A105C87CB202
There is an integer $g_0$ such that for every even $g\geq g_0$ there exists a compact smooth embedded orientable surface $\Sigma _ g\subset \mathbb R^3$ of genus $g$ satisfying $\vec H=-X^\perp /2$. Its flow is $M_ t=\sqrt{-t}\, \Sigma _ g$, $t<0$. This is an existence construction by desingularization, not an explicit formula and not an existence assertion for every positive genus.
Each surface is complete and proper, has no ends, and is invariant under a dihedral group of order $2g$.
Along the constructed family as $g\to \infty $, the surfaces converge in Hausdorff distance to the union of a round shrinking sphere and a rotational shrinking torus. Convergence is locally smooth away from their two intersection circles ([Moller2014], pp. 1–2). The limiting union is singular and is not one of the smooth members.
These surfaces are not mean convex and are entropy unstable: apply [CM12], Theorems 0.17 and 0.12, to the compact embedded non-spherical surfaces. Their homothetic flows have Type I extinction at the origin.
The cited construction supplies no exact coordinates or exact entropy value and makes no uniqueness assertion for a specified genus.
The illustration follows Figures 1–2 of [Moller2014]: a sphere of radius $2$ intersects a non-circular rotational shrinking torus along two circles. In cylindrical coordinates $(r,\theta ,z)$, the displayed numerical torus meets the sphere near $r=1.783$, $z=\pm 0.906$, consistent with Proposition 2.1. These are numerical coordinates, not exact formulas. Rotation, a quarter cutaway, and a meridian view expose both sheets and the circles where the desingularization takes place. The removed sector is only a viewing aid. As in the paper's Figure 1, this is the singular configuration before handle insertion, not a computed smooth surface $\Sigma _ g$.
The second illustration shows how the two intersection circles can be replaced by handles, using gluing schematics of genera $8$, $16$, and $24$. Their mesh topology is checked, but these models do not solve the self-shrinker equation and are not certified approximations to $\Sigma _ g$. The chosen genera illustrate the topology; the cited theorem does not give an explicit $g_0$ that would establish these particular values as members of its constructed family.
Comments
Comments are not enabled yet.