Search for: All records

ABSTRACT We present detailed morphology measurements for 8.67 million galaxies in the DESI Legacy Imaging Surveys (DECaLS, MzLS, and BASS, plus DES). These are automated measurements made by deep learning models trained on Galaxy Zoo volunteer votes. Our models typically predict the fraction of volunteers selecting each answer to within 5–10 per cent for every answer to every GZ question. The models are trained on newly collected votes for DESI-LS DR8 images as well as historical votes from GZ DECaLS. We also release the newly collected votes. Extending our morphology measurements outside of the previously released DECaLS/SDSS intersection increases our sky coverage by a factor of 4 (5000–19 000 deg2) and allows for full overlap with complementary surveys including ALFALFA and MaNGA. 

ABSTRACT Astronomers have typically set out to solve supervised machine learning problems by creating their own representations from scratch. We show that deep learning models trained to answer every Galaxy Zoo DECaLS question learn meaningful semantic representations of galaxies that are useful for new tasks on which the models were never trained. We exploit these representations to outperform several recent approaches at practical tasks crucial for investigating large galaxy samples. The first task is identifying galaxies of similar morphology to a query galaxy. Given a single galaxy assigned a free text tag by humans (e.g. ‘#diffuse’), we can find galaxies matching that tag for most tags. The second task is identifying the most interesting anomalies to a particular researcher. Our approach is 100 per cent accurate at identifying the most interesting 100 anomalies (as judged by Galaxy Zoo 2 volunteers). The third task is adapting a model to solve a new task using only a small number of newly labelled galaxies. Models fine-tuned from our representation are better able to identify ring galaxies than models fine-tuned from terrestrial images (ImageNet) or trained from scratch. We solve each task with very few new labels; either one (for the similarity search) or several hundred (for anomaly detection or fine-tuning). This challenges the longstanding view that deep supervised methods require new large labelled data sets for practical use in astronomy. To help the community benefit from our pretrained models, we release our fine-tuning code zoobot. Zoobot is accessible to researchers with no prior experience in deep learning. 

Abstract We have obtained sensitive dust continuum polarization observations at 850 μ m in the B213 region of Taurus using POL-2 on SCUBA-2 at the James Clerk Maxwell Telescope as part of the B -fields in STar-forming Region Observations (BISTRO) survey. These observations allow us to probe magnetic field ( B -field) at high spatial resolution (∼2000 au or ∼0.01 pc at 140 pc) in two protostellar cores (K04166 and K04169) and one prestellar core (Miz-8b) that lie within the B213 filament. Using the Davis–Chandrasekhar–Fermi method, we estimate the B -field strengths in K04166, K04169, and Miz-8b to be 38 ± 14, 44 ± 16, and 12 ± 5 μ G, respectively. These cores show distinct mean B -field orientations. The B -field in K04166 is well ordered and aligned parallel to the orientations of the core minor axis, outflows, core rotation axis, and large-scale uniform B -field, in accordance with magnetically regulated star formation via ambipolar diffusion taking place in K04166. The B -field in K04169 is found to be ordered but oriented nearly perpendicular to the core minor axis and large-scale B -field and not well correlated with other axes. In contrast, Miz-8b exhibits a disordered B -field that shows no preferred alignment with the core minor axis or large-scale field. We found that only one core, K04166, retains a memory of the large-scale uniform B -field. The other two cores, K04169 and Miz-8b, are decoupled from the large-scale field. Such a complex B -field configuration could be caused by gas inflow onto the filament, even in the presence of a substantial magnetic flux.