Is Pantheon biased?
New publication on Arxiv with Pablo Lemos and Ofer Lahav! We used the weak lensing estimator developed in our previous paper to look at whether cosmological parameters estimated from the Pantheon SN Ia dataset are biased.
Here’s a brief summary of the paper. It’s a well-known argument (due to Weinberg) that lensing conserves flux. The logic is that lensing neither creates nor destroys photons. So, if one throws a sphere at constant distance around a source, while some parts of the sphere may see higher intensity, others must therefore see lower intensity as the photon flux through the sphere remains unchanged (more analysis can be found in Kaiser and Peacock).
But what if the observer sits preferentially at regions of higher or lower flux? This might happen if the observation strategy avoids high-density regions, for example due to obscuration or a desire to avoid foregrounds in spectroscopy. Conversely at the magnitude limit of the survey, the magnified SN Ia are more likely to be seen. While bias corrections try to fix this, they assume a statistical ensemble that may not be realised in practice.
Somewhat surprisingly, no-one seems to have checked this before. In fact, many authors just assume flux is conserved and start working from there. You can find our results here.
