. Her research has run on several systems at the National Energy Research Scientific Computing Center, including its newest supercomputer, a Cray XC40 named Cori.

There’s more work to do. Harris admits she chose to tackle one of the CSM problem’s easiest parts: the low-density scenario. Now she must move on to the more difficult job of modeling SNe Ia interaction with high-density CSMs. With that, she also can explore interactions between CSMs and other supernova types, such as superluminous supernovae, rare exploding stars up to 100 times brighter than other kinds. “We have no idea what’s exploding. It has insane energies and there are theories all over the place.”

Harris expects to graduate in 2018 and plans to seek research fellowships so she can pursue these and other interesting and amusing questions.