Kobayashi-Royden pseudometric of contractible surfaces

Shulim Kaliman · 2025

Read Kobayashi-Royden pseudometric of contractible surfaces by Shulim Kaliman (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-06874.txt.

Introduction

A family of smooth contractible surfaces, unlike the plain plane, is described through a special gauge by Shulim Kaliman in a solo 2025 geometry paper. The Kobayashi-Royden pseudometric gauges how a complex surface resists holomorphic maps. Kaliman's surfaces admit dominant holomorphic maps from the squared complex plane, yet their geometry is tightly constrained. On each surface there is one line such that, away from it, the pseudometric vanishes nowhere on the tangent bundle.

Catalog record

CREATORShulim Kaliman
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE24 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-06874

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Kobayashi-Royden pseudometric of contractible surfaces Thanks: 2020 Mathematics Subject Classification Primary 14 Shulim Kaliman Abstract . We describe a family of smooth contractible algebraic surfaces X X different from ℂ 2 \mathbb{C}^{2} such that X X admits dominant holomorphic maps from ℂ 2 \mathbb{C}^{2} and there is a unique line E E in X X for which the Kobayashi-Royden pseudometric vanishes on the tangent bundle over X ∖ E X\setminus E . 1. Introduction All varieties in this paper are considered over the field of complex numbers. In 1971 Ramanujam [ Ra71 ] constructed the first smooth contractible affine algebraic surface different from \mathbb{C}^{2} . Such surfaces X are often called Ramanujam surfaces [ Za90 ] . Fujita [ Fu79 ] and Miyanishi and Sugie [ MiSu80 ] showed that X\times\mathbb{C}^{n} is not isomorphic to \mathbb{C}^{n+2} for every n>0 , thus, proving the Zariski cancellation conjecture in dimension 2. Later Fujita [ Fu82 ] developed an effective method of constructing Ramanujam surfaces. But his construction (as well as the Ramanujam original paper) produces only Ramanujam surfaces X with logarithmic Kodaira dimension \bar{\kappa}(X)=2 . Gurjar and Miyanishi [ GuMi88 ] discovered Ramanujam surfaces X with \bar{\kappa}(X)=1 and found their classification. They also showed that there is no Ramanujam surface X with \bar{\kappa}(X)=-\infty , whereas the result of Pełka and Raźny [ PeRa21 ] implies that X cannot have \bar{\kappa}(X)=0 (see Proposition 2.3 ). An elegant description of Ramanujam surfaces X with \bar{\kappa}(X)=1 was found later by Petrie and tom Dieck [ tDiPe90 ] (actually, all of them can be presented as explicit hypersurfaces in \mathbb{C}^{3} [ KaML96 ] ). On the other hand, the result of Choudary and Dimca [ ChDi94 ] implies that X\times\mathbb{C}^{n} is diffeomorphic to \mathbb{R}^{2n+4} for n\geq 1 . So, it is natural to ask whether X\times\mathbb{C}^{n} is biholomorphic to \mathbb{C}^{n+2} . A partial answer to this question is based on the paper of Sakai [ Sa77 ] who proved that every complex algebraic variety of general type is measure hyperbolic (that is, its top Eisenman measure is almost nowhere zero). Using this fact Zaidenberg [ Za90 ] , [ Za93 ] showed that X\times\mathbb{C}^{n} is not biholomorphic to \mathbb{C}^{n+2} when \bar{\kappa}(X)=2 . However, Ramanujam surfaces X with \bar{\kappa}(X)=1 are not measure hyperbolic and the question whether X\times\mathbb{C}^{n} is not biholomorphic to \mathbb{C}^{n+2} remains open. The answer to this question would be positive is X possesses a nontrivial Kobayashi-Royden pseudometric which is also the first Eisenman measure [ Ka94 ] (for the definition of Eisenman measures we refer to [ Ko76 ] and [ Ei70 ] ). In particular, one may hope that as in the case of the top Eisenman measures for varieties of general type the Kobayashi-Royden pseudometric is almost nowhere zero on tangent bundle TX . An unsuccessful attempt to prove this fact was the original motivation for our paper. Actually, we find a family of Ramanujam surfaces X with \bar{\kappa}(X)=1 and Kobayashi-Royden pseudometric vanishing on a dense open subset of TX (see Section 4 ). Furthermore, we show that there a holomorphic map \mathbb{C}^{2}\to X such that its Jacobian determinant is not identically zero (a surface admitting such map is called holomorphically dominable by \mathbb{C}^{2} ), whereas there is no dominant morphism from \mathbb{C}^{2} to X . The surfaces holomorphically dominable by \mathbb{C}^{2} were extensively studied by Buzzard and Lu [ BuLu00 ] . However, the families we discovered are not in their list. 2. Preliminaries Throughout this paper a notation of the form \mathbb{C}_{z_{1},\ldots,z_{n}}^{n} means \mathbb{C}^{n} equipped with a coordinate system (z_{1},\ldots,z_{n}) (in particular, \mathbb{C}_{z} is a line equipped with a coordinate z ). As usual, for a complex manifold X its tangent bundle is denoted by TX and T_{x}X is the tangent space at x\in X . Definition 2.1 . For r>0 let \Delta_{r}\subset\mathbb{C}_{z} be the disc of radius r with the center at the origin 0 and \operatorname{{\rm Hol}}(\Delta_{r},X) be the set of holomorphic maps from \Delta_{r} to X . Given a vector \nu\in T_{x}X its Kobayashi-Royden pseudometric is K_{X}(\nu)=\inf\{1/r\,|\,\varphi\in\operatorname{{\rm Hol}}(\Delta_{r},X)\text{ with }\varphi(0)=x\text{ and }{\rm d}\varphi\big(\frac{{\rm d}}{{\rm d}z}\big|_{0}\big)=\nu\}. In particular, if one has a holomorphic map \varphi:\mathbb{C}\to X with \varphi(0)=x and \varphi\big(\frac{{\rm d}}{{\rm d}z}\big|_{0}\big)=\nu , then K_{X}(\nu)=0 . Further in this paper every complex manifold X will be the complement to an effective divisor D=D_{1}+\ldots+D_{s} of simple normal crossing (SNC) type in a projective algebraic manifold {\bar{X}} . Let n=\dim X and \Omega_{n}(X) be the sheaf of germs of holomorphic n -forms on X . In particular, the space \Gamma(X,\Omega_{n}(X)) of holomorphic sections of \Omega_{n}(X) is the space of top holomorphic differential forms. For every point of {\bar{X}} there exists a local coordinate system (z,w)=(z_{1},\ldots,z_{l},w_{1},\ldots,w_{n-l}) such that z_{1}\cdots z_{l}=0 defines the germ of D around this point ( 0\leq l\leq n and when l=0 this point does not belong to D ). In this case \Omega_{n}(X) contains the subsheaf {\mathcal{L}}=\Omega_{n}(\bar{X},D) of logarithmic n -forms along D . The germs of these forms can be written as \sum_{r+q=k}c_{I,J}(z,w){\frac{dz_{i(1)}}{z_{i(1)}}}\wedge\ldots{\frac{dz_{i(r)}}{z_{i(r)}}}\wedge{\frac{dw_{j(1)}}{w_{j(1)}}}\wedge\ldots{\frac{dw_{j(q)}}{w_{j(q)}}}, where c_{I,J}(z,w) is the germ of a holomorphic function, and the indices of summation are I=(i(1),\ldots,i(r)) and J=(j(1),\ldots,j(q)) . Definition 2.2 . If for every m>0 the m th power {\mathcal{L}}^{m} of {\mathcal{L}} does not have a non-trivial global section, then the logarithmic Kodaira dimension of X is \bar{\kappa}(X)=-\infty . Otherwise, \bar{\kappa}(X)=\lim_{m\to+\infty}\sup{\frac{\log\dim\Gamma(X,{\mathcal{L}}^{m})}{\log m}}. The logarithmic Kodaira dimension does not depend of the choice of SNC completion {\bar{X}} of X and one also has \bar{\kappa}(X)\leq n ( [ Ii77 ] ). Every smooth contractible affine algebraic surface X is factorial [ Fu82 ] and the ring \mathbb{C}[X]^{*} of invertible functions on it consists of constants only. Gurjar and Miyanishi [ GuMi88 ] classified all smooth factorial surfaces X with \mathbb{C}[X]^{*}=\mathbb{C}^{*} and logarithmic Kodaira dimension equal to \bar{\kappa}(X)=1 (this yields, in particular a classification of Ramanujam surfaces with logarithmic Kodaira dimension 1, whereas for negative dimension such Ramanujam surfaces do not exist). However, their classification for the case \bar{\kappa}(X)=0 contains a flaw. Freudenburg, Kojima, and Nagamine [ FrKoNa19 ] found an infinite collection of smooth affine factorial surfaces X with \mathbb{C}[X]^{*}=\mathbb{C}^{*} and \bar{\kappa}(X)=0 that are not in the list presented in [ GuMi88 ] . Later Pełka and Raźny [ PeRa21 ] obtained a complete classification of such surfaces. The next proposition is an immediate consequence of the latter result. Proposition 2.3 . There are no Ramanujam surfaces X with \bar{\kappa}(X)=0 . Proof. Let X be a smooth affine factorial surface with \mathbb{C}[X]^{*}=\mathbb{C}^{*} and \bar{\kappa}(X)=0 . By [ PeRa21 , Theorem 1.1] there exist monic polynomials p_{1}(t) and p_{2}(t) such that X is isomorphic to the spectrum of the ring \mathbb{C}[x_{1},x_{2}][(x_{2}x_{1}^{-\mathop{\rm deg}p_{1}}-p_{1}(x_{1}^{-1}))x_{1}^{-1},(x_{1}x_{2}^{-\mathop{\rm deg}p_{2}}-p_{2}(x_{2}^{-1}))x_{2}^{-1}]. This implies that the restriction of the natural morphism \varphi:X\to\mathbb{C}^{2}_{x_{1},x_{2}} is isomorphism over the torus T=\{(x_{1},x_{2})\in\mathbb{C}^{2}|\,x_{1},x_{2}\neq 0\} , the image of \varphi is T\cup\{(1,0),(0,1)\} , and \varphi^{-1}((1,0))\simeq\varphi^{-1}((0,1))\simeq\mathbb{C} . Thus, X can be stratified into a disjoint union of T and two lines. Since the Euler characteristic of T is 0 and of \mathbb{C} is 1, the Euler characteristic of X is 2 by [ Du87 ] which yields the desired conclusion. ∎ 3. Ramification Let L be the curve in \mathbb{C}^{2}_{u,v} given by u^{k}-v^{l}=1 where k and l are fixed relatively prime natural numbers with k>l\geq 2 and L_{0}\subset\mathbb{C}^{2}_{u,v} be given by

Read the complete text