We present LOFAR High-Band Array (HBA) observations of the Herschel-ATLAS North Galactic Pole survey area. The survey we have carried out, consisting of four pointings covering around 142 square degrees of sky in the frequency range 126-173 MHz, does not provide uniform noise coverage but otherwise is representative of the quality of data to be expected in the planned LOFAR wide-area surveys, and has been reduced using recently developed 'facet calibration' methods at a resolution approaching the full resolution of the datasets (∼ 10 × 6 arcsec) and an rms off-source noise that ranges from 100 µJy beam −1 in the centre of the best fields to around 2 mJy beam −1 at the furthest extent of our imaging. We describe the imaging, cataloguing and source identification processes, and present some initial science results based on a 5-σ source catalogue. These include (i) an initial look at the radio/far-infrared correlation at 150 MHz, showing that many Herschel sources are not yet detected by LOFAR; (ii) number counts at 150 MHz, including, for the first time, observational constraints on the numbers of star-forming galaxies; (iii) the 150-MHz luminosity functions for active and starforming galaxies, which agree well with determinations at higher frequencies at low redshift, and show strong redshift evolution of the star-forming population; and (iv) some discussion of the implications of our observations for studies of radio galaxy life cycles.

Galaxy morphology is a fundamental quantity, that is essential not only for the full spectrum of galaxy-evolution studies, but also for a plethora of science in observational cosmology (e.g. as a prior for photometric-redshift measurements and as contextual data for transient lightcurve classifications). While a rich literature exists on morphological-classification techniques, the unprecedented data volumes, coupled, in some cases, with the short cadences of forthcoming 'Big-Data' surveys (e.g. from the LSST), present novel challenges for this field. Large data volumes make such datasets intractable for visual inspection (even via massively-distributed platforms like Galaxy Zoo), while short cadences make it difficult to employ techniques like supervised machine-learning, since it may be impractical to repeatedly produce training sets on short timescales. Unsupervised machine learning, which does not require training sets, is ideally suited to the morphological analysis of new and forthcoming surveys. Here, we employ an algorithm that performs clustering of graph representations, in order to group image patches with similar visual properties and objects constructed from those patches, like galaxies. We implement the algorithm on the Hyper-Suprime-Cam Subaru-Strategic-Program Ultra-Deep survey, to autonomously reduce the galaxy population to a small number (160) of 'morphological clusters', populated by galaxies with similar morphologies, which are then benchmarked using visual inspection. The morphological classifications (which we release publicly) exhibit a high level of purity, and reproduce known trends in key galaxy properties as a function of morphological type at z < 1 (e.g. stellar-mass functions, rest-frame colours and the position of galaxies on the star-formation main sequence). Our study demonstrates the power of unsupervised machine learning in performing accurate morphological analysis, which will become indispensable in this new era of deep-wide surveys.

Aims. We aim to study the far-infrared radio correlation (FIRC) at 150 MHz in the local Universe (at a median redshift z ∼ 0.05) and improve the use of the rest-frame 150-MHz luminosity, L 150 , as a star-formation rate (SFR) tracer, which is unaffected by dust extinction. Methods. We cross-match the 60-µm selected Revised IRAS Faint Source Survey Redshift (RIFSCz) catalogue and the 150-MHz selected LOFAR value-added source catalogue in the Hobby-Eberly Telescope Dark Energy Experiment (HETDEX) Spring Field. We estimate L 150 for the cross-matched sources and compare it with the total infrared (IR) luminosity, L IR , and various SFR tracers. Results. We find a tight linear correlation between log L 150 and log L IR for star-forming galaxies, with a slope of 1.37. The median q IR value (defined as the logarithm of the L IR to L 150 ratio) and its rms scatter of our main sample are 2.14 and 0.34, respectively. We also find that log L 150 correlates tightly with the logarithm of SFR derived from three different tracers, i.e., SFR Hα based on the Hα line luminosity, SFR 60 based on the rest-frame 60-µm luminosity and SFR IR based on L IR , with a scatter of 0.3 dex. Our best-fit relations between L 150 and these SFR tracers are, log L 150 (L ) = 1.35(±0.06) × log SFR Hα (M /yr) + 3.20(±0.06), log L 150 (L ) = 1.31(±0.05) × log SFR 60 (M /yr) + 3.14(±0.06), and log L 150 (L ) = 1.37(±0.05) × log SFR IR (M /yr) + 3.09(±0.05), which show excellent agreement with each other.

The radio and far-infrared luminosities of star-forming galaxies are tightly correlated over several orders of magnitude; this is known as the far-infrared radio correlation (FIRC). Previous studies have shown that a host of factors conspire to maintain a tight and linear FIRC, despite many models predicting deviation. This discrepancy between expectations and observations is concerning since a linear FIRC underpins the use of radio luminosity as a star-formation rate indicator. Using LOFAR 150 MHz, FIRST 1.4 GHz, and Herschel infrared luminosities derived from the new LOFAR/H-ATLAS catalogue, we investigate possible variation in the monochromatic (250 µm) FIRC at low and high radio frequencies. We use statistical techniques to probe the FIRC for an optically-selected sample of 4,082 emission-line classified star-forming galaxies as a function of redshift, effective dust temperature, stellar mass, specific star formation rate, and mid-infrared colour (an empirical proxy for specific star formation rate). Although the average FIRC at high radio frequency is consistent with expectations based on a standard power-law radio spectrum, the average correlation at 150 MHz is not. We see evidence for redshift evolution of the FIRC at 150 MHz, and find that the FIRC varies with stellar mass, dust temperature and specific star formation rate, whether the latter is probed using MAGPHYS fitting, or using mid-infrared colour as a proxy. We can explain the variation, to within 1σ , seen in the FIRC over mid-infrared colour by a combination of dust temperature, redshift, and stellar mass using a Bayesian partial correlation technique.