After Julien found me on Tuesday morning, I was invited to attend the LPNHE group meeting. The format is fairly relaxed, a combination of quick journal club presentations on recent papers and progress reports on students’ research. Nicolas, Julien and Pierre Astier were all present, as well as Reynald Pain, the director of LPNHE, whom I know from SNfactory work.
Fang Huang, a visiting student from Beijing, presented her poster from the MPA meeting on the use of SNe IIP for cosmology. These are core-collapse supernovae with thick hydrogen envelopes, the explosions of red supergiant stars. They stay bright for a period of a month or two (the “plateau” phase for which the P in “SNe IIP” stands), then very suddenly become faint after the hydrogen envelope has cooled and the inner ejecta (mostly carbon, oxygen and heavier elements) have been revealed. Fang and her group report a relation between how quickly this switch-off happens and the plateau luminosity, which allows them to use SNe IIP as standard candles in much the same way (and, they claim, to similar precision) as SNe Ia. Other groups have developed different methods to calibrate supernovae of this type, using combinations of photometric and spectroscopic observations, and sometimes sophisticated radiation transfer modeling, to derive a distance to each supernova.
SNe IIP shouldn’t quit their day job yet, though. They are on average ten times fainter than SNe Ia, meaning you need a telescope three times bigger to make a SN IIP Hubble diagram competitive with one made from SNe Ia at the same distances. (In fact, the sample that Fang showed were so close to us that their host galaxies all had Cepheid or Tully-Fisher distances.) Their spectra may not be as uniform as SNe Ia, so that a mean spectral template to derive K-corrections may not work well, and some other, more probabilistic method may be needed to get accurate distances from SNe IIP at higher redshift. For the same reason, their colors are not necessarily uniform or well-understood1, so that estimates of extinction by dust will be harder to obtain than for SNe Ia.
I was then invited on the spot to share my work on bolometric light curves. I gave my Parisian colleagues the same report I gave at last month’s mini-meeting at the RSAA. I had plenty of questions from Julien and Reynald about the robustness and accuracy of the method. I argued that the dominant systematic error in using bolometric light curves to measure masses for SNe powered by radioactivity relates to uncertainty about the distribution of radioactive elements in the expanding ejecta: for a given mass of radioactive material (say, 56Ni), a supernova in which that material is buried deep in the ejecta will release the radioactive energy more slowly, and hence be brighter and slower to fade, than a supernova in which it lives near the surface. Right now we just assume a plausible functional form to parametrize this aspect of our model. Perhaps with further study we can figure out how to constrain the 56Ni distribution with observables, or marginalize over possible distributions to get representative uncertainties that are, as they say, good enough for government work.
1Actually, on the plateau phase the photosphere will have a particular temperature, set by the way ionized hydrogen atoms recombine with their electrons as they cool. But other things, such as the density and composition of the SN’s environment, might drive differences in the spectra which aren’t as pronounced for SNe Ia.