MINT: a library that conserves vorticity or lateral fluxes when interpolating vector fields.

As part of the LFRic effort [1] to develop next generation weather and climate forecasting tools, we present a C++/Python package (MINT) designed to interpolate and regrid vector fields with staggered components. MINT implements an interpolation method that conserves line and flux integrals in two dimensions. Mathematical identities such as ∇ × ∇ = ∇ · ∇× = 0 are preserved to near machine accuracy after regridding. The method is immune to grid singularities, supports partially valid cells and the grid can be an unstructured collection of distorted quadrilateral cells.