FINITE-TIME BLOW-UP OF THE HEAT FLOW OF HARMONIC MAPS FROM SURFACES
Kung-Ching Chang,Wei Ding,Rugang Ye
DOI: https://doi.org/10.4310/jdg/1214448751
1992-01-01
Journal of Differential Geometry
Abstract:This note is concerned with the singular behaviour of solutions to the heat equation of harmonic maps between Riemannian manifolds.Eells and Sampson [5] showed that for C 1 initial values the solutions exist locally in time and asked whether they exist for all time or they may blow up in finite time.Recently, Coron-Ghidaglia [3], Ding [4] and Chen-Ding [2] constructed many examples of finite-time blow-up of the solutions in the case where the domain manifold has dimension greater than two.However, it has been widely believed that finite-time singularities would not occur in dimension two.Such a belief is seemingly supported by a number of phenomena special for dimension two.For example, it has been noted that even if a given smooth map from a surface belongs to a homotopy class containing no harmonic maps, the solution to the heat equation with the given map as initial value may exist globally (see [1] and [6]).Such a phenomenon is unknown in higher dimensions and may be viewed as an indication that probably all solutions exist globally in dimension two.Another phenomenon, observed first by R. Hamilton, is as follows.Suppose that for some point x 0 of the domain surface and t 0 > 0 one has e(t) = \\e(u(t, ))II C °W ^ Γ^7' f o r '<*(>>where u is a solution of the heat equation, e(u(t, •)) denotes the energy density of u at ί, and D δ is a small neighborhood of x Q .Then, using rescaling analysis (cf.[8]) it is not hard to show that (t Q , x Q ) is a regular point of the solution u.Roughly speaking, this implies that if a solution blows up at some finite time t 0 , then the velocity of the blow-up cannot be very high.Indeed, by letting λ(t) = l/e(t), the above inequality is equivalent to λ(t)>C-\t o -t) 9