What it is
A submanifold L of half the dimension of a symplectic manifold is Lagrangian if the symplectic form vanishes on it. Smoczyk (1996) proved that, in a Kähler–Einstein ambient manifold, mean curvature flow preserves this condition. In a Calabi–Yau manifold each Lagrangian has a Lagrangian angle θ, and the mean curvature vector is J∇θ. Under the flow, θ solves the heat equation.
The fixed points are special Lagrangians, those with constant angle. They are calibrated, so they minimize volume in their homology class. Thomas and Yau (2002) conjectured that a Lagrangian satisfying a stability condition flows for all time and converges to a special Lagrangian in its Hamiltonian isotopy class. Joyce later refined the picture in terms of Bridgeland stability on the Fukaya category, with the flow continued through singularities by surgeries.
Why it matters
Special Lagrangians are central to mirror symmetry: the SYZ conjecture describes Calabi–Yau manifolds as fibred by special Lagrangian tori. Lagrangian mean curvature flow is the natural way to construct them, as Ricci flow constructs Einstein metrics. It also has a feature Ricci flow lacks: the Lagrangian condition rules out some singularities. Neves showed that for zero-Maslov Lagrangians the tangent flows are unions of special Lagrangian cones, not the cylinders seen in hypersurface flows. Yet finite-time singularities cannot be avoided in general: Neves (2013) found them in every Hamiltonian isotopy class of a Lagrangian surface in a Calabi–Yau surface.
Where it stands
The Thomas–Yau conjecture is open in general. It has been proved in settings with extra symmetry, such as circle-invariant Lagrangians in hyperkähler 4-manifolds (Lotay–Oliveira). Current work classifies Lagrangian self-shrinkers, translators and ancient solutions (see self-shrinkers and translators), studies flows of graphs, and develops the surgeries Joyce's programme needs.