During the tenure of his Hubble Fellowship, Lifan pioneered the new field of supernova spectropolarimetry. Polarization measures departures from spherical symmetry. For a spherically symmetric object, the orientation of the electric vectors of electromagnetic radiation will cancel out on average, giving no net polarization. If finite polarization is detected, then the radiating object cannot be spherically symmetric.

This was a bold project to undertake because the observations are challenging, the data analysis subtle and the interpretation difficult. Lifan overcame all those impediments and delivered both data and interpretations that, I believe, will shake the foundations of our thinking about supernovae and, perhaps, of gamma ray bursts.

While discussions of the possible utility of spectropolarimetry go back to the early 80s, there was no systematic attempt to acquire regular spectropolarimetry of supernovae until Lifan began his Hubble program at Texas. Then Lifan began to collect data at McDonald Observatory. For various technical reasons, he was forced to use the 78.7-inch (2.1-meter) Struve telescope. This is a very small telescope compared to the 315- and 394-inch (8- and 10-meter) giants coming online. To make up for the lack of aperture, Lifan had to make heroic, long, multi-night exposures.

Despite this handicap, he quickly doubled, then tripled, the number of supernovae for which polarization data was available. He obtained data on every known type of supernova. The sample is still relatively small, about 15 events, but it is clear that Lifan has basically written the book on supernova spectropolarimetry.

The next question becomes the source of the asymmetry. It could be an asymmetric density distribution. The density could be spherical, but the source of luminosity asymmetrically distributed. The light emerging from the supernova could be unpolarized, only to become so by passing though an asymmetric circumstellar medium ejected from the star before it exploded. The asymmetry could have something to do with the interstellar medium and nothing to do with the supernova at all. Lifan's next great contribution was to bring order to this welter of possible interpretations.

The first breakthrough in this work came as the data set slowly grew. Lifan quickly recognized that there was a dramatic trend, heretofore unrecognized. There are two basic types of supernova explosion mechanisms: one is the thermonuclear combustion of a white dwarf that explodes the star, leaving no compact remnant; the other is based on the collapse of an inner iron core to produce a neutron star (or perhaps a black hole in some special circumstances). Lifan discovered that the former were scarcely polarized, but that the latter were substantially polarized. That trend continues to this day as the sample continues to grow. Every core-collapse event ever observed by Lifan, or, by now, other teams, has displayed finite polarization and hence finite asymmetry. This has led to some dramatic rethinking of the theory of core collapse.

In addition to the basic discovery that all core-collapse events are asymmetric, Lifan discovered another critical trend. He found that the less mass that is contained in the outer hydrogen layers when the star explodes, the greater the polarization and the greater the asymmetry. He also discovered that for a given supernova, the longer one watched, the higher was the polarization and the asymmetry. In the former case, there is a smaller protective blanket and one is looking deep into the heart of the supernova. In the latter case, the effect is that at later times, the supernova matter thins out, so one is observing deeper into the supernovae. Both of these observations are telling the same story. The deeper into the supernova one looks, the greater the asymmetry. What this is saying is that the asymmetry is not some casual feature of the supernova environment, but coming from deep in the heart of the core-collapse explosion. It is the explosion process itself that is strongly asymmetric.

The traditional theory for this sort of explosion is that the iron-core collapse forms a neutron star and a burst of neutrinos. The neutrinos carry off 99 percent of the energy of the collapse, and the remaining 1 percent is enough to explode the star. So far, however, no one has been able to make a robust explosion, using computer models, with this mechanism. Even if it worked, it is very unlikely that the explosion would create the type of strong, long-lasting, asymmetries that Lifan's data revealed. Rather, the explosion must produce asymmetric streams of energy and hold those streams in place long enough to impose a permanent asymmetry. The explosion requires a jet.

This insight set off a rush of work on different explosion mechanisms. It has been established that energetic jets from the newly born neutron star can trigger the explosion and produce the observed asymmetries with no action from the neutrinos at all (although they may play some role in practice). There are now efforts to see whether magneto-rotational effects can produce those jets. This is a major change in 30-year's worth of thinking about the core-collapse mechanism. It is fed by yet another profound discovery by Lifan, that many core-collapse supernovae show not only asymmetry, but an asymmetry that is aligned on a single axis. The supernovae are bipolar, enhancing the argument for a jet-like explosion mechanism.

Most recently, Lifan has combined his insights with data from the Hubble Space Telescope on SN 1987A to show that the explosion is completely consistent with the type of bipolar configuration one expects from models of jet-induced explosions.

Hubble Fellows are chosen from among the best and the brightest. Most of them do excellent work. Few of them pioneer entirely new fields of study with far-reaching ramifications. Lifan Wang has done so.