Dr Becky Smethurst

Abstract:

We use Bayesian convolutional neural networks and a novel generative model of Galaxy Zoo volunteer responses to infer posteriors for the visual morphology of galaxies. Bayesian CNN can learn from galaxy images with uncertain labels and then, for previously unlabelled galaxies, predict the probability of each possible label. Our posteriors are well-calibrated (e.g. for predicting bars, we achieve coverage errors of 11.8 per cent within a vote fraction deviation of 0.2) and hence are reliable for practical use. Further, using our posteriors, we apply the active learning strategy BALD to request volunteer responses for the subset of galaxies which, if labelled, would be most informative for training our network. We show that training our Bayesian CNNs using active learning requires up to 35–60 per cent fewer labelled galaxies, depending on the morphological feature being classified. By combining human and machine intelligence, Galaxy zoo will be able to classify surveys of any conceivable scale on a time-scale of weeks, providing massive and detailed morphology catalogues to support research into galaxy evolution.

Abstract:

Abstract Recent observations and simulations have revealed the dominance of secular processes over mergers in driving the growth of both supermassive black holes (SMBH) and galaxy evolution. Here we obtain narrowband imaging of AGN powered outflows in a sample of 12 galaxies with disk-dominated morphologies, whose history is assumed to be merger-free. We detect outflows in 10/12 sources in narrow band imaging of the $\mathrm{\left[ O \small {III}\right] }$ $5007~\mathring{\rm A}$ emission using filters on the Shane-3m telescope. We calculate a mean outflow rate for these AGN of $0.95\pm 0.14~\rm {M}_{\odot }~\rm {yr}^{-1}$. This exceeds the mean accretion rate of their SMBHs ($0.054\pm 0.039~\rm {M}_{\odot }~\rm {yr}^{-1}$) by a factor of ∼18. Assuming that the galaxy must provide at least enough material to power both the AGN and the outflow, this gives a lower limit on the average inflow rate of $\sim 1.01\pm 0.14~\rm {M}_{\odot }~\rm {yr}^{-1}$, a rate which simulations show can be achieved by bars, spiral arms and cold accretion. We compare our disk dominated sample to a sample of nearby AGN with merger dominated histories and show that the black hole accretion rates in our sample are 5 times higher (4.2σ) and the outflow rates are 5 times lower ( 2.6σ). We suggest that this could be a result of the geometry of the smooth, planar inflow in a secular dominated system, which is both spinning up the black hole to increase accretion efficiency and less affected by feedback from the outflow, than in a merger-driven system with chaotic quasi-spherical inflows. This work provides further evidence that secular processes are sufficient to fuel SMBH growth.

Abstract:

We use classifications provided by citizen scientists though Galaxy Zoo to investigate the correlation between bulge size and arm winding in spiral galaxies. Whilst the traditional spiral sequence is based on a combination of both measures, and is supposed to favour arm winding where disagreement exists, we demonstrate that, in modern usage, the spiral classifications Sa–Sd are predominantly based on bulge size, with no reference to spiral arms. Furthermore, in a volume limited sample of galaxies with both automated and visual measures of bulge prominence and spiral arm tightness, there is at best a weak correlation between the two. Galaxies with small bulges have a wide range of arm winding, while those with larger bulges favour tighter arms. This observation, interpreted as revealing a variable winding speed as a function of bulge size, may be providing evidence that the majority of spiral arms are not static density waves, but rather wind-up over time. This suggests the ‘winding problem’ could be solved by the constant reforming of spiral arms, rather than needing a static density wave. We further observe that galaxies exhibiting strong bars tend to have more loosely wound arms at a given bulge size than unbarred spirals. This observations suggests that the presence of a bar may slow the winding speed of spirals, and may also drive other processes (such as density waves) that generate spiral arms. It is remarkable that after over 170 years of observations of spiral arms in galaxies our understanding of them remains incomplete.