1: \documentclass[10pt]{amsart}
2:
3: \title{The Crepant Resolution Conjecture}
4: \date{December 22nd, 2006}
5: \address{
6: Dept of Math, Univ. of British Columbia,
7: Vancouver, BC, Canada
8: }
9: \email{jbryan@math.ubc.ca}
10: \address{
11: Dept of Math, California Institute of Technology, Pasadena, CA
12: }
13: \email{graber@caltech.edu}
14:
15:
16:
17: \author{Jim Bryan and Tom Graber}
18:
19:
20:
21:
22:
23:
24:
25: %\usepackage{diagrams}
26: %\usepackage{eepic,epic}
27:
28: \usepackage{amsmath}
29: \usepackage{amsmath,amsthm,amsfonts}
30: %\usepackage{amstex}
31: \usepackage[all]{xypic}
32:
33: \DeclareMathOperator{\Aut}{Aut}
34: \DeclareMathOperator{\Hilb}{Hilb}
35: \DeclareMathOperator{\Sym}{Sym}
36: \DeclareMathOperator{\GHilb}{G-Hilb}
37:
38: \newcommand{\comment}[1]{}
39: \newcommand{\fz}{{\mathfrak{z}}}
40: \newcommand{\hZ}{\hat{Z}}
41: \newcommand{\E}{{\mathbb E}}
42: \newcommand{\C}{{\mathbb C}}
43: \newcommand{\Q}{{\mathbb Q}}
44: \newcommand{\cnums} {{\mathbb C}} % complex numbers
45: \newcommand{\nnums} {{\mathbb N}} % natural numbers
46: \newcommand{\rnums} {{\mathbb R}} % real numbers
47: \newcommand{\znums} {{\mathbb Z}} % integers
48: \newcommand{\Z}{{\mathbb Z}}
49: \newcommand{\qnums} {{\mathbb Q}} % rationals
50: \newcommand{\DD}{{\mathcal D}}
51: \newcommand{\bD}{{\bar{D}}}
52: \newcommand{\smcup}{\cup}
53: \newcommand{\smcap}{\cap}
54: \newcommand{\Hom}{\operatorname{Hom}}
55: \newcommand{\Ker}{\operatorname{Ker}}
56: \newcommand{\End}{\operatorname{End}}
57: \newcommand{\Tr}{\operatorname{tr}}
58: \newcommand{\tr}{\operatorname{tr}}
59: \newcommand{\Coker}{\operatorname{Coker}}
60: \newcommand{\im}{\operatorname{Im}}
61: \newcommand{\sym}{\operatorname{Sym}}
62: \newcommand{\hilb}{\operatorname{Hilb}}
63: \newcommand{\point}{\mathrm{pt}}
64: \newcommand{\vertline}{\operatorname{|}}
65:
66:
67: \newcommand{\cK}{{\mathcal{K}}}
68: \newcommand{\cC}{{\mathcal{C}}}
69: \newcommand{\oh}{{\mathcal{O}}}
70: \renewcommand{\O}{\oh}
71: \newcommand{\proj}{{\mathbb{P}}}
72: \renewcommand{\P}{\mathbb{P}}
73: \newcommand{\M}{\overline{{M}}}
74: \newcommand{\smargin}[1]{\marginpar{\tiny{#1}}}
75: \newcommand{\X}{\mathcal{X}}
76: \newcommand{\orb}{\mathit{orb}}
77: \newcommand{\tw}{\mathit{tw}}
78: \newcommand{\vir}{{\mathrm{vir}}}
79: \newcommand{\IX}{{I\X}}
80:
81: \newcommand{\NS}{\widehat{NS}}
82: \newcommand{\hbeta}{{\widehat{\beta}}}
83: \newcommand{\hF}{{\widehat{F}}}
84:
85: \newcommand{\Hn}{\hilb^n(\C^2)}
86: \newcommand{\bC}{\bar{C}}
87: \newcommand{\barf}{\bar{f}}
88: \newcommand{\hC}{\hat{C}}
89: \newcommand{\barH}{{\overline{H}}}
90: \newcommand{\Irr}{\operatorname{Irr}}
91: \newcommand{\Conj}{\operatorname{Conj}}
92: \newcommand{\age}{\operatorname{age}}
93:
94:
95: \newtheorem{thm}{Theorem}[section]
96: \newtheorem{hyp}{Hypothesis}[section]
97: \newtheorem{theorem}[thm]{Theorem}
98: \newtheorem{cor}[thm]{Corollary}
99: \newtheorem{lem}[thm]{Lemma}
100: \newtheorem{lemma}[thm]{Lemma}
101: \newtheorem{prop}[thm]{Proposition}
102: \newtheorem{proposition}[thm]{Proposition}
103: \newtheorem{ax}{Axiom}
104: \newtheorem{defn}[thm]{Definition}
105: \newtheorem{definition}[thm]{Definition}
106: \newtheorem{warning}[thm]{Warning}
107: \newtheorem{questions}{Question}
108: \newtheorem{conjecture}[thm]{Conjecture}
109: \newtheorem{conjecture*}{Conjecture}
110: \newtheorem{assumption}{Assumption}
111: %\newtheorem{remark}[thm]{Remark}
112:
113:
114: % There has got to be a better way of making Remarks non italicized,
115: % but the following seems to work:
116:
117: \newtheorem{rem1}[thm]{Remark}
118: \newenvironment{rem}{\begin{rem1}\em}{\end{rem1}}
119: \newenvironment{remark}{\begin{rem1}\em}{\end{rem1}}
120:
121:
122:
123:
124:
125: \newcommand{\bpf}{\mathsc{Proof:}}
126: \newcommand{\epf}{\qed}
127:
128: \begin{document}
129: \maketitle
130:
131: \begin{abstract}
132: For orbifolds admitting a crepant resolution and satisfying a hard
133: Lefschetz condition, we formulate a conjectural equivalence between
134: the Gromov-Witten theories of the orbifold and the resolution. We
135: prove the conjecture for the equivariant Gromov-Witten theories of
136: $\sym ^{n}\cnums ^{2}$ and $\hilb ^{n}\cnums ^{2}$.
137: \end{abstract}
138: %\thanks{The first author \dots The second author is supported by an
139: %Alfred P. Sloan Research Fellowship and NSF grant .}
140:
141: %\markboth{???} {???}
142: %\renewcommand{\sectionmark}[1]{}
143:
144:
145: %\tableofcontents
146: %\pagebreak
147:
148:
149: \section{Introduction}
150:
151: \subsection{Overview} Gromov-Witten theory is the mathematical
152: counterpart of topological string theory in physics. A well known
153: principle in physics states that string theory on an orbifold is
154: equivalent to string theory on a crepant resolution
155: \cite{Vafa-orbifold-numbers,Zaslow}. In their ground breaking paper
156: \cite{Chen-Ruan}, Chen and Ruan define orbifold cohomology using an
157: orbifold version of Gromov-Witten theory. Orbifold Gromov-Witten
158: theory was developed in the algebro-geometric context in
159: \cite{Ab-Gr-Vi-2,Ab-Vi,Ab-Gr-Vi} using Abramovich and Vistoli's notion
160: of twisted stable maps to a Deligne-Mumford stack.
161:
162: The Gromov-Witten invariants of a projective manifold $Y$ are
163: multilinear functions $\left\langle \dotsb \right\rangle_{g,\beta
164: }^{Y}$ on its cohomology $H^{*} (Y)$. The Gromov-Witten invariants of
165: a orbifold $\X$ are multilinear functions $\left\langle \dotsb
166: \right\rangle_{g,\beta }^{\X }$ of the orbifold cohomology
167: $H^{*}_{\orb} (\X)$. Orbifold cohomology is by definition the ordinary
168: cohomology of the inertia stack $I\X $ with a shifted grading
169: \cite{Ab-Gr-Vi,Chen-Ruan}. $H_{\orb }^{*} (\X )$ contains the usual
170: cohomology $H^*(X)$ as a subspace and its orthogonal complement is
171: referred to as the space of twisted sectors. If $\X $ is a Gorenstein
172: orbifold whose coarse moduli scheme $X$ admits a crepant resolution $Y
173: \to X$, Yasuda has proven that $H^{*}_{\orb} (\X ,\C)$ and $H^{*} (Y,
174: \C)$ are isomorphic as graded vector spaces \cite{Yasuda}. Yasuda's
175: proof provides an equality of Betti numbers but does \emph{not}
176: provide any natural choice of isomorphism. Nevertheless, it has been
177: suggested by Ruan \cite{Ruan-crepant} that there should be such an
178: isomorphism which identifies the Gromov-Witten theories. He proposes
179: that specializing the values of certain quantum parameters of the
180: small quantum cohomology of the resolution will recover the orbifold
181: cohomology of the orbifold. In this paper, we formulate an analogous
182: conjecture at the level of the full genus zero quantum potentials, and
183: we explore its consequences. In particular, we show this conjecture
184: allows one to essentially recover the Gromov-Witten theory of the
185: resolution in terms of that of the orbifold. We confirm the validity
186: of our conjecture for some examples including the case of $\X =\sym
187: ^{n}\cnums ^{2}$, $Y=\hilb^{n}\cnums ^{2} $.
188:
189: Recent work of Coates, Corti, Iritani, and Tseng, \cite{CCIT} strongly
190: suggests that for orbifolds failing the hard Lefschetz condition, the
191: relationship between the Gromov-Witten theories of the orbifold and
192: its resolution is more complicated and is better expressed in the more
193: sophisticated framework of Givental's Lagrangian formalism. We are
194: grateful to them for bringing the hard Lefschetz condition to our
195: attention.
196:
197: %\begin{hyp}\label{age condition} Every component $A \subset I\X$ of
198: %the inertia stack satisfies
199: %\[
200: %\age(A)=\frac12(\dim(\X)-\dim(A)).
201: %\]
202: %\end{hyp}
203: %It is easy to verify that this condition is always satisfied by
204: %orbifolds with a holomorphic symplectic form.
205:
206: \subsection{Notation} By an \emph{orbifold}, we will mean a smooth
207: algebraic Deligne-Mumford stack over $\cnums $. An orbifold $\X $ is
208: said to be \emph{Gorenstein} if $\X $ has generically trivial
209: stabilizers and the canonical bundle of $\X $ pulls back from a line
210: bundle on the coarse moduli space $X$ (equivalently, for every $x\in
211: \X $, the action of the isotropy group on the canonical line bundle is
212: trivial). A resolution of singularities $\pi :Y\to X$ is called
213: \emph{crepant} if $K_{Y}=\pi ^{*}K_{X}$.
214:
215: Let $\X $ be a Gorenstein orbifold and let
216: \[
217: \pi :Y\to X
218: \]
219: be a crepant resolution of the coarse moduli space $X$.
220: \comment{For
221: simplicity of exposition, we will assume $\X $ and $Y$ have no odd
222: cohomology.} We say that an integer basis for the second homology group
223: of a variety is \emph{positive} if the cone generated by the basis
224: contains the Mori cone.
225:
226: Let $\{\beta_1, \ldots , \beta_r \}$ be a positive basis of $H_2(Y)$
227: such that $\{\beta_{s+1}, \ldots ,\beta_r \}$ is a basis for the
228: kernel of $\pi _{*}:H_2(Y) \to H_2(X)$. Note that $\{\pi _{*}\beta
229: _{1},\dotsc ,\pi _{*}\beta _{s} \}$ is a positive basis for $H_{2}
230: (X)$. We choose a basis $\{\gamma_0, \ldots, \gamma_a\}$ for
231: $H^*(Y)$, with $\gamma_0=1$ and $\gamma_1, \ldots, \gamma_r$ the basis
232: for $H^2(Y)$ dual to the $\beta_i$.
233:
234:
235: The genus zero Gromov-Witten invariants of $Y$ are multilinear
236: functions $\left\langle \dotsb \right\rangle_{\beta }^{Y}$ on $H^{*}
237: (Y)$, defined by cohomological evaluations against $[\M _{0,n}
238: (Y,\beta )]^{\vir }$, the virtual fundamental class of the moduli space
239: of stable maps \cite{coxkatz,mirrorbook}. The invariants are encoded
240: in the potential function,
241: \[
242: F^{Y} (y_{0},\dotsc ,y_{a},q_{1},\dotsc ,q_{r}) =\sum _{n_0,\ldots , n_a=0}^{\infty }\sum _{\beta } \left\langle
243: \gamma_0^{n_0}\cdots\gamma_a^{n_a} \right\rangle_{\beta }^{Y}
244: \frac{y_0^{n_0}}{n_0!}\cdots\frac{y_a^{n_a}}{n_a!}q_1^{d_1 }\cdots
245: q_r^{d_r}
246: \]
247: where $\beta =d_{1}\beta _{1}+\dotsb +d_{r}\beta _{r}$ is summed over
248: all non-negative $(d_{1},\dotsc ,d_{r})$.
249:
250: Similarly, the genus zero Gromov-Witten invariants of $\X $ are
251: multilinear functions $\left\langle \dotsb \right\rangle_{\beta }^{\X
252: }$ on $H^{*} _{\orb }(\X )$, defined by cohomological evaluations
253: against $[\M _{0,n} (\X ,\beta )]^{\vir }$, the virtual fundamental
254: class of the moduli space of \emph{twisted} stable maps
255: \cite{Ab-Gr-Vi}. We choose a basis $\{\delta _{0},\dotsc ,\delta _{a}
256: \}$ for $H^{*}_{\orb } (\X )$ and we define the potential function for
257: $\X $:
258: \[
259: F^{\X} (x_{0},\dotsc ,x_{a},u_{1},\dotsc ,u_{s}) =\sum
260: _{n_0,\ldots , n_a=0}^{\infty }\sum _{\beta } \left\langle
261: \delta_0^{n_0}\cdots\delta_a^{n_a} \right\rangle_{\beta }^{\X}
262: \frac{x_0^{n_0}}{n_0!}\cdots\frac{x_a^{n_a}}{n_a!}u_1^{d_1 }\cdots
263: u_s^{d_s}
264: \]
265: where $\beta =d_{1}\pi _{*}\beta _{1}+\dotsb +d_{s}\pi _{*}\beta _{s}$
266: is summed over all non-negative $(d_{1},\dotsc ,d_{s})$.
267:
268: The inertia stack $I\X $ of an orbifold $\X $ is defined to be the
269: fibered product of $\X $ with itself over the diagonal in $\X \times
270: \X $. The points of $I\X $ are pairs $(x,g)$ where $x\in \X $ and
271: $g\in \Aut _{\X } (x)$. There is an involution $I$ of $I\X $ taking
272: $(x,g)$ to $(x,g^{-1})$. To each component $\X _{i}$ of $I\X $ we
273: assign a rational number $\age (\X _{i})$ as follows. Let $(x,g)$ be a
274: point in $\X _{i}$. Then $g$ acts on $T_{x}\X $ with eigenvalues
275: $(\alpha _{1},\dotsc ,\alpha _{n})$ where $n=\dim \X $. Let $r$ be the
276: order of $g$ and define $s_{j}\in {0,\dotsc ,r-1}$ by $\alpha
277: _{j}=\exp (2\pi i\frac{s_{j}}{r})$. Then age is defined by
278: \[
279: \age (\X _{i}) = \frac{1}{r}\sum _{j=1}^{n}s_{j}.
280: \]
281: Age is well defined and is integral for Gorenstein orbifolds.
282:
283: As a graded vector space, the orbifold cohomology of $\X $ is the
284: cohomology of $I\X $ with the grading shifted by twice the age:
285: \[
286: H^{*}_{\orb } (\X ) = \bigoplus _{\X _{i}\subset I\X }H^{*+2\age (\X _{i})} (\X _{i}).
287: \]
288:
289: Suppose that the coarse moduli space $X$ is projective with hyperplane
290: class $\omega $. In \cite{Fernandez}, Fernandez asked if the hard
291: Lefschetz isomorphism holds in orbifold cohomology, namely if the
292: operator $L_{\omega }$ given by multiplication by $\omega $ in the
293: \emph{orbifold cohomology ring}, induces isomorphisms
294: \[
295: L^{p}_{\omega }:H^{n-p}_{\orb } (\X )\to H^{n+p}_{\orb } (\X ).
296: \]
297: Fernandez proved that $L^{p}_{\omega }$ is an isomorphism for all
298: $\omega $ if and only if the age is invariant under the involution
299: $I$. We call this condition (also defined for non-projective
300: orbifolds) the hard Lefschetz condition.
301: \begin{defn}\label{defn: hard lefschetz}
302: An orbifold $\X $ is said to satisfy the hard Lefschetz condition if the involution
303: \[
304: I:I\X \to I\X
305: \]
306: preserves the age.
307: \end{defn}
308:
309: Note that this condition is satisfied by holomorphic symplectic orbifolds.
310:
311: \subsection{The Conjecture}
312:
313: Our main conjecture relates the two
314: potential functions $F^{Y}$ and $F^{\X }$.
315:
316:
317: \begin{conjecture}[Crepant Resolution Conjecture]\label{mainconjecture}
318: Given an orbifold $\X$ satisfying the hard Lefschetz condition and
319: admitting a crepant resolution $Y$, there exists a graded linear
320: isomorphism
321: \[
322: L : H^*(Y) \to
323: H_\orb^*(\X)
324: \]
325: and roots of unity $c_{s+1}, \ldots , c_r $ such that the
326: following conditions hold.
327: \begin{enumerate}
328: \item The inverse of $L$ extends the map $\pi ^{*}: H^*(\X)\to H^{*}
329: (Y)$.
330: \item Regarding the potential function $F^Y$ as a power series in
331: $y_{0},\dotsc ,y_{a},q_{1},\dotsc ,q_{s}$, the coefficients
332: admit analytic continuations from $(q_{s+1},\dotsc ,q_{r}) =
333: (0,\dotsc ,0)$ to $(q_{s+1},\dotsc ,q_{r}) = (c_{s+1},\dotsc
334: ,c_{r})$.
335: \item The potential functions $F^\X$ and $F^{Y}$ are equal after
336: the substitution
337: \begin{align*}
338: y_i &= \sum _{j} L^j_ix_j\\
339: q_i &= \begin{cases} c_i & \text{when } i> s\\
340: u_i & \text{when } i\leq s. \end{cases}
341: \end{align*}
342: \end{enumerate}
343: \end{conjecture}
344:
345:
346: \begin{remark}\label{rem: FY and FX contain equiv info}
347: The cohomological parameters in the potential functions,
348: $\{x_{0},\dotsc x_{a} \}$ and $\{y_{0},\dotsc ,y_{a} \}$ are equal in
349: number by Yasuda's result. However, the number of quantum parameters,
350: $\{u_{1},\dotsc ,u_{s} \}$ and $\{q_{1},\dotsc ,q_{r} \}$ differ, and
351: so na\"ively, the potential function $F^{Y}$ appears to have more
352: information than $F^{\X }$. However, the divisor equation implies
353: that the potential function $F^{Y}$ contains redundant information. In
354: fact, given $L$, $c_{s+1},\dotsc ,c_{r}$, and $F^{\X }$, one can
355: essentially recover $F^{Y}$. This will be made more clear in
356: section~\ref{sec: degree in twisted sectors and QH}, where we will
357: present an alternative but equivalent formulation of the conjecture
358: which is particularly convenient when studying the small quantum
359: cohomology ring. Rather than resolving the difference in the number
360: of $q$'s and $u$'s by setting some of the $q$'s to constants, one can
361: adjoin extra $u$ variables to the orbifold partition function, by
362: defining a generalized notion of degree for orbifold curves with
363: unmarked twisted points.
364: \end{remark}
365:
366: \comment{\begin{remark}
367: The analytic continuation condition above is implicitly assumed in the
368: physical literature on the subject, and it is certainly reasonable to
369: believe that this even stronger convergence properties are true for
370: the potential functions in general, outside the context of crepant
371: resolutions of orbifolds.
372: \end{remark}}
373:
374: \begin{remark}\label{rem: unstable terms}
375: A finite set of coefficients in these potential functions are not well
376: defined since certain degenerate moduli spaces do not exist. Namely,
377: terms of degree zero in the quantum parameters and of degree less than
378: three in the cohomological variables are undefined. We are not making
379: any conjectures about these coefficients. To get a precise equality,
380: one needs to either take triple derivatives of the series on both
381: sides, or choose compatible assignments of values to the unstable
382: coefficients. It would be interesting to find a meaningful way of
383: defining these unstable invariants.
384: \end{remark}
385:
386: \begin{remark}\label{rem: poincare pairing is preserved by L}
387: It is a consequence of the conjecture that the linear map $L$ must
388: preserve the (orbifold) Poincar\'e pairing.
389: \end{remark}
390:
391:
392: \begin{remark}\label{rem: can add descendants etc to conj}
393: If $\X$ admits an action of an algebraic torus $T$ and $Y$ is a
394: $T$-equivariant crepant resolution, then we can extend the conjecture
395: to include equivariant parameters. In fact, this equivariant
396: version of the conjecture follows immediately from the absolute
397: version, by considering the conjecture applied to finite dimensional
398: approximations to the homotopy quotients $Y_{T}\to X_{T}$.
399: \end{remark}
400:
401:
402:
403: \begin{remark}\label{rem: galois automorphisms}
404: The coefficients of $F^{Y}$ and $F^{\X }$ are rational
405: numbers, but in general the linear transformation $L$ may be have to
406: be defined over some extension of $\qnums $. A consequence of the
407: conjecture is that there is a symmetry of $F^{Y}$ given by the
408: action of the Galois group of the extension on the change of
409: variables. In practice, this is often a highly non-trivial symmetry.
410: \end{remark}
411:
412: \begin{remark}\label{rem: CRC in higher genus}
413: Our conjecture may also hold as stated for higher genus
414: potentials. There is very little evidence in positive
415: genus, although Maulik's computation of the full Gromov-Witten potential for
416: $A_{n}$ surface resolution \cite{Maulik-An} does provide some positive
417: evidence. The relationship of the higher genus Gromov-Witten
418: potentials for Gorenstein orbifolds failing the hard Lefschetz
419: condition is expected to be more complicated involving a mixing of
420: different genera. For a physical account, see
421: \cite{Aganagic-Bouchard-Klemm}. See also the discussion in section~5
422: of \cite{CCIT}.
423: \end{remark}
424:
425:
426: \subsection{The noncompact case}\label{subsec: noncompact case}
427:
428: Although Gromov-Witten theory is best known in the compact setting,
429: the simple examples we want to focus on are non-compact, so we
430: observe that there is a large class of noncompact examples where
431: there is a well defined version of the conjecture.
432:
433: The most convenient hypothesis here is to assume that $X$ is
434: projective over an affine scheme and that $Y$ is projective over $X$
435: and hence also projective over an affine. (In fact, in our examples
436: $X$ will actually be affine.) In this setting, although the spaces
437: of stable (twisted) maps need not be proper, the evaluation maps
438: from the space of maps to $Y$ (or $I\X$) will be proper. Thus we
439: have well-defined Gromov-Witten classes
440: \[
441: \langle \gamma_1, \ldots , \gamma_n, *\rangle_\beta
442: \]
443: defined as in \cite{Ab-Gr-Vi-2} by pushing forward from the space of $n+1$
444: pointed (twisted) stable maps to $Y$ (or $I\X$).
445:
446: If the target is projective, then because of the formula
447: \[
448: \langle
449: \gamma_1, \ldots , \gamma_{n+1}\rangle_{\beta} =
450: \gamma_{n+1} (\langle \gamma_1, \ldots, \gamma_n, *\rangle_{\beta})
451: \]
452: these homology
453: valued invariants contain equivalent information to the numerical
454: Gromov-Witten invariants, and moreover, the conjecture as stated
455: implies immediately a conjecture for a generating series of
456: homology valued invariants. In the noncompact setting, where one
457: cannot reduce the homology classes to numbers in this way, we can
458: then use these invariants to make a meaningful version of the
459: conjecture.
460:
461: In fact, we will not need to pursue a careful language for these
462: refined invariants, since our examples have another useful feature.
463: They all admit a torus action with compact fixed locus. Because of
464: this, there is a perfect pairing on the $T$-equivariant cohomology
465: given by formally applying the Bott residue formula. While this
466: pairing takes values in $H_T^*(pt, \C)_{\mathfrak m}$, rather than
467: $\C$, it still allows us to do calculations at the level of the
468: familiar generating functions for numerical Gromov-Witten invariants
469: with the slight novelty that some of these numbers will be rational
470: functions in the equivariant parameters.
471:
472: \comment{A further advantage of this
473: equivariant case is that standard facts about equivariant cohomology
474: make it easy to conclude that if $\X$ admits an equivariant
475: compactification $\overline\X$ with equivariant crepant resolution
476: $\overline Y$ extending $Y$, then the conjecture for $\overline Y \to
477: \overline X$ implies this conjecture for $Y\to X$. Since it is hard
478: to believe that Gromov-Witten theory would detect an obstruction to
479: the existence of such a resolution of singularities, it seems
480: reasonable to assume that if the conjecture holds for all projective
481: $X$, it will also hold for this nice class of quasi-projective
482: examples.}
483:
484: \section{Degree in twisted sectors and Quantum cohomology}\label{sec: degree in twisted sectors and QH}
485:
486:
487: In this section we extend the definition of $\left\langle \dotsb
488: \right\rangle^{\X }_{g,\beta }$ to allow for $\beta $ to be a ``curve
489: class'' in the twisted sector. Consequently, the corresponding
490: Gromov-Witten potential of $\X $ includes quantum parameters
491: corresponding to twisted sectors. This allows us to formulate an
492: alternative version of the Crepant Resolution Conjecture where the
493: number of variables for $\X $ and for $Y$ are the same. In particular,
494: the large and small quantum cohomology rings of $Y$ and $\X $ have the
495: same number of deformation parameters and are isomorphic (in a certain
496: sense -- see subsection~\ref{subsec: orbifold quantum cohomology})
497: when the Crepant Resolution Conjecture holds.
498:
499:
500:
501:
502: \subsection{The orbifold Neron-Severi group and twisted degrees}
503:
504: We define an enlarged Neron-Severi group for a Gorenstein orbifold
505: $\X$ as follows. Let $T^1(\X)$ be the twisted part of
506: $H^2_{orb}(\X,\Z)$. As this is generated by fundamental classes of
507: certain irreducible components of the inertia stack, it comes with a
508: canonical (unordered) basis and is a free Abelian group of rank $r-s$.
509:
510: \begin{defn}
511: We define the \emph{orbifold Neron-Severi group} $\NS _1(\X) $ by
512: \[
513: \NS _1(\X) = NS_1(\X) \oplus T^1(\X)^\vee.
514: \]
515: That is, an element $\hbeta \in\NS _1(\X)$ is a curve class $\beta $
516: in $\X$ together with a function $\hbeta (i)$ assigning an integer to
517: each age one component of $I\X$. An
518: element of $\NS _1(\X)$ will be considered {\em effective} if the
519: underlying curve class is effective, and the function is nonnegative.
520: \end{defn}
521:
522:
523: Recall that evaluation at the $i$th point of a twisted stable map
524: takes values in the inertia stack and defines a virtual morphism
525: $e_{i}:\M _{g,n} (\X ,\beta )\to I\X $. (We are using different
526: conventions here than those of \cite{Ab-Gr-Vi} or \cite{Ab-Gr-Vi-2}
527: --- our $\M$ corresponds to $\cK$ and our $e_{i}$ corresponds to
528: $\tilde{e}_{i} $ of Proposition~6.1.4 of \cite{Ab-Gr-Vi-2}.)
529:
530: \begin{defn}\label{defn: moduli space for twisted degrees}
531: Given an effective class $\hat \beta \in \NS_1(\X)$, we define
532: $\M_{g,n}(\X,\hat\beta)$ to be the moduli space parameterizing genus
533: $g$ twisted stable maps to $\X$ with degree $\beta$ with $n$ ordered
534: marked points and with $\hat\beta(i)$ \emph{unordered} twisted points which
535: map to $D_{i}$, the $i$th component of the inertia stack. Precisely,
536: if we consider the following fiber product:
537: $$\xymatrix{
538: \M \ar[r] \ar[d] &
539: \overline{I}\X^n \times \overline{D}_1^{\hbeta(1)} \times \cdots \times \overline{D}_{r-s}^{\hbeta(r-s)} \ar[d]\\
540: \M_{g,n+\sum \hat\beta(i)}(\X,\beta)\ar[r] & \overline{I}\X^n \times
541: \overline{I}\X^{\hbeta(1)} \times \cdots \times
542: \overline{I}\X^{\hbeta(r-s)}}$$ then we define $\M_{g,n}(\X,\hbeta)$
543: to be the quotient $[\M/S_{\hbeta(1)} \times \cdots \times
544: S_{\hbeta(r-s)}].$ Here $\overline{I}\X $ is the rigidified stack and
545: $\overline{D}_{i}$ is the $i$th component.
546: \end{defn}
547:
548: \begin{remark}\label{rem: counting interpretation of degree in twisted sector}
549: One interpretation of the usual degree is as counting the number of
550: times a curve intersects some fixed divisor. Similarly, we can
551: interpret the degree in the twisted sector as counting the number of
552: times some curve ``intersects'' some twisted divisor, namely it gives
553: the number of (unmarked, non-nodal) stacky points that get mapped to
554: the corresponding age one component of the inertia stack. The reason
555: for not including nodal stacky points in the count is so that the
556: degree will be locally constant in families. We ignore the marked
557: points so that degree is additive when gluing smooth curves together
558: to form nodal ones, and therefore the boundary of the moduli spaces
559: $\M _{g,n} (\X ,\hbeta )$ have a product description analogous to the
560: usual one for the ordinary stable map moduli spaces.
561: \end{remark}
562:
563:
564: \subsection{Gromov-Witten invariants for degrees in twisted sectors
565: and the divisor equation}
566:
567: Since the right hand vertical arrow of the diagram in
568: Definition~\ref{defn: moduli space for twisted degrees} is simply an
569: inclusion of a union of connected components, so is the left hand
570: vertical arrow, which means that the perfect obstruction theory and
571: virtual fundamental class for the usual space of twisted stable maps
572: immediately give one on $\M$, and by descent, we get a virtual
573: fundamental class on $\M_{g,n}(\X,\hbeta)$.
574:
575: We can thus define Gromov-Witten invariants for curves with degrees
576: defined in the twisted sectors using these moduli
577: spaces. Correspondingly, we define the genus zero \emph{extended
578: Gromov-Witten potential} of $\X $ by
579: \begin{align*}
580: \hF^{\X } (x_{0},\dotsc ,x_{a},u_{1},\dotsc ,u_{r}) &=\\
581: \sum
582: _{n_0,\ldots , n_a=0}^{\infty }\sum _{\hbeta }& \left\langle
583: \delta_0^{n_0}\cdots\delta_a^{n_k} \right\rangle_{\hbeta }^{\X}
584: \frac{x_0^{n_0}}{n_0!}\cdots\frac{x_a^{n_a}}{n_a!}u_1^{d_1 }\cdots
585: u_s^{d_s}u_{s+1}^{\hbeta (1)}\dotsb u_{r}^{\hbeta (r-s)}
586: \end{align*}
587:
588: The extended invariants do not contain any new information, since we
589: have the following obvious formula:
590: \begin{equation}\label{eqn: div eqn for twisted divisors}
591: \langle D_1^{\hbeta(1)} \cdots D_{r-s}^{\hbeta(r-s)}\alpha _1 \cdots
592: \alpha _{n} \rangle_{\beta }^{\X } = \hbeta(1)!\dotsb \hbeta (r-s)! \cdot
593: \langle \alpha_1 \cdots \alpha_n\rangle_{\hbeta }^{\X }
594: \end{equation}
595: which immediately reduces the calculation of these ``new'' invariants
596: to the calculation of the standard orbifold invariants.
597: We think of this as the analog of the divisor equation for the
598: ``twisted divisors'' $D_i$, since it formally allows us to remove
599: the $D_i$ from invariants. Note, however, that this equation
600: is different in form from the usual divisor equation.
601:
602: It is useful to see what the divisor equation tells us about the form
603: of the potential function. For $Y$, it is well known that repeated
604: application of the divisor equation implies that (up to unstable
605: terms) we have
606:
607:
608: \comment{
609: \[
610: F^Y = \sum _{n_{r+1},\ldots , n_a=0}^{\infty
611: }\sum _{\beta } \left\langle
612: \gamma_{r+1}^{n_{r+1}}\cdots\gamma_a^{n_k} \right\rangle_{g,\beta }^{Y}
613: \frac{y_{r+1}^{n_{r+1}}}{n_{r+1}!}\cdots\frac{y_a^{n_a}}{n_a!}
614: \left(q_{1}e^{y_{1}} \right)^{d_{1}}\dotsb \left(q_{r}e^{y_{r}} \right)^{d_{r}}
615: \]}
616:
617: \[
618: F^Y=F^Y(y_0,0,0,\ldots, 0,y_{r+1},\ldots,y_{a},q_1e^{y_1},\ldots , q_re^{y_r}).
619: \]
620:
621:
622: In other words, the potential function depends on the variables in the
623: combinations
624: \[
625: q_1e^{y_1},\dotsc ,
626: q_re^{y_r}, y_{r+1},\dotsc , y_n.
627: \]
628: We can apply this to only the exceptional classes, giving the form more useful to us here:
629: \[
630: F^Y=F^Y(y_0,\ldots, y_s,0,\ldots, 0,y_{r+1},\ldots,y_{a},q_1,\ldots ,
631: q_s,q_{s+1}e^{y_{s+1}},\ldots, q_re^{y_r})
632: \]
633: For the orbifold invariants, the analogous result is that the extended
634: potential function depends on the variables only in the combinations
635: \[
636: u_1e^{x_1},\dotsc , u_se^{x_s}, (u_{s+1} + x_{s+1}),\dotsc ,
637: (u_r+x_r), x_{r+1},\dotsc ,x_n.
638: \]
639: More precisely, equation~\eqref{eqn: div eqn for twisted divisors}
640: implies the identity
641:
642: \[
643: \hF^\X= F^\X(x_0 , \ldots , x_s, (x_{s+1}+u_{s+1}),
644: \ldots , (x_r+u_r), x_{r+1}, \ldots , x_n,u_{1},\dotsc ,u_{s}).
645: \]
646:
647: So, assuming Conjecture \ref{mainconjecture}, we see that we get the
648: equality $\hF^\X = F^Y$ for the extended potential function after the
649: change of variables:
650:
651: \begin{align*}
652: y_i &= \sum _{j} L^j_ix_j\\
653: q_i &= \begin{cases} c_ie^{L_i^ju_j} & \text{when } i> s\\
654: u_i & \text{when } i\leq s. \end{cases}
655: \end{align*}
656:
657: Since this change of variables is invertible up to the discrete
658: choices of branches of certain logarithms, it shows that one can
659: essentially recover the Gromov-Witten theory of $Y$ from that of $\X$.
660: Moreover, in this form it is especially clear that the existence of
661: the standard divisor equation on $Y$ gives a very strong and
662: mysterious prediction about the potential for $\X$ -- it should depend
663: on the new $u$ variables (or equivalently some of the original $x$
664: variables) only in terms of certain exponentials.
665:
666: \subsection{Orbifold Quantum Cohomology}\label{subsec: orbifold
667: quantum cohomology} Another application of the above formalism is to
668: define a quantum product for an orbifold that is equivalent to the
669: quantum product of its crepant resolution by a method completely
670: parallel to the usual definition. We will discuss here only the small
671: quantum cohomology. Of course, one can use the derivatives of the
672: genus zero potential function to define a big quantum cohomology ring
673: for orbifolds and everything we say can be applied there as well.
674:
675: Assume $\X$ is a Gorenstein orbifold with projective coarse moduli scheme.
676: We consider the three evaluation maps from $\M_{0,3}(\X,\hbeta)$ to
677: $\IX$, and given classes $\delta$ and $\gamma$ in $H^*(\IX)$, we
678: define
679: \[
680: \delta * \gamma= \sum_\hbeta \left(\langle \delta, \gamma, *\rangle_{\hbeta }^{\X }
681: \right)^{\vee }u^\hbeta
682: \]
683: where $(\cdot )^{\vee }$ denotes dual with respect to the orbifold
684: Poincar\'e pairing. The same proof of
685: associativity holds for this product as for the one considered in
686: \cite{Ab-Gr-Vi}.
687:
688: We can express the quantum product in a basis, using the orbifold
689: Poincar\'e pairing $ g_{ij}$ on $H^{*} (\IX )$ as
690: \[
691: \gamma * \delta = \sum \langle \gamma, \delta, \gamma_i
692: \rangle_{\hbeta}\, g^{ij}\gamma_j u^\hbeta.
693: \]
694:
695:
696: Hence, it is an immediate consequence of
697: Conjecture~\ref{mainconjecture} that the products agree in the sense
698: that if we identify $H^*(Y)$ and $H^*(I\X)$ using $L$, then the
699: structure constants for the quantum product are related by the change
700: of variables:
701: \begin{align*}
702: q_i &\mapsto \begin{cases} c_ie^{L_i^ju_j} & \text{when } i> s\\
703: u_i & \text{when } i\leq s. \end{cases}
704: \end{align*}
705:
706: \begin{remark}As in subsection~\ref{subsec: noncompact case}, this
707: definition of the quantum product makes sense using only the
708: hypothesis that $X$ is projective over an affine scheme. The argument
709: reducing the equivalence of the quantum products of $\X$ and $Y$ to
710: the equivalence of the potential functions uses the perfectness of the
711: Poincar\'e pairing, which we do have in the torus equivariant
712: setting provided that the fixed locus is compact.
713:
714: \end{remark}
715:
716: %\begin{remark}\label{rem: divisor eqn and exp form of change of vars}
717: %The divisor equation for the Gromov-Witten invariants of $Y$ can be
718: %formulated as the assertion that the operator
719: %\[
720: %\frac{\partial }{\partial y_{i}} - q_{i}\frac{\partial }{\partial q_{i}}
721: %\]
722: %annihilates the $\beta \neq 0$ part of $F^{Y}$. For the
723: %\end{remark}
724:
725: \begin{remark}
726: The definition of small quantum cohomology given in \cite{Chen-Ruan}
727: or \cite{Ab-Gr-Vi} can be recovered from this one by setting the new
728: parameters equal to zero. It follows then, that one recovers that
729: quantum cohomology ring of $\X$ from the quantum cohomology ring of
730: $Y$ by simply setting some of the $q$'s to roots of unity. The idea of
731: setting quantum parameters on the resolution equal to roots of unity
732: first appears in the mathematics literature in the work of Ruan
733: \cite{Ruan-crepant} where he observes that in some examples, one needs
734: to set $q=-1$ to recover the orbifold cohomology of $\X $.
735: \end{remark}
736:
737:
738: \section{Examples}
739:
740: To provide evidence for our conjecture we consider orbifolds of the
741: form
742: \[
743: \X =\left[V/G \right]
744: \]
745: where $G\subset SL (V)$ is a finite subgroup.
746:
747: When the dimension of $V$ is 2 or 3, there is a canonical crepant
748: resolution given by the $G$-Hilbert scheme \cite{BKR}:
749: \[
750: Y=\GHilb (V).
751: \]
752: The diagonal $\cnums ^{\times}$ action on $V$ commutes with $G$ and the
753: induced action on $X$ lifts to $Y$. Thus the crepant resolution
754: conjecture can be considered $\cnums ^{\times}$ equivariantly.
755:
756: By \cite{BKR}, there is a canonical basis for $H^{*}_{\cnums ^{\times}}
757: (\GHilb V) $ indexed by $R\in \Irr (G)$, irreducible representations
758: of $G $. On the other hand, there is a canonical basis of
759: $H^{*}_{\cnums ^{\times},\orb } ([V/G])$ indexed by $(g)\in \Conj (G)$,
760: conjugacy classes of $G$. Denote the corresponding cohomology variables by
761: \[
762: \left\{y_{R} \right\}_{R\in \Irr (G)} \text{ and }\left\{x_{(g)}
763: \right\}_{(g)\in \Conj (G)}
764: \]
765: respectively. Let $y_{0}$ and $x_{0}$ be the variables corresponding
766: to the trivial representation and the trivial conjugacy class
767: respectively.
768:
769: \subsection{Polyhedral and Binary polyhedral groups.}
770:
771: A finite subgroup $G$ of $ SO (3)$ (respectively $SU (2)$) is called a
772: \emph{polyhedral} (respectively \emph{binary polyhedral}) group. Such
773: groups are classified by ADE Dynkin diagrams and they come with a
774: natural representation $V$ of dimension 3 (respectively 2). For these
775: groups, the equivariant quantum cohomology of $\GHilb (V)$ has been
776: completely described in terms of the root theory of the corresponding
777: ADE root system by Bryan-Gholampour
778: \cite{Bryan-Gholampour2,Bryan-Gholampour3}. They conjecture that the
779: change of variables for the crepant resolution is a certain
780: modification of the character table.
781:
782: \begin{conjecture}\label{conj: change of vars for polyhedral G}
783: The change of variables for the crepant resolution conjecture in the
784: case of
785: \[
786: \GHilb (V) \to V/G
787: \]
788: where $G$ is a polyhedral or binary
789: polyhedral group is given by
790: \begin{align*}
791: y_{0}&=x_{0},\\
792: y_{R}& = \frac{1}{|G|}\sum _{g\in G} \sqrt{\chi _{V} (g)-\dim V}\quad
793: \chi _{R} (g) \, x_{(g)},\\
794: q_{R}&=\operatorname{exp}\left(\frac{2\pi i \dim R}{|G|} \right)
795: \end{align*}
796: where $R$ runs over the \emph{non-trivial } irreducible
797: representations of $G$.
798: \end{conjecture}
799: Note that as a consequence of $V$ being the natural representation of
800: a polyhedral or binary polyhedral group, the orbifold $\X =[V/G]$
801: satisfies the hard Lefschetz condition. Moreover, all non-trivial
802: conjugacy classes have age one, and so the above linear transformation
803: preserves the grading. In fact, these are the only faithful group
804: representations that have the property that all non-trivial elements
805: have age one.
806:
807:
808: Using the root theoretic formula for the Gromov-Witten potential of
809: $\GHilb (V)$ given in \cite{Bryan-Gholampour2,Bryan-Gholampour3} and
810: applying the crepant resolution to the above change of variables, one
811: arrives at a prediction for the orbifold Gromov-Witten potential
812: $F^{\X }$. This prediction has not been verified in general, but it
813: does pass some strong tests of its validity. Namely, it can be shown
814: to exhibit various vanishing properties and to have the correct
815: classical terms.
816:
817: The complete determination of $F^{\X }$, and hence the verification of
818: the crepant resolution conjecture, has been done for $G$ equal to
819: \begin{align*}
820: \znums _{2}&\subset SU (2) \text{ in the next subsection,}\\
821: \znums _{3}&\subset SU (2) \text{ in \cite{Bryan-Graber-Pandharipande},}\\
822: \znums _{4}&\subset SU (2) \text{ in \cite{Bryan-Jiang},}\\
823: \znums _{2}\times \znums _{2}&\subset SO (3) \text{ in \cite{Bryan-Gholampour1}, and}\\
824: A_{4}&\subset SO (3) \text{ in \cite{Bryan-Gholampour1}.}
825: \end{align*}
826:
827:
828:
829:
830: \subsection{The case of the rational double point.}\label{subsec:
831: C2modZ2 case} We consider the case where $V=\cnums ^{2}$ and $G=\{ \pm
832: 1\}\subset SU (2)$ so that
833: \[
834: \X =[\cnums ^{2}/\{\pm 1\}],\quad Y=T^{*}\P ^{1}.
835: \]
836: This is the simplest nontrivial example and it already provides a very
837: interesting case study. Here we will establish the equivariant version
838: of the conjecture.
839:
840:
841:
842: Let $T=\C^{\times } \times \C^{\times }$ so that
843: \[
844: H^{*}_{T} (\point )\cong \qnums [t_{1},t_{2}].
845: \]
846: The natural $T$ action on $\cnums ^{2}$ induces a $T$ action on $Y$,
847: the minimal resolution of the quotient $X=\cnums ^{2}/\{\pm 1 \}$. $Y$
848: is isomorphic to $T^{*}\P ^{1}$, the total space of the cotangent
849: bundle of $\P ^{1}$. There are two fixed points of the $T$ action on
850: $Y$ having weights
851: \[
852: (2t_{1},t_{2}-t_{1})\text{ and } (2t_{2},t_{1}-t_{2}).
853: \]
854:
855: First we will compute the genus zero potential function for $Y$. We
856: take our generator for $H_2(Y)$ to be the class of the zero section,
857: $[E]$. We let $\gamma \in H^2_T(Y)$ be the dual of $[E]$. It is given
858: by the first Chern class of an equivariant line bundle with weights
859: $-t_{1}$ and $-t_{2}$ at the fixed points.
860:
861: The degree zero invariants are simply given by triple intersections in
862: equivariant cohomology
863: \[
864: \left\langle a,b,c \right\rangle_{0} = \int _{Y}a\cup b\cup c
865: \]
866: which are computed by localization. The results are:
867:
868: \[
869: \left\langle1,1,1\right\rangle_{0}= \frac{1}{2t_1t_2},\quad
870: \left\langle\gamma, 1, 1\right\rangle_{0} = 0,\quad
871: \left\langle\gamma, \gamma, 1\right\rangle_{0} = -\frac12,\quad
872: \left\langle\gamma,\gamma,\gamma\right\rangle_{0}=\frac12 (t_1+t_2)
873: \]
874:
875: To compute the invariants in positive degrees, we first observe
876: that the image of any nonconstant morphism from a curve to $Y$
877: must lie in $E$. Thus we have a natural isomorphism
878: \[
879: \M_{0,n}(Y,d[E]) \cong \M_{0,n}(\proj^1, d),
880: \]
881: however, the virtual fundamental classes on the two sides differ.
882: Under the above identification, it is well known that
883: \[
884: [\M_{0,n}(Y,d[E])]^\vir = e(R^1\pi_*f^*N_{E/Y})
885: \]
886: where $\pi :\mathcal{C}\to \M _{0,n} (\P ^{1},d)$ and
887: $f:\mathcal{C}\to Y$ are the universal curve and the universal map
888: respectively.
889:
890: Since $E$ is a -2 curve, we have an isomorphism of
891: $N_{E/Y} \cong \oh(-2)$. Consider the standard
892: Euler sequence on $\proj^1$,
893: \[
894: 0 \to \oh(-2) \to \oh(-1)\oplus\oh(-1) \to \oh \to 0 .
895: \]
896: Pulling this sequence back to $\cC$ and taking the associated long
897: exact sequence of derived pushforwards gives us
898: $$
899: 0\to \oh \to R^1\pi_*f^*(\oh(-2)) \to R^1\pi_*f^*(\oh(-1)\oplus\oh(-1)) \to 0.
900: $$
901: An analysis of the weights shows that the action of $T$ on the left
902: hand term in this sequence is given by $t_{1}+t_{2}$. We conclude
903: that
904: \[
905: e (R^1\pi_*f^*N_{E/Y})=(t_1+t_2)e(R^1\pi_*(\oh(-1)+\oh(-1))).
906: \]
907: The integral is then
908: evaluated using the famous Aspinwall-Morrison formula:
909: \[
910: \left\langle\, \right\rangle_{d}= (t_{1}+t_{2})\int _{[\M _{0,0} (\P
911: ^{1},d)]}e (R^{1}\pi _{*}f^{*} (\oh (-1)\oplus \oh
912: (-1)))=\frac{t_{1}+t_{2}}{d^{3}}.
913: \]
914:
915: Let $y_{0}$ and $y_{1}$ denote the variables corresponding to $1$ and
916: $\gamma $. Combining the above formulas with the divisor equation and
917: the point axiom, we have shown the following.
918: \begin{proposition}\label{prop: FY for TP1}
919: The genus zero Gromov-Witten potential function of $Y$ is given by:
920: \[
921: F^Y = \frac{1}{12t_1t_2} y_0^3 - \frac14 y_0y_1^2 +
922: \frac{t_1+t_2}{12}y_1^3+ (t_{1}+t_{2})\sum_{d>0} \frac{1}{d^3} q^d e^{dy_1}.
923: \]
924: \end{proposition}
925:
926: We now consider the invariants for the orbifold $\X$. Let $1$ and $D$
927: be generators for $H^{0}_{\orb } (\X )$ and $H^{2}_{\orb } (\X )$ and
928: let $x_{0}$ and $x_{1}$ be the corresponding variables.
929:
930: Since the coarse
931: moduli space for $\X$ is affine, every stable map is constant. If the
932: source curve has any twisted points, the image of the map is forced to
933: be the unique point of $\X$ with nontrivial stabilizer. Thus we see
934: that with the exception of
935: \[
936: \langle 1, 1, 1 \rangle=\frac{1}{2t_1t_2},
937: \]
938: every invariant naturally arises as an integral over $\M_{0,n}(B\znums
939: _2)$. By the point axiom and monodromy considerations, the only other
940: invariant involving 1 is
941: \[
942: \left\langle 1,D,D \right\rangle=\frac{1}{2}.
943: \]
944:
945:
946: Since the only remaining non-vanishing invariants are then $\langle
947: D^n \rangle$ we actually need only consider the connected component of
948: $\M _{0,n} (B\znums _{2})$ where all the evaluation maps go to the
949: twisted sector. Setting $n=2g+2$, we denote this space as
950: $\barH_g^{ord}$. Concretely, it is the usual compactified moduli space
951: of hyperelliptic curves (with ordered branch points). The virtual
952: class on $\barH _{g}^{ord}$ is given by
953: \[
954: e (R^{1}\pi _{*}f^{*} (L\oplus L))
955: \]
956: where $f$ and $\pi $ are the universal map and universal curve for $\M
957: _{0,n} (B\znums _{2})$ and
958: \[
959: L\oplus L\to B\znums _{2}
960: \]
961: is two copies of the non-trivial line bundle over $B\znums _{2}$ with
962: the torus acting with weight $t_{1}$ on the first factor and with
963: weight $t_{2}$ on the second factor. The bundle $R^{1}\pi _{*}f^{*}L$
964: is in fact isomorphic to $\E ^{\vee }$, the dual of the Hodge bundle
965: pulled back by the map $\barH _{g}^{ord}\to \M _{g}$. We conclude that
966: that for $n=2g+2>0$, we can write
967: \begin{align*}
968: \left\langle D^{n} \right\rangle &=\int _{\barH
969: _{g}^{ord}}e (\E ^{\vee }\oplus \E ^{\vee })\\
970: &= - (t_{1}+t_{2})\int _{\barH _{g}^{ord}} \lambda _{g}\lambda _{g-1}.
971: \end{align*}
972:
973: The generating function for these integrals was computed in
974: \cite[Corollary 2]{Faber-Pandharipande-logarithmic}. Applying that
975: computation, we obtain:
976: \begin{proposition}\label{prop: F for X=C2modZ2}
977: The potential function of $\X =[\cnums ^{2}/\{\pm 1 \}]$ is given by
978: \[
979: F^{\X } (x_{0},x_{1}) =
980: \frac{1}{12t_{1}t_{2}}x_{0}^{3}+\frac{1}{4}x_{0}x_{1}^{2} -
981: (t_{1}+t_{2})x_{1}^{2}H (x_{1})
982: \]
983: where (following the notation of
984: \cite{Faber-Pandharipande-logarithmic}) $H (x_{1})$ satisfies
985: \[
986: (x_{1}^{2}H (x_{1}))''' = \frac{1}{2}\tan \left(\frac{x_{1}}{2} \right).
987: \]
988: \end{proposition}
989: \begin{cor} The crepant resolution conjecture holds for the pair
990: $(Y,\X )$. That is, the potential functions $F^{Y} (y_{0},y_{1},q)$
991: and $F^{\X } (x_{0},x_{1})$ agree, up to unstable terms, under the
992: change of variables (c.f. Conjecture~\ref{conj: change of vars for
993: polyhedral G})
994: \[
995: y_{0}=x_{0},\quad y_{1}=ix_{1},\quad q=-1.
996: \]
997: \end{cor}
998: \textsc{Proof:} Clearly the terms of $F^{Y}$ and $F^{\X } $ which have
999: $y_{0}$ and $x_{0}$ match up. And since we are only interested in
1000: stable terms, it suffices to check that
1001: \[
1002: \left(\frac{d}{dx_{1}} \right)^{3}F^{Y} (x_{0},ix_{1},-1) =
1003: \left(\frac{d}{dx_{1}} \right)^{3}F^{\X } (x_{0},x_{1}).
1004: \]
1005: The right hand side is given by
1006: \[
1007: - (t_{1}+t_{2})\frac{1}{2}\tan \left(\frac{x_{1}}{2} \right),
1008: \]
1009: whereas the left hand side is
1010: \begin{align*}
1011: &(t_{1}+t_{2})\left[\frac{i^{3}}{2}+\sum _{d=1}^{\infty }i^{3} (-e^{ix_{1}})^{d} \right]\\
1012: =&(t_{1}+t_{2})\frac{1}{2i}\left[\frac{1-e^{ix_{1}}}{1+e^{ix_{1}}} \right]\\
1013: =&(t_{1}+t_{2})\frac{1}{2}\tan \left(\frac{-x_{1}}{2} \right).
1014: \end{align*}
1015: \qed
1016:
1017:
1018:
1019:
1020: \subsection{The case of the Hilbert scheme}\label{subsec: the case of
1021: Hilb}
1022: We consider the case where
1023: \[
1024: \X =\Sym^{n} (\cnums ^{2})\text{ and } Y=\Hilb^{n} (\cnums ^{2}).
1025: \]
1026: This is one of the best known and most studied examples of a crepant
1027: resolution of singularities of a Gorenstein orbifold.
1028:
1029: We will show in this section that by matching the Nakajima basis for
1030: the cohomology of the Hilbert scheme with the natural basis for the orbifold
1031: cohomology of the symmetric product we verify
1032: Conjecture~\ref{mainconjecture} in this case.
1033:
1034: Because the Hilbert scheme is holomorphically symplectic, there are no
1035: interesting Gromov-Witten invariants unless one works equivariantly.
1036: An analogous fact is true on the orbifold side. Thus, to verify the
1037: conjecture for the nonequivariant theory it suffices to compare the
1038: ring structure on the ordinary cohomology of the Hilbert scheme with
1039: the orbifold cohomology of the symmetric product. This is done in
1040: \cite{Vasserot,Lehn-Sorger} (see also \cite{Fantechi-Gottsche,
1041: Uribe}).
1042:
1043: The nontrivial, fully equivariant genus 0 Gromov-Witten theory of $Y$ is
1044: determined in
1045: \cite{Ok-Pan-Hilb}. We will determine the genus 0 equivariant
1046: Gromov-Witten theory of $\X$ and verify that it matches their result after
1047: the appropriate change of variables.
1048:
1049:
1050: First, let us describe the inertia stack $I\X$. We use the standard
1051: correspondence between conjugacy classes of $S_n$ and partitions of
1052: $n$. Given such a partition $\mu$, the corresponding component of the
1053: inertia stack $I_\mu$ can be described by choosing a representative
1054: permutation $\sigma$ and taking the stack quotient $[\C^{2n}_\sigma /
1055: C(\sigma)]$ where $\C^{2n}_\sigma$ denotes the invariant part of
1056: $\C^{2n}$ under the action of $\sigma$ and $C(\sigma)$ denotes the
1057: centralizer of $\sigma$ in $S_n$. The dimension of $\C^{2n}_\sigma$
1058: is $2l(\mu)$ and the age of $\mu$ is $n-l(\mu)$. The quotient of a
1059: vector space by a finite group has no higher cohomology groups, so we
1060: conclude that a basis for the orbifold cohomology of $\X$ as a
1061: $\Q[t_1,t_2]$-module is given by
1062: \[
1063: [I_\mu]\subset H^{2n-2l(\mu)}_{T,\orb } (\X )
1064: \]
1065:
1066: Since each element of $S_n$ is conjugate to its inverse, the
1067: equivariant Poincar\'e pairing on $H^*_{T,\orb}(\X)$ is diagonal in
1068: this basis. It is easily computed by localization, since the fixed
1069: points of the $T$ action on $I\X$ are isolated --- there is a single
1070: fixed point in each irreducible component $I_\mu$. This point has
1071: automorphism group equal to the centralizer of a representative
1072: element, which has order
1073: \[
1074: \fz (\mu) = |\Aut (\mu)|
1075: \prod \mu_i.
1076: \]
1077: It follows that the pairing is given by
1078: \[
1079: ( [I_\mu],[I_\mu])=\frac{1}{\fz(\mu)}(t_1t_2)^{-l(\mu)}.
1080: \]
1081:
1082: It is straightforward to check that the orbifold product here is a slight
1083: modification of the usual multiplication on $Z\qnums [S_{n}]$, the
1084: center of the group ring of $S_n$ obtained by inserting factors of
1085: $t_1t_2$ to make that product respect the grading by age. In
1086: particular, the limit $t_1=t_2=1$ gives the standard product on
1087: $Z\qnums [S_{n}]$.
1088:
1089: On the Hilbert scheme, there is an analogous description of the equivariant
1090: cohomology, given by the Nakajima basis. Given a partition $\mu$, the
1091: corresponding class
1092: \[
1093: N_\mu \in H^{2n-2l (\mu )}_T(\Hn)
1094: \]
1095: is given
1096: by $\frac{1}{\prod \mu_i}[C_\mu]$ where $C_\mu$ is the
1097: subvariety of $\Hn$ whose general point parameterizes a length $n$
1098: subscheme composed of $l(\mu)$ irreducible components
1099: of lengths $\mu_i$. The $T$-equivariant Poincar\'e pairing
1100: in the Nakajima basis is also diagonal with
1101: \[
1102: ( N_\mu , N_\mu) = \frac{(-1)^{n-l(\mu)}}{\fz(\mu)}(t_1t_2)^{-l(\mu)}.
1103: \]
1104:
1105: This gives us an obvious candidate for the map $L$ identifying
1106: the orbifold cohomology of $\Sym^n(\C^2)$ with the cohomology
1107: of $\Hn$. Namely, we define $L$ by
1108: \begin{equation}\label{eqn: change of vars for HilbC2}
1109: L([I_\mu])= i^{l(\mu)}N_\mu.
1110: \end{equation}
1111: Note also, that since there exists a unique partition of length $n-1$,
1112: the partition corresponding to a 2-cycle, which we will denote $(2)$,
1113: there is only one divisor class, and so a single constant $c$ to
1114: choose to finish determining the change of variables. The correct
1115: choice of $c$ turns out to be $-1$. Thus the predicted change of
1116: variables for the quantum parameters is
1117: \[
1118: q=-e^{iu}.
1119: \]
1120:
1121:
1122: To establish the full equality of the genus zero Gromov-Witten
1123: potentials, it will be extremely convenient to use the formalism
1124: introduced for the small quantum product as a bookkeeping device. Let
1125: $c^{\nu }_{\mu }$ be the structure constants for quantum
1126: multiplication by $[I_{(2)}]$:
1127: \[
1128: [I_{(2)}]*[I_\mu] = \sum_\nu c^\nu_\mu [I_\nu].
1129: \]
1130: Here the $c_\mu^\nu$ are elements of $\Q[t_1,t_2][[u]]$ where $u$ is
1131: the quantum parameter associated to the twisted sector as defined in
1132: Section~\ref{sec: degree in twisted sectors and QH}. If we let
1133: $c_\mu^\nu(d)$ denote the coefficient of $u^d$ in $c^\nu_\mu$, then we
1134: have the formula
1135: \begin{equation}\label{gwforc}c_\mu^\nu(d)=\fz(\nu)(t_1t_2)^{l(\nu)}\langle
1136: [I_\mu], [I_\nu], [I_{(2)}]\rangle_d.
1137: \end{equation}
1138: Note that the above Gromov-Witten invariant is an element of
1139: $\Q(t_1,t_2)$, whereas $c^{\nu }_{\mu } (d)$ is a polynomial. This
1140: fact will be essential for the degree arguments that follow. While
1141: the polynomality is an immediate consequence of the existence of the
1142: equivariant quantum product referred to in Section~\ref{subsec:
1143: noncompact case}, the reader can also check that it follows directly
1144: from the explicit localization formula we will give in the next
1145: section.
1146:
1147:
1148: By degree considerations, we see that $c_\mu^\nu$ vanishes if
1149: $l(\mu)\geq l(\nu)+1$. Since Equation \ref{gwforc} gives a symmetry,
1150: we also have the inequality $l(\nu)\geq l(\mu) +1$. We will see below
1151: that if $|l(\nu) - l(\mu)|=1$ then the only contribution to
1152: $c_\mu^\nu$ is in degree zero where we just see the classical term
1153: corresponding to multiplication in the group ring of $S_n$. Aside
1154: from these classical terms, the matrix $c^\nu_\mu$ is diagonal.
1155:
1156: \begin{lemma}\label{diagonal} If $l(\mu)=l(\nu)$, but $\mu \neq \nu$,
1157: then $c_\mu^\nu =0$.
1158: \end{lemma}
1159: \textsc{Proof:} Let $\M _{0,(\lambda ,\mu ,\nu )} (\X ,d)$ denote the
1160: component(s) of $\M _{0,3} (\X ,d)$ given by $e_{1}^{-1} (I_{\lambda
1161: })\cap e_{2}^{-1} (I_{\mu })\cap e_{3}^{-1} (I_{\nu })$.
1162:
1163: By definition, we have
1164: \[
1165: c_\mu^\nu (d)[I_{\nu }] =
1166: e_{3*}([\M_{0,((2),\mu,\nu)}(\X,d)]^\vir)^{\vee }.
1167: \]
1168: By degree consideration, this must be a codimension 1
1169: class in $I_\nu$. However it is easy to see that the codimension of
1170: $e_{3}(\M_{0,((2),\mu,\nu)}(\X,d))$ is at least 2, since the
1171: intersection of the images of $I_\mu$ and $I_\nu$ in $\X$ has
1172: codimension at least 2 in each. The lemma follows immediately. \qed
1173:
1174: To finish the determination of the structure of the quantum cohomology
1175: ring, we use a localization calculation. Because $\Sym^n(\C^2)$ is
1176: affine, every twisted stable map is constant at the level of coarse
1177: moduli schemes. It follows that we have a canonical identification of
1178: the $T$ fixed locus of the space of maps to $\X$ with the space of
1179: maps to $BS_n$ (the fixed locus of the action of $T$ on $\X$). The
1180: normal bundle to this fixed locus decomposes naturally as a sum of two
1181: rank $n$ vector bundles. These two bundles come with $T$ weights
1182: $t_1$ and $t_2$, but are otherwise identical, each corresponding to
1183: the standard $n$-dimensional representation of $S_n$ under the usual
1184: correspondence between sheaves on $BG$ and representations of $G$. We
1185: will use $V$ to denote this bundle. There is another way to think of
1186: $V$ which is convenient for us here. Consider the morphism
1187: $i:BS_{n-1} \to S_n$ induced by the standard inclusion of $S_{n-1}
1188: \hookrightarrow S_n$. Then $V$ is simply the pushforward of the
1189: structure sheaf.
1190:
1191: Since the coarse moduli scheme of $BS_n$ is a point, the moduli
1192: space of twisted maps $\M_{0,r}(BS_n)$ is smooth of dimension $r-3$.
1193: By the results of \cite{Gr-Pa} we can identify the equivariant
1194: virtual fundamental class of $\M_{0,r}(\Sym^n(\C^2))$ with the pushforward
1195: from this fixed locus of the class
1196: \[
1197: e(-R^{\bullet } \pi_*f^*(V\oplus V)) .
1198: \]
1199: where the torus acts on the two factors of $V$ are with weights $t_1$ and
1200: $t_2$, and where $f$ and $\pi$ are the universal maps in the universal
1201: diagram
1202: $$\xymatrix{\mathcal{C} \ar[r]^f \ar[d]^\pi & BS_n \\
1203: \M_{0,r}(BS_n).}
1204: $$
1205: To give a description of the virtual class in more familiar terms, we extend
1206: the above diagram to
1207:
1208: $$\xymatrix{ \tilde{\mathcal{C}} \ar@/_1pc/[dd]_p \ar[r]^{\tilde{f}} \ar[d]^g & BS_{n-1} \ar[d]^i \\
1209: \mathcal{C} \ar[r]^f \ar[d]^\pi & BS_n \\
1210: \M_{0,r}(BS_n)}.
1211: $$
1212:
1213: The curve $\tilde{\mathcal{C}}$ here is the degree $n$ covering of
1214: $\mathcal{C}$ corresponding to the map to $BS_n$ via the usual correspondence
1215: between principal $S_n$ bundles and degree $n$ \'etale covers.
1216:
1217: We know that $V=i_*\oh_{BS_{n-1}}$. Since $i$ is a finite morphism,
1218: we have $f^*(i_*\oh)=g_*\tilde{f}^*\oh_{BS_{n-1}}=g_*\oh_{\tilde{C}}$.
1219: Since $g_*$ is exact, we conclude that the virtual class can be
1220: rewritten as the pushforward from the fixed locus of
1221: \[
1222: e(-R^{\bullet } p_* (
1223: \oh_{\tilde{\mathcal{C}}} \oplus \oh_{\tilde{\mathcal{C}}})).
1224: \]
1225: In other words, if we think of the space of maps to $BS_n$ as
1226: parameterizing the family of $n$-sheeted covers of $\proj^1$ given by
1227: $\tilde{\mathcal{C}}$, then the invariants we want to compute are
1228: expressed in terms of the Chern classes of the Hodge bundle. Thus if
1229: we let
1230: \[
1231: \E^\vee=R^1p_*\oh_{\tilde{\mathcal{C}}}
1232: \]
1233: and we let $s$ denote the locally constant function on
1234: $\M_{0,r}(BS_n)$ recording the number of connected components of the
1235: fibers of $\tilde{\mathcal{C}}$, we obtain the following formula.
1236:
1237: \begin{lemma}\label{locform} The $r$ point, degree zero invariants of
1238: $\X $ are given by
1239: \[
1240: \langle \mu_1 , \ldots, \mu_r\rangle_{0} =\int_{[\M_{0,(\mu_1, \ldots
1241: , \mu_r)}(BS_n)]^\vir} (t_1t_2)^{-s}c_{top}(\E_{t_1}^\vee \oplus
1242: \E_{t_2}^\vee)
1243: \]
1244: where $\M _{0,(\mu _{1}\dotsb \mu _{r})} (BS_{n})$ is the component(s) of
1245: $\M _{0,r} (BS_{n})$ where $e_{i}$ maps to the component of $IBS_{n}$
1246: corresponding to $\mu _{i}$. We also must interpret the above
1247: integral as a sum over connected components of the moduli space. The
1248: rank of $\E$ and the integer $s$ can vary from component to component.
1249: \end{lemma}
1250:
1251: Since the orbifold $\X $ only has divisor classes in the twisted
1252: sector, the higher degree invariants are determined by the degree zero
1253: invariants by the divisor equation. Thus the above lemma determines
1254: all the invariants.
1255:
1256: \begin{lemma} If $d>0$, then $c_\mu^\nu(d)$ is divisible by $(t_1+t_2)$.
1257: \end{lemma}
1258: This is a consequence of Mumford's relation that $c(\E \oplus \E^\vee)
1259: = 1$. If we set $t_2=-t_1$ and use the fact that
1260: $c_{top}(\E^\vee_{-t_1}) = \pm c_{top}(\E_{t_1})$, we find that the
1261: integrand in Lemma \ref{locform} is simply a power of $t_1$. The
1262: hypothesis $d>0$ implies that the moduli space is positive
1263: dimensional, so the result follows. \qed
1264:
1265: We remark that on the Hilbert scheme side, this divisibility is
1266: related to the existence of a holomorphic symplectic structure on
1267: $\Hilb^n(\C^2)$.
1268:
1269: \begin{cor}If $d>0$ and $\mu \neq \nu$, we have $c^\nu_\mu=0$.
1270: \end{cor}
1271: Given Lemma \ref{diagonal} and the discussion just before it, we see
1272: that the only interesting case here is if $l(\mu) = l(\nu)+1$.
1273: However, in this case, we know that the degree of $c_\mu^\nu$ is zero,
1274: so the divisibility constraint forces this invariant to vanish.\qed
1275:
1276:
1277: This reduces our task to the calculation of the invariants
1278: $c^\mu_\mu$. We can further reduce to the case where the partition
1279: $\mu$ has just one part, by using the following lemma.
1280:
1281: \begin{lemma}
1282: $$c^\mu_\mu =\frac{1}{\fz(\mu)} \sum_{i=1}^{l(\mu)} \mu_i
1283: c_{(\mu_i)}^{(\mu_i)}.$$
1284: \end{lemma}
1285:
1286: The right hand side of this formula is easily seen to be the
1287: contribution from those components of $\M_{0,((2),\mu,\mu)}(\X,d)$
1288: where the corresponding branched cover $C$ consists of $l(\mu)$
1289: connected components, all but one of which is a smooth genus zero
1290: curve branched only at 0 and $\infty$. To prove the lemma, we need to
1291: show that the other components make no contribution. We will do this
1292: by means of the formula of Lemma \ref{locform}, so we will always be
1293: considering the space of maps to $BS_n$ rather than to $\X$, and we
1294: will denote the two distinguished points of the source curve
1295: corresponding to $\mu$ as 0 and $\infty$. These are the only points
1296: over which the associated branched cover of $\proj^1$ has non-simple
1297: branching.
1298:
1299: \smallskip
1300:
1301: \noindent {\bf Step 1}: Suppose we have a component where the
1302: associated cover is connected. Then, by the Riemann-Hurwitz formula,
1303: it will have genus $g=\frac{d+3}{2} - l(\mu)$. In order for the
1304: integral in Lemma \ref{locform} not to vanish, it is obviously
1305: necessary that $2g\geq d$ since $d$ is the dimension of the moduli
1306: space. This inequality is satisfied only if $l(\mu)=1$ (in which case
1307: the Lemma is vacuously true). Otherwise, we conclude that a component
1308: of the moduli space can contribute to this invariant only if it
1309: parameterizes disconnected covers.
1310:
1311: \smallskip
1312:
1313: \noindent {\bf Step 2}: Suppose we consider a component $\M'$ of the
1314: moduli space where the corresponding branched cover is disconnected.
1315: We get a natural map $\Psi: \M' \to (\prod_a \M_a)/\Aut$ where the
1316: $\M_a$ are some moduli spaces of lower degree branched covers with
1317: certain branching conditions and the group $\Aut$ is acting by
1318: permuting factors with identical parameters. We do not need a very
1319: careful description here, since we will use just two crude facts.
1320: First, if two different factors of the target space parameterize covers
1321: with branching away from zero and $\infty$, then $\Psi$ has positive
1322: dimensional fibers, since we can independently act by $\C^{\times }$ on
1323: different components. Since $\E$ is pulled back under $\Psi$ this
1324: immediately kills contributions from any such component of $\M$.
1325:
1326: If a branched cover has all the simple branch points on a single
1327: connected component, then the other components are necessarily genus
1328: zero curves ramified only at zero and infinity. Now Step 1 will apply
1329: to the remaining interesting component, showing that this component of
1330: moduli space makes no contribution to the integral unless we are in
1331: the maximally disconnected case. \qed
1332:
1333: We remark that the argument in this lemma extends to give an alternate
1334: proof of Lemma \ref{diagonal}.
1335:
1336:
1337: We see then, that we will have completely determined the quantum
1338: multiplication by $[I_{(2)}]$ once we calculate the invariants
1339: $c_{(n)}^{(n)}$ for all $n$. Here we can give an explicit formula.
1340:
1341: \begin{lemma} We have the following:
1342: \[
1343: \sum_d \langle [I_{(n)}], [I_{(n)}], [I_{(2)}] \rangle_d u^d =
1344: -\frac{i}{2}\frac{t_1+t_2}{t_1t_2}(n\cot(\frac{nu}{2}) - \cot(\frac{u}{2})).
1345: \]
1346: \end{lemma}
1347:
1348: \textsc{Proof:} This follows from the same argument as \cite[Theorem
1349: 6.5]{Br-Pa-local-curves}. The restriction of $\tilde{\mathcal{C}}$ to
1350: $e_{1}^{-1} (I_{(n)})\subset \M_{0,3} (\X ,d)$ is necessarily a family
1351: of connected curves of genus $g$ where $d=2g-1$. Applying
1352: Lemma~\ref{locform} and the divisor equation, we get the formula
1353: \[
1354: \langle [I_{(n)}], [I_{(n)}], [I_{(2)}] \rangle_d =
1355: -\frac{t_1+t_2}{t_1t_2}\frac{1}{(2g-1)!}\int_{[\M_{0,((n)(n)(2)\dotsb (2))}
1356: (BS_{n})]}\lambda_g\lambda_{g-1}.
1357: \]
1358: The map to $\M_{g,2}$ induced by the family $\tilde{\mathcal{C}}$ is
1359: generically finite of degree $(2g)!$ onto its image, which is the set
1360: of curves admitting a degree $n$ map to $\proj^1$ totally ramified at
1361: the two marked points. The image of this map is called
1362: $\overline{H}_{d}\subset \M _{g,2}$ in \cite{Br-Pa-local-curves} and
1363: the pairing of $[\overline{H}_{g}]$ against $\lambda_g\lambda_{g-1}$
1364: is explicitly evaluated in \cite{Br-Pa-local-curves} to yield the
1365: series above. \epf
1366:
1367: Having completely determined the $c^\nu_\mu$ we can deduce our main result.
1368: \begin{thm} \label{thm: QH(Sym)=QH(Hilb))}
1369: After making the change of variables given by
1370: equation~(\ref{eqn: change of vars for HilbC2}) and relating the
1371: quantum parameters by
1372: \[
1373: q=-e^{iu},
1374: \]
1375: the genus zero Gromov-Witten potential of $\Hilb^n(\C^2)$ is equal to
1376: the (extended) genus zero Gromov-Witten potential of
1377: $\Sym^n(\C^2)$. Hence the crepant resolution conjecture holds in this
1378: case.
1379: \end{thm}
1380:
1381: By direct inspection, the matrix of multiplication by $[I_{(2)}]$ in
1382: $QH^*(\Sym^n(\C^2))$ matches with the matrix of multiplication by
1383: $i[N_{(2)}]$ in $QH^*(\Hilb^n(\C^2))$ calculated in \cite{Ok-Pan-Hilb}
1384: (equations (6) and (8), see also \cite{Br-Pa-local-curves} equations
1385: (19) and (29)) under the change of variables $q=-e^{iu}$. As is
1386: observed there, the fact that this matrix has distinct eigenvalues
1387: implies that after extending the scalars to $\Q(t_1,t_2)$ the quantum
1388: cohomology is generated by the divisor class $[I_{(2)}]$. Thus the
1389: entire ring structure is encoded in this multiplication matrix.
1390:
1391: Finally, since the small quantum cohomology is generated by divisors,
1392: a variant of the reconstruction theorem of Kontsevich-Manin shows that
1393: one can use the WDVV equation to reduce arbitrary genus zero
1394: Gromov-Witten invariants to invariants with only two insertions
1395: (c.f. \cite{Rose}). As these are already encoded in the small quantum
1396: product, the proof of the theorem is complete. \qed
1397:
1398:
1399:
1400:
1401:
1402:
1403: \comment
1404: {If $\lambda, \mu, \nu$ are partitions of $d$, let
1405: $\M^\bullet_g(\proj^1 \times \C^2,\lambda,\mu,\nu)$ be the moduli
1406: space of genus $g$ relative stable maps to $\proj^1 \times \C^2$
1407: relative to the divisors over $0,1,$ and $\infty$. We allow the
1408: source to be disconnected, but the map must be nonconstant on each
1409: connected component. Using the action of the torus on $\C^2$ with
1410: compact fixed point set $\proj^1$, we can define via localization a
1411: basic invariant here by simply integrating the virtual fundamental
1412: class.
1413:
1414: To formulate the relationship between stable relative maps and admissible
1415: covers that we need, we set
1416: $\M_{g}(\lambda,\mu,\nu) \subset \M_{0,2g+2}(BS_d)$
1417: to be
1418: $$ev_1^{-1}(I_\lambda)\cap ev_2^{-1}(I_\mu)
1419: \cap ev_3^{-1}(I_\nu) \bigcap_4^{2g+2}ev_i^{-1}I_{(2)}.$$
1420:
1421: \begin{lemma}There is a natural $T-$equivariant morphism
1422: $$\Phi: \M_g(\lambda,\mu,\nu) \to \M^\bullet_g(\proj^1,n)_{\lambda,\mu,\nu}$$
1423: which restricts to an isomorphism from a
1424: dense open subset of $\M_g(\lambda, \mu, \nu)$ onto its image.
1425: \end{lemma}
1426:
1427: \bpf
1428:
1429: Consider the representable morphism
1430: from $BS_{n-1} \to BS{n}$ associated
1431: to the standard inclusion of $S_{n-1}$ in $S_d$.
1432: Given a twisted stable map
1433: $f: \DD \to BS_d$, we form the fiber product
1434:
1435: Because the map $BS_{n-1} \to BS_d$ is \'etale of degree $d$, we find that
1436: $C$ is a nodal curve and the vertical arrow $g$ is a degree
1437: $d$ \'etale morphism to a twisted curve whose coarse moduli scheme $D$
1438: is a stable $2g+2$ pointed genus zero curve. If we let $g'$
1439: denote the composition of $g$ with the map to $D$, the conditions
1440: on the evaluation of the three marked points, implies that the
1441: ramification type of $g'$
1442: over the first three markings are $\lambda, \mu,$ and $\nu$.
1443: We can use the first three
1444: markings to give a canonical morphism $C \to \proj^1$ taking them to
1445: $0,1,\infty$. The key point here is that there exists a curve $\bD$
1446: intermediate between $D$ and $\proj^1$, such that
1447: $C \to \bD \to \proj^1$ is a relative stable morphism. $\bD$
1448: is obtained by simply contracting all components of $D$ which would not
1449: be included in a shortest path joining two
1450: of the first three markings.
1451:
1452: That this construction behaves well in families is an easy consequence
1453: of Hassett's comprehensive work on moduli spaces of weighted pointed
1454: curves. This is because the contraction desired is the one associated
1455: to giving the first three markings weight 1, and the remaining markings
1456: a small weight $\epsilon$. In other words, if we employ as a stability
1457: condition for our pointed curve that the $\Q$ divisor $\omega_D(p_1+p_2+p_3
1458: +\epsilon\sum_4^? p_i)$ be ample, then the stable model for $D$ will be
1459: the desired curve $\bD$. The naturality of this construction is just
1460: a special case of Theorem 4.1 of \cite{hassett}.
1461:
1462: It is obvious that the morphism $\Phi$ is an isomorphism
1463: over the locus where the source curve is smooth.
1464:
1465: \epf}
1466:
1467: \pagebreak
1468:
1469:
1470: \subsection{Equivalence with other theories}
1471:
1472: There are two other theories which are equivalent to the quantum
1473: cohomologies of $\sym ^{n} (\cnums ^{2})$ and $\Hilb ^{n} (\cnums
1474: ^{2})$. By computing the equivariant Gromov-Witten partition function
1475: (in all genus) for the degree $n$ invariants of $\P ^{1}\times \cnums
1476: ^{2}$ relative to $\{0,1,\infty \}\times \cnums ^{2}$, one obtains the
1477: structure constants of an associative Frobenius algebra
1478: \cite{Br-Pa-local-curves}. Similarly, on obtains a Frobenius algebra
1479: from the partition function for the degree $n$ equivariant
1480: Donaldson-Thomas invariants of $\P ^{1}\times \cnums ^{2}$ relative to
1481: $\{0,1,\infty \}\times \cnums ^{2}$
1482: \cite{Okounkov-Pandharipande-dtlc}. Theorem~\ref{thm:
1483: QH(Sym)=QH(Hilb))} completes the following tetrahedron of
1484: equivalences.
1485:
1486: \vspace{1.5in}
1487: \begin{center}
1488: \scriptsize
1489: \begin{picture}(200,75)(-30,-50)
1490: \put(-100 ,-115 ){\line(1 ,0 ){340}}
1491: \put(-100 ,-115 ){\line(0 ,1 ){225}}
1492: \put(-100 ,110 ){\line(1 ,0 ){340}}
1493: \put(240 ,-115 ){\line(0 ,1 ){225}}
1494: \thicklines
1495: \put(25,25){\line(1,1){50}}
1496: \put(25,25){\line(1,-1){50}}
1497: \put(125,25){\line(-1,1){50}}
1498: \put(125,25){\line(-1,-1){50}}
1499: \put(75,-25){\line(0,1){100}}
1500: \put(25,25){\line(1,0){45}}
1501: \put(80,25){\line(1,0){45}}
1502: \put(75,95){\makebox(0,0){Equivariant quantum}}
1503: \put(75,85){\makebox(0,0){cohomology of $\Hilb (\cnums ^{2})$ }}
1504: \put(75,-35){\makebox(0,0){Equivariant orbifold quantum }}
1505: \put(75,-45){\makebox(0,0){cohomology of $\operatorname{Sym} (\cnums ^{2})$ }}
1506: \put(160,35){\makebox(0,0){Equivariant}}
1507: \put(160,25){\makebox(0,0){Gromov-Witten}}
1508: \put(160,15){\makebox(0,0){theory of $\P ^{1}\times \cnums ^{2}$}}
1509: \put(-15,35){\makebox(0,0){Equivariant}}
1510: \put(-15,25){\makebox(0,0){Donaldson-Thomas}}
1511: \put(-15,15){\makebox(0,0){theory of $\P ^{1}\times \cnums ^{2}$}}
1512: \end{picture}
1513: \end{center}
1514: \begin{quote}
1515: \scriptsize{The above four theories are equivalent. The southern and
1516: eastern theories have parameter $u$, while the northern and western
1517: theories have parameter $q=-e^{iu}$. The vertical equivalence is the
1518: equivariant Crepant Resolution Conjecture for $\Hilb \cnums ^{2}\to
1519: \operatorname{Sym}\cnums ^{2}$. The horizontal equivalence is the
1520: equivariant DT/GW correspondence for $\P ^{1} \times \cnums ^{2}$. The
1521: four corners are computed in \cite{Br-Pa-local-curves,
1522: Okounkov-Pandharipande-dtlc,Ok-Pan-Hilb} and the present paper. }
1523: \end{quote}
1524: \vspace{.5in}
1525:
1526:
1527:
1528: \subsection{Acknowledgments.} We warmly acknowledge helpful
1529: discussions with Mina Aganagic, Renzo Cavalieri, Tom Coates, Amin Gholampour,
1530: Yunfeng Jiang, Rahul Pandharipande, Michael Thaddeus, and Hsian-Hua
1531: Tseng. We acknowledge support from NSERC, NSF, the Sloan Foundation,
1532: and IHES.
1533:
1534:
1535:
1536:
1537: \bibliography{mainbiblio}
1538: \bibliographystyle{plain}
1539:
1540: \end{document}
1541:
1542: