The LOFAR Two-metre Sky Survey: VII. Third Data Release

Abstract: We present the third data release of the LOFAR Two-metre Sky Survey (LoTSS-DR3). The survey images cover 88% of the northern sky and were created from 12,950 hrs of data (18.6 PB) accumulated over 10.5 years. The images were produced through direction-independent and direction-dependent calibration pipelines that correct for instrumental effects as well as spatially and temporally varying ionospheric distortions. In our 120-168 MHz continuum mosaic images with an angular resolution of 6 arcsec (9 arcsec below declination 10$^\circ$) we catalogue 13,667,877 sources, formed from 16,943,656 Gaussian components. The scatter in the astrometric precision approximately follows the expected noise-like behaviour but with an additional systematic component of at least 0.24 arcsec that is likely due to calibration imperfections. The random flux density scale error is 6%, while the systematic offset was previously shown to be within 2%. The median sensitivity of our mosaics is 92$μ$Jy beam$^{-1}$. Completeness simulations, accounting for realistic source models, time- and bandwidth-smearing effects, and astrometric errors, indicate that we detect more than 95% of compact sources with integrated flux densities exceeding 9 times the local root mean square (RMS) noise. However, the recovered source counts in a particular integrated flux density bin do not match the injected counts until flux densities exceed 45 times the local RMS noise. The Euclidean-normalised differential source counts derived from the survey constrain the radio source population over five orders of magnitude and are in good agreement with previous deep and wide-area surveys. All data products are publicly available, including catalogues, individual-field Stokes I, Q, U, and V images, mosaicked Stokes I images, and $uv$ data with associated direction-dependent calibration solutions.

• The paper presents the DR3 survey covering 88% of the northern sky with nearly 14M radio sources, emphasizing refined astrometry and flux calibration.
• It employs advanced direction-dependent calibration and HEALPix-based mosaic construction to achieve a median sensitivity of 92μJy/beam and robust source extraction.
• The release sets a new benchmark for low-frequency radio astronomy, enhancing studies of galaxy evolution, radio AGN feedback, and diffuse emission.

The LOFAR Two-metre Sky Survey: Third Data Release

Survey Scope, Observational Design, and Data Processing

The third data release (DR3) of the LOFAR Two-metre Sky Survey (LoTSS) represents an extensive 120–168 MHz imaging program covering 88% of the northern sky, corresponding to 19,035 deg$^2$. Utilizing 12,950 hours of LOFAR HBA array data over a decade, the release catalogs 13,667,877 radio sources, detected with a typical angular resolution of $6''$ above declination $10^\circ$ ($9''$ below $10^\circ$). The survey infrastructure, comprising 48 core, 14 remote, and up to 14 international stations, facilitated high spectral and spatial sensitivity. The observations were meticulously scheduled to optimize elevation and minimize station area projection effects, yielding improved sensitivity in higher declination regions.

Data processing leveraged both direction-independent (LINC) and direction-dependent (DDF-pipeline) calibration workflows. These pipelines were orchestrated across distributed computing infrastructures, cumulatively consuming 20 million core hours. The direction-dependent calibration, which is essential at LOFAR frequencies due to severe ionospheric and primary beam effects, utilized multi-facet self-calibration (typically 45 directions per field) and iterative imaging cycles for robust removal of instrumental artifacts.

Figure 1: Computational breakdown for a typical DDF-pipeline run, highlighting the dominant expense of direction-dependent calibration.

Mosaic Construction, Astrometric Alignment, and Flux Calibration

Mosaic images were generated via weighted combinations of individual pointings, integrating RMS-based and primary beam weights. The DR3 mosaicking workflow adopted HEALPix-based segmentation to maximize sky coverage and facilitate robust cataloguing. Crucial refinements included astrometric correction using Pan-STARRS DR2 as the anchoring optical reference, yielding median RA/Dec offsets of $0.05''$/$0.06''$ (standard deviations: $0.24''$, $0.19''$) relative to FIRST for individual pointings, and even tighter alignment in mosaics.

Figure 2: Astrometric offset distributions between LoTSS-DR3 and FIRST, demonstrating the efficacy of alignment procedures.

The flux density scale calibration was anchored to NVSS, with systematic offsets kept below $2\%$ and random errors reduced to $6\%$ following iterative planar corrections across pointings. Inter-pointing alignment mitigated spatial gradients due to beam and calibration imperfections, advancing uniformity. Simulated analyses confirmed these corrections reduced pairwise flux discrepancies between sources from $11\%$ to $6\%$ globally, and to $2\%$ at smaller angular separations.

Figure 3: Iterative flux density scale alignment across LoTSS-DR3, illustrating convergence to low spatial variance.

Image Quality and Sensitivity

DR3 achieves a median mosaic sensitivity of $92\mu$Jy beam$^{-1}$, with central RMS ranging from $29\mu$Jy to $438\mu$Jy beam$^{-1}$ across fields (median $112\mu$Jy beam$^{-1}$, $\sigma=46\mu$Jy beam$^{-1}$). Sensitivity degrades at lower declinations due to projection effects but is consistently maintained elsewhere. The large dynamic range and coverage enable extragalactic and Galactic science at both full ($6''$/$9''$) and low ($20''$) resolution, with additional deep mosaics and Stokes Q, U, V products.

Figure 4: Matched RMS tracking per epoch, mapping noise evolution and solar cycle impact.

Figure 5: Declination-dependent sensitivity distribution and noise histogram for LoTSS-DR3 mosaics.

Compact Source Completeness and Systematics

Completeness simulations accounted for time/bandwidth smearing, astrometric errors, and source model realism. For compact sources, detection completeness exceeds 95% at flux densities $>9\times$ local RMS, but the recovered source counts accurately reflect intrinsic distributions only at $>45\times$ local RMS. PyBDSF-based source extraction, with completeness corrections derived from simulations, underpins robust catalogue construction and differential source count analyses.

Figure 6: Completeness curves for compact source detection as a function of integrated flux density, with corrections for smearing and astrometric uncertainty.

Differential Source Counts and Population Constraints

Euclidean-normalised differential source counts were derived across five orders of magnitude, leveraging both DR3 mosaics and LoTSS Deep fields. Counts are corrected via completeness simulations and match prior wide-area surveys (e.g., TGSS-ADR1, NVSS) at bright flux densities, with deep counts consistent with T-RECS population synthesis models and COSMOS field results at fainter levels. The faint-end spectral index ($\alpha\approx-0.6$) reflects increasing dominance of star-forming galaxies.

Figure 7: Euclidean-normalised 144 MHz source counts for LoTSS-DR3 and deep mosaics, corrected for completeness and systematics.

Data Products and Future Directions

DR3 offers an unprecedented radio catalogue and image collection: 2,551 individual fields, 1,580 mosaics, full Stokes imaging, $uv$ datasets with calibration solutions, and extensive Gaussian-component catalogues. All products are publicly accessible with persistent identifiers. The paper outlines prospects for further DR4 expansion post-LOFAR2.0 upgrade and international-station exploitation for sub-arcsecond imaging across the northern hemisphere (ILoTSS).

Figure 8: All-sky projection of DR3 mosaic and RMS coverage, highlighting expansion over DR1 (2%) and DR2 (27%) to DR3 (88%).

Figure 9: Typical extragalactic region in DR3, illustrating source multiplicity and radio morphological diversity.

Figure 10: Mosaic of the Galactic plane, revealing the density and structure of supernova remnant populations.

Implications and Theoretical Considerations

LoTSS-DR3 establishes a foundational, high-fidelity low-frequency survey with broad-reaching implications. The precision in astrometry and flux calibration enables systematic studies of radio AGN feedback, cosmological large-scale structure, star formation, diffuse emission, and polarisation science. The scale and completeness of the catalogue will inform galaxy evolution modeling, cross-survey population synthesis, and statistical analysis of faint radio populations. With imminent upgrades, LoTSS will facilitate expanded multi-wavelength database correlation and VLBI-scale imaging in low-frequency regimes.

Conclusion

The LOFAR Two-metre Sky Survey: Third Data Release (DR3) constitutes the single largest radio survey in both source number and area, with comprehensive control over systematics in flux density and astrometry. The resulting data products, completeness corrections, and source counts form a benchmark for low-frequency radio astronomy, providing critical input for both observational and theoretical studies. Future expansions and technological enhancements portend deeper insights and finer angular resolution, further broadening the scientific utility of the LoTSS archive (2602.15949).