We study the impact of cluster environment on the morphology of a sample of 4304 extended radio galaxies from Radio Galaxy Zoo. A total of 87% of the sample lies within a projected 15 Mpc of an optically identified cluster. Brightest cluster galaxies (BCGs) are more likely than other cluster members to be radio sources, and are also moderately bent. The surface density as a function of separation from cluster center of non-BCG radio galaxies follows a power law with index −1.10 ± 0.03 out to 10 r ₅₀₀ (∼7 Mpc), which is steeper than the corresponding distribution for optically selected galaxies. Non-BCG radio galaxies are statistically more bent the closer they are to the cluster center. Within the inner 1.5 r ₅₀₀ (∼1 Mpc) of a cluster, non-BCG radio galaxies are statistically more bent in high-mass clusters than in low-mass clusters. Together, we find that non-BCG sources are statistically more bent in environments that exert greater ram pressure. We use the orientation of bent radio galaxies as an indicator of galaxy orbits and find that they are preferentially in radial orbits. Away from clusters, there is a large population of bent radio galaxies, limiting their use as cluster locators; however, they are still located within statistically overdense regions. We investigate the asymmetry in the tail length of sources that have their tails aligned along the radius vector from the cluster center, and find that the length of the inward-pointing tail is weakly suppressed for sources close to the center of the cluster.

In this work, we introduce attention as a state-of-the-art mechanism for classification of radio galaxies, using convolutional neural networks. We present an attention-based model that performs on par with previous classifiers while using more than 50 per cent fewer parameters than the next smallest classic CNN application in this field. We demonstrate quantitatively how the selection of normalization and aggregation methods used in attention-gating can affect the output of individual models, and show that the resulting attention maps can be used to interpret the classification choices made by the model. We observe that the salient regions identified by the our model align well with the regions an expert human classifier would attend to make equivalent classifications. We show that while the selection of normalization and aggregation may only minimally affect the performance of individual models, it can significantly affect the interpretability of the respective attention maps and by selecting a model which aligns well with how astronomers classify radio sources by eye, a user can employ the model in a more effective manner.

In this work, we examine the classification accuracy and robustness of a state-of-the-art semi-supervised learning (SSL) algorithm applied to the morphological classification of radio galaxies. We test if SSL with fewer labels can achieve test accuracies comparable to the supervised state of the art and whether this holds when incorporating previously unseen data. We find that for the radio galaxy classification problem considered, SSL provides additional regularization and outperforms the baseline test accuracy. However, in contrast to model performance metrics reported on computer science benchmarking data sets, we find that improvement is limited to a narrow range of label volumes, with performance falling off rapidly at low label volumes. Additionally, we show that SSL does not improve model calibration, regardless of whether classification is improved. Moreover, we find that when different underlying catalogues drawn from the same radio survey are used to provide the labelled and unlabelled data sets required for SSL, a significant drop in classification performance is observed, highlighting the difficulty of applying SSL techniques under data set shift. We show that a class-imbalanced unlabelled data pool negatively affects performance through prior probability shift, which we suggest may explain this performance drop, and that using the Fréchet distance between labelled and unlabelled data sets as a measure of data set shift can provide a prediction of model performance, but that for typical radio galaxy data sets with labelled sample volumes of $\mathcal {O}(10^3)$, the sample variance associated with this technique is high and the technique is in general not sufficiently robust to replace a train–test cycle.

Radio galaxies can extend far beyond the stellar component of their originating host galaxies, and their radio emission can consist of multiple discrete components. Furthermore, the apparent source structure will depend on survey sensitivity, resolution and the observing frequency. Associated discrete radio components and their originating host galaxy are typically identified through a visual comparison of radio and mid-infrared survey images. We present the first data release of Radio Galaxy Zoo, an online citizen science project that enlists the help of citizen scientists to cross-match extended radio sources from the Faint Images of the Radio Sky at Twenty Centimeters (FIRST) and the Australia Telescope Large Area Survey (ATLAS) surveys, often with complex structure, to host galaxies in 3.6 $\mu$m infrared images from the Wide-field Infrared Survey Explorer (WISE) and the SpitzerSpace Telescope. This first data release consists of 100 185 classifications for 99 146 radio sources from the FIRST survey and 583 radio sources from the ATLAS survey. We include two tables for each of the FIRST and ATLAS surveys: (1) the identification of all components making up each radio source and (2) the cross-matched host galaxies. These classifications have an average reliability of 0.83 based on the weighted consensus levels of our citizen scientists. The reliability of the DR1 catalogue has been further demonstrated through several parallel studies which used the pre-release versions of this catalogue to train and prototype machine learning-based classifiers. We also include a brief description of the radio source populations catalogued by RGZ DR1.

We present a novel natural language processing (NLP) approach to deriving plain English descriptors for science cases otherwise restricted by obfuscating technical terminology. We address the limitations of common radio galaxy morphology classifications by applying this approach. We experimentally derive a set of semantic tags for the Radio Galaxy Zoo EMU (Evolutionary Map of the Universe) project and the wider astronomical community. We collect 8486 plain English annotations of radio galaxy morphology, from which we derive a taxonomy of tags. The tags are plain English. The result is an extensible framework, which is more flexible, more easily communicated, and more sensitive to rare feature combinations, which are indescribable using the current framework of radio astronomy classifications.

We present a model for generating postage stamp images of synthetic Fanaroff–Riley Class I and Class II radio galaxies suitable for use in simulations of future radio surveys such as those being developed for the Square Kilometre Array. This model uses a fully connected neural network to implement structured variational inference through a variational autoencoder and decoder architecture. In order to optimize the dimensionality of the latent space for the autoencoder, we introduce the radio morphology inception score (RAMIS), a quantitative method for assessing the quality of generated images, and discuss in detail how data pre-processing choices can affect the value of this measure. We examine the 2D latent space of the VAEs and discuss how this can be used to control the generation of synthetic populations, whilst also cautioning how it may lead to biases when used for data augmentation.