GMIMS-DRAGONS: A Faraday Depth Survey of the Northern Sky Covering 350-1030 MHz

Polarized synchrotron emission at meter to centimeter wavelengths provides an effective tracer of the Galactic magnetic field. Calculating Faraday depth, the most useful parameter for mapping the line-of-sight magnetic field, requires observations covering wide frequency bands with many channels. As part of the Global Magneto-Ionic Medium Survey (GMIMS), we have observed polarized emission spanning 350-1030 MHz over the northern sky, in the declination range ${-20^{\circ}}\leqδ\leq{90^{\circ}}$. We used the 15 m telescope at the Dominion Radio Astrophysical Observatory (DRAO), equipped to receive orthogonal circular polarizations, with the Onsala Space Observatory band 1 feed developed for the SKA Project. Angular resolution varies across the band from $1.3^{\circ}$ to $3.6^{\circ}$. A digital spectrometer provided 42 kHz frequency resolution. Data were taken with the telescope moving rapidly in azimuth and are absolutely calibrated in intensity. Approximately 25% of the data were lost due to radio-frequency interference. The resolution in Faraday depth is $\sim6$ rad m$^{-2}$, and features as wide as $\sim38$ rad m$^{-2}$ are represented. The median sensitivity of the Faraday depth cube is 11 mK. Approximately 55% of sight-lines in this survey show Faraday complexity. This dataset, called ``DRAO GMIMS of the Northern Sky’’ (DRAGONS), is the first to probe Faraday depth of the northern sky in its frequency range and will support many scientific investigations. The data will be used to calibrate surveys with higher angular resolution, particularly Galactic foreground maps from the Canadian Hydrogen Intensity Mapping Experiment, and to provide information on large structures for aperture-synthesis telescopes, particularly the DRAO Synthesis Telescope. The data are available through the Canadian Astronomy Data Centre.

💡 Research Summary

This paper presents the results of the DRAO GMIMS of the Northern Sky (DRAGONS) survey, a comprehensive Faraday‑depth mapping effort covering the declination range –20° ≤ δ ≤ 90° at frequencies from 350 MHz to 1030 MHz. The observations were carried out with the 15 m offset Gregorian reflector at the Dominion Radio Astrophysical Observatory (DRAO), equipped with a quad‑ridged, flared‑horn feed originally developed for the SKA‑Band 1 project at the Onsala Space Observatory. The feed delivers orthogonal left‑ and right‑hand circular polarizations (L,R), which are then cross‑correlated to obtain Stokes Q and U with high stability, avoiding the large systematic errors that arise when deriving Q from the difference of two total‑power signals.

Key instrumental features include the application of the Mizuguchi technique to shape the reflector surfaces, reducing the L–R beam offset to less than 5 % of the beam width across the band. The half‑power beam width (HPBW) varies from 3.6° at 350 MHz to 1.3° at 1030 MHz; for Faraday synthesis all channel maps are convolved to a common 3.6° resolution. The receiver chain provides a clean passband from the feed cutoff at ∼350 MHz to a low‑pass filter at ∼1030 MHz, with system temperatures ranging from ~12 K (350 MHz) to ~35 K (1030 MHz). A digital spectrometer records the signal with 42 kHz channel spacing, but the final data products are binned to 1 MHz to improve signal‑to‑noise and to facilitate RFI excision.

Observations employed a rapid azimuth‑scan strategy, enabling absolute intensity calibration. Approximately 25 % of the raw data were lost to radio‑frequency interference (RFI); the remaining data were processed through an automated RFI mask, supplemented by manual inspection. The processing pipeline consists of (1) basic flagging and RFI removal, (2) absolute calibration including frequency‑dependent gain corrections, (3) generation of Stokes I, Q, and U maps, (4) smoothing all frequency channels to a uniform 3.6° beam, (5) RM synthesis using a Fourier transform of the complex polarization as a function of λ², and (6) a CLEAN‑like deconvolution to mitigate the limited λ² sampling window. The resulting Faraday‑depth cube has a resolution of ~6 rad m⁻², a maximum recoverable scale of ~38 rad m⁻², and a median sensitivity of 11 mK. More than 55 % of sight‑lines exhibit Faraday complexity, i.e., multiple components in the Faraday spectrum, indicating a richly structured magneto‑ionic medium along the line of sight.

The survey fills a unique niche in the Global Magneto‑Ionic Medium Survey (GMIMS) suite. Its λ² coverage (0.085–0.73 m²) is the widest among all GMIMS components, providing unprecedented sensitivity to Faraday‑depth structure at low frequencies. The overlap with the GMIMS‑Low‑Band‑South (300–480 MHz) allows cross‑validation, while the frequency overlap with the Canadian Hydrogen Intensity Mapping Experiment (CHIME; 400–800 MHz) makes DRAGONS an ideal dataset for calibrating CHIME’s polarization maps and supplying the missing short‑baseline information for interferometric surveys. Likewise, the data will support the DRAO Synthesis Telescope and other interferometers (e.g., LOFAR, POSSUM) by supplying large‑scale polarized emission that is filtered out in interferometric imaging.

All data products—including calibrated Stokes cubes, the Faraday‑depth cube, RFI masks, and detailed metadata—are publicly available through the Canadian Astronomy Data Centre (CADC) in FITS format. The authors outline several forthcoming science papers that will exploit DRAGONS for (i) three‑dimensional mapping of the Galactic magnetic field, (ii) studies of magnetic field morphology in star‑forming regions and molecular clouds, (iii) investigations of the warm ionized, warm partially ionized, and low‑ionization neutral media via their distinct Faraday signatures, and (iv) synergy with high‑resolution surveys to construct multi‑scale magnetic field models.

In summary, the DRAGONS survey demonstrates that a single‑dish, broadband, low‑resolution instrument can deliver high‑fidelity Faraday‑depth information across most of the northern sky. By providing a well‑calibrated, wide‑band polarization dataset, it establishes a crucial bridge between all‑sky single‑dish surveys and high‑resolution interferometric observations, thereby advancing our ability to reconstruct the three‑dimensional magneto‑ionic structure of the Milky Way.