Not yet reviewed by Pith; the record is open.

Magnetic fields are an energetically important component of star-formation galaxies, but it is often difficult to measure their properties from observations. One of the complexities stems from the fact that the magnetic fields, especially in spiral galaxies, have a two-scale nature: a large-scale field, coherent over ${\rm kpc}$ scales and a small-scale, random field with a scale of $\lesssim$ $100~{\rm pc}$. Moreover, it is known that the strength of small- and large-scale fields are comparable and this makes it even harder to find their imprints in radio polarisation observations such as the Faraday rotation measure, ${\rm RM}$, which is the integral over the path length of the product of the thermal electron density and the parallel component of the magnetic field to the line of sight. Here, we propose and demonstrate the use of second-order structure functions of ${\rm RM}$ computed with multiple higher-order stencils as a powerful analysis to separate the small- and large-scale magnetic field components. In particular, we provide new methods and calibrations to compute the scale and the strength of the large-scale magnetic field in the presence of small-scale magnetic fluctuations. We then apply the method to find the scale of large-scale magnetic fields in the nearby galaxies M51 and NGC 6946, using archival data and further discuss the need for computing the ${\rm RM}$ structure functions with higher-order stencils. With multiple modern radio polarisation observatories and eventually the Square Kilometre Array, ${\rm RM}$ observations will significantly improve in quantity and quality, and the higher-order stencil structure function techniques developed here can be used to extract information about multiscale magnetic fields in galaxies.