Abstract. We examine some practical aspects of using a mushy-layer Rayleigh number for the interpretation of sea-ice-core data. In principle, such analysis should allow one to determine convectively active regions within the ice core by identifying those regions in which the mush-Rayleigh number is super-critical. In practice, however, a quantitative analysis is complicated by uncertainties regarding the specific formulation of both the mush-Rayleigh number itself and of the sea-ice permeability that is crucial for quantifying the Rayleigh number. Additionally, brine loss from highly permeable sections of the ice core, in particular close to the ice–ocean interface, and typically weekly ice core sampling, limit the practical applicability of the Rayleigh number for ice-core interpretation. We here quantify these uncertainties, suggest a standard method for the computation of the Rayleigh number for sea ice and discuss possibilities and limitations of ice-core interpretation based on the Rayleigh number.