What it is
The minimal model program (MMP) of birational geometry simplifies a projective variety X step by step. It contracts curves on which the canonical bundle KX is negative, through divisorial contractions and flips, until either KX is nef (a minimal model) or the variety fibres over a lower-dimensional base with Fano fibres (a Mori fibre space).
The normalized Kähler–Ricci flow ∂tω=−Ric(ω)−ω moves the Kähler class along the straight line displayed above, towards 2πc1(KX). Tian and Zhang proved that the flow exists exactly as long as that class stays Kähler. So singularities form at a time predicted by cohomology, and the classes reached there are those that the MMP contracts.
- If KX is nef, the flow exists for all time. If KX is also big, it converges to the singular Kähler–Einstein metric on the canonical model (Tsuji; Tian–Zhang).
- If KX is semi-ample, Song and Tian proved that the flow collapses onto the canonical model, converging to a generalized Kähler–Einstein metric twisted by the geometry of the fibres. They also showed the scalar curvature stays bounded.
- At finite-time singularities, Song and Weinkove showed that the flow can contract exceptional divisors in the Gromov–Hausdorff sense, and Song and Tian constructed weak flows on varieties with log terminal singularities that continue uniquely through divisorial contractions and flips.
Why it matters
The programme gives Ricci flow singularities an algebraic meaning: a finite-time singularity is a contraction or a flip, and long-time collapse is a fibration. Conversely, it offers a geometric route to the canonical metrics of birational geometry. It is the setting where singularities of Ricci flow are best understood beyond dimension three.
Where it stands
Open problems include Gromov–Hausdorff convergence at every finite-time singularity, the structure of the flow near flips, collapsing along Fano fibrations, and the abundance conjecture, on which some convergence results depend. The feed below tracks all of them. See also Kähler–Ricci flow and Kähler–Einstein and K-stability.