Adaptive Mesh Refinement in gPLUTO on staggered grids

High-resolution magnetohydrodynamic (MHD) simulations rely on Adaptive Mesh Refinement (AMR) to resolve localized features efficiently, yet maintaining the magnetic field divergence constraint across refinement boundaries remains non-trivial.

In this talk, I will present an overview of general AMR strategies before describing the implementation being developed in gPLUTO with particular emphasis on staggered grids. On these grids, magnetic fields are stored on cell faces and electric fields on cell edges.

I will discuss the resulting challenges in communication and synchronization across refinement levels, compare divergence cleaning with constrained transport, and explain the role of flux and electromotive-force corrections at coarse–fine interfaces to prevent the development of magnetic monopoles at AMR boundaries. The talk will conclude with the current status of the implementation and ongoing validation.