Recurrent novae are compact binaries where Roche lobe overflow is spilling material onto a white dwarf, where the matter accumulates until it is hot enough to start a thermonuclear runaway that creates a nova eruption at least once per century. The recurrence time is so fast due to the white dwarf being near the Chandrasekhar mass and due to a high accretion rate. If recurrent novae have massive white dwarves with material being piled onto them at a fast rate, then they must inevitably collapse when the mass accumulates to the Chandrasekhar limit and then produce a Type Ia supernova. With this logic, recurrent novae are a leading contender as being the progenitor system for Type Ia SNe. Various other types of systems have been proposed as progenitors and this has resulted in a still-unresolved 'Type Ia progenitor problem' that has been prominent for over 40 years. This problem is now critical to large enterprises of broad importance since Type Ia events are being used as standard candles for cosmology to determine the age and fate of the Universe. The solution of the progenitor problem is required to calculate the evolution with red shift of the supernovae peak brightnesses.
Here, Dr. Schaefer, along with undergraduates and a high school student, will attempt to answer the fundamental question "Are recurrent novae the progenitors of Type Ia supernovae?" by performing two significant tests. The first is whether or not the white dwarfs in such systems are gaining or losing mass. To answer this, the mass ejected during eruption will be measured for two recurrent novae by timing their orbital period changes across recent eruptions. The rate of accretion for these stars will be derived from UBVRIJHK photometry and light curve modeling. An exhaustive search of archive plates dating back to 1889 will determine the rates of eruption. Thus it will be possible to assess the mass accumulated between eruptions to compare with the mass lost. The second question is whether there are enough recurrent novae in our Galaxy and the Local Group to allow for the observed Type Ia events. This will be accomplished through various tests of the discovery frequency, which will result in a realistic estimate of the total number of recurrent novae in our Galaxy, the LMC, and in M31. The novae lifetimes will be derived using their measured accretion rates and the death rate will be the total number divided by the lifetime. This can then be directly compared with the Type Ia event rate.
As part of this project, Dr. Schaefer will continue his active public outreach work. This program will also involve undergraduate students and a high school student in front-edge research as well as make up the bulk of a Ph.D. dissertation for a graduate student at Louisiana State University.
A long-standing problem in astrophysics is the nature of the progenitor systems that explode as Type Ia supernovae (SNe Ia). This progenitor problem is of importance, both as a keystone for stellar evolution and because the progenitor must be known to allow calculations of evolution for precision cosmology. Indeed, the recent Decadal Survey of the National Academy of Sciences has identified the progenitor question as one of the top nine questions in all astrophysics. Many progenitor systems have been proposed, most involving a degenerate carbon-oxygen white dwarf. The best candidate progenitors are the recurrent novae (RNe) because they are certainly systems where the white dwarf is near the maximum limit (the Chandrasekhar mass) and where material is being piled on at a very high rate so as to push the white dwarf over the top. However, two problems arise in the case for RNe as progenitors, the first is whether there are enough RNe so as to allow for the observed rate of SNe Ia, and the second is whether the white dwarfs are really increasing in mass due to the large amount of material ejected during each eruption. The results of this NSF grant make large advances in answering both of these questions. The first problem (the number of RNe) was initially in poor shape, with prior answers having (now) known orders-of-magnitude errors. For this program, a massive demographics survey was made. The bulk of the world’s archival plates were exhaustively searched for missed eruptions, and indeed six were discovered. All existing eruption light curve data was collected and analyzed so as to produce definitive light curves for calculating discovery probabilities. Detailed surveys of the plate archives and many visual data bases were made so as to derive the eruption discovery probabilities for each RN for every year since 1890. Detailed and exhaustive plate searches were made of many apparently ‘classical novae’ (i.e., with only one known eruption though there might be more that were simply missed) to seek hidden RNe, and indeed one new RN (V2487 Oph) was discovered. An exhaustive catalog of nova light curves was created, with orders of magnitude more data than all prior catalogs, so as to evaluate the rate of hidden RNe. A literature survey of all known ‘classical novae’ was made for any of four properties newly identified as pointing to RNe. The result of this exhaustive work is that RNe are roughly 100X more common than all prior estimates, with the difference being that prior work had ignored the discovery efficiency for RN eruptions. Based on this, the preliminary answer is that there are enough RNe to supply the observed number of SNe Ia. The second problem (whether the white dwarfs are gaining mass) had previously had only estimates for the ejected mass with several orders-of-magnitude uncertainty. For this NSF program, the first accurate estimates were made of the mass ejected by two RN eruptions (U Sco in 1999 and CI Aql in 2000). The new method was to time the eclipses over long periods of time before and after the eruption so as to get accurate pre-eruption and post-eruption orbital periods. By Kepler’s Law, the change in orbital period across the eruption directly gives the mass ejected by the eruption. The result is that both eruptions were measured to eject a very small amount of material, and in the case of CI Aql, the mass of the material was substantially less than had been accreted in the prior inter-eruption interval. This means that the white dwarfs are gaining in mass on average over time, so they will soon reach the critical mass and become SNe Ia. The result of this NSF program have been to answer the two outstanding questions for the most promising SN Ia progenitor candidate. The conclusions are that there are enough RNe to provide all the observed SNe Ia and that the white dwarfs are increasing in mass so that a SN Ia explosion is inevitable. With this, RNe must be the dominant progenitor.
