The second author is partially supported by NSF..
Tautological equation in
via invariance conjectures
D. Arcara
Y.-P. Lee
Department of Mathematics, University of Utah, 155 S. 1400 E., Room 233, Salt Lake City, UT 84112-0090, USA E-mail address : arcara@math.utah.edu E-mail address : yplee@math.utah.edu
-
Abstract.
We use the Invariance Conjectures in tautological relations explained in [1, 8, 9, 10] to find a tautological relation of
in codimension 3.
0 Introduction
This work is a continuation of [1] . We apply the same technique to codimension 3 strata in
and find a tautological equation.
Our motivation was quite simple. In the earlier work [6, 1] , the conjectural framework is shown to be valid in genus one and two. While it is satisfactory to learn that all previous equations can be derived in this framework, the Conjectures predicts the possibility of finding all tautological equations under this framework. This work is set to be the first step towards this goal.
The choice of codimension 3 in
is almost obvious. First of all, the Invariance Conjectures asks for an inductive process. Given what we know about genus one and two, it is only reasonable to proceed to either
for
or
.
One also knows from the Theorem of Graber–Vakil [5] that
on
is rational equivalent to a sum of boundary strata containing at least one (geometrical) genus zero component. It seems a reasonable place to start. As an excercise, the reader might amuse himself with the following
Lemma.
There is no relation among the boundary strata and
classes in codimension 2 in
.
This can be shown, for example, by the same technique used in this paper. Let
denotes a basis of these strata. When one sets a hypothetical equation
and imposes the invariance condition, the only solution is
for all
.
The main result of this paper is the following:
Theorem.
The invariance conjecture can be used to find the following tautological equation for codimension 3 strata in
.
| |
Remark.
While our paper is under preparation, a preprint by T. Kimura and X. Liu [
7]
appeared on the arxiv. There are two major differences between our results.
First, their choice of basis of codimension 3 strata in
is different. They use
instead of (3) below. We have checked that their equation is equivalent to ours. Second, their approach is “traditional”: knowing there must be a relation from Graber–Vakil, they can then proceed to find the coefficients based on the evaluation of the Gromov–Witten invariants of
.
Our approach is quite different. While the result of Graber–Vakil serves as a motivation, our scheme is completely independent of their result. Our scheme is also simpler: only linear algebra is involved. The drawback of course is that the framework is still conjectural.
Acknowledgement.
We are thankful to E. Getzler and R. Vakil for the useful discussions.
This paper is dedicated to Matteo S. Arcara, who was born while the paper was at its last phase of preparation.
1 Strata of
We start with picking a basis for codimension 3 strata in
.
Out of the several strata (allowing descendences) of codimension
in
, many of them can be written in terms of the others using WDVV, TRR's, Mumford's, Getzler's or the Belorousski-Pandharipande relations. After using those relations, we can write all of the strata in terms of the following ones:
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
2 Derivatives for
Let us now calculate the derivatives in the case
. Let
.
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
We now start with the hypothetical equation
and apply the operator
. See the first equation in Appendix A with
for the form of
. Note that it is shown in [9] the contributions of the sum of the terms produced by the first term there is zero. Simiarly the second term contributes only if the operator acts on the edges.
The Invariance Conjecture predicts
In each strata, we will pick a basis for the tautological group and set the coefficients of the basis equal to 0. Some of the strata will be disconnected, i.e. the dual graph is disconnected. They are easier to deal with as they involve less relations (e.g. WDVV). Some of them involve no relations at all and can be used as our basis in the lower strata.
So let us start with the disconnected terms.
-
The terms of the form
give the equation
-
The terms of the form
give the equation
|
(2)
|
-
The terms of the form
give the equation
|
(3)
|
-
The terms of the form
give the equation
|
(4)
|
-
The terms of the form
give the equation
|
(5)
|
-
The terms of the form
give the equation
|
(6)
|
-
The terms of the form
give the equation
|
(7)
|
-
The terms of the form
give the equation
|
(8)
|
-
The terms of the form
give the equation
|
(9)
|
-
The terms of the form
give the equation
|
(10)
|
-
The terms of the form
give the equation
|
(11)
|
-
The terms of the form
give the equation
|
(12)
|
-
The terms of the form
give the equation
|
(13)
|
-
The terms of the form
give the equation
|
(14)
|
-
The terms of the form
give the equation
|
(15)
|
-
The terms of the form
give the equation
|
(16)
|
-
The terms of the form
give the equation
-
The terms of the form
give the equation
|
(18)
|
-
The terms of the form
give the equation
|
(19)
|
-
The terms of the form
give the equation
|
(20)
|
-
The terms of the form
give the equation
|
(21)
|
-
The terms of the form
give the equation
|
(22)
|
-
The terms of the form
give the equation
|
(23)
|
-
The terms of the form
give the equation
|
(24)
|
-
The terms of the form
give the equation
|
(25)
|
-
The terms of the form
give the equation
|
(26)
|
-
The terms of the form
give the equation
|
(27)
|
-
There are several terms containing
. If we remove the
, they become terms in
of codimension
, and there is a relation between them which we can find by using Getzler's relation with a two descendences on
and one descendence on
(with
). The relation is1
| |
| |
| |
| |
| |
and it does not involve terms of the form
or
. We are going to solve this relation for the term
and find an equation for all of the other terms. Since the terms of the form
have a coefficient of
in our derivative, we will be adding the appropriate multiple of this number to the coefficient of the other terms.
-
The terms of the form
give the equation
|
(28)
|
-
The terms of the form
give the equation
|
(29)
|
-
The terms of the form
give the equation
|
(30)
|
-
The terms of the form
give the equation
|
(31)
|
-
There are seven different terms with genus
part
, and they are related by WDVV. To simplify the notation, we shall call the seven terms
,
,
,
,
,
, and
. Here is a list of each term with the new notation for it and its coefficient in the derivative.
-
–
The term
has coefficient
-
–
The term
has coefficient
-
–
The term
has coefficient
-
–
The term
has coefficient
-
–
The term
has coefficient
-
–
The term
has coefficient
-
–
The term
has coefficient
Using WDVV, we can write the seven terms in terms of
,
,
, and
as follows:
After these substitutions, the terms of the form
give the equation
|
(32)
|
The terms of the form
give the equation
|
(33)
|
The terms of the form
give the equation
|
(34)
|
The terms of the form
give the equation
|
(35)
|
-
The terms of the form
give the equation
|
(36)
|
-
The terms of the form
give the equation
|
(37)
|
-
The terms of the form
give the equation
|
(38)
|
The other terms are all connected, and they are strata of codimension
in
. We found relations between them in two different ways2
: First of all, by taking Getzler's relation in
, adding another marked point, and then identifying either two of the first four marked points or the fifth marked point with one of the others. Secondly, by taking the Belorousski-Pandharipande relation, adding a descendence at the marked point
, and then simplifying. The relations we obtain are the following:
| |
| |
| |
| |
| |
| |
| |
| |
and
| |
| |
| |
| |
Here is a table of the terms we shall solve for, each with their coefficient in the derivative:
| |
| |
| |
| |
After solving the four relations above for these four terms, we obtain the following equations for the other terms.
-
The terms of the form
give the equation
|
(39)
|
-
The terms of the form
give the equation
|
(40)
|
-
The terms of the form
give the equation
|
(41)
|
-
The terms of the form
give the equation
|
(42)
|
-
The terms of the form
give the equation
|
(43)
|
-
The terms of the form
give the equation
|
(44)
|
-
The terms of the form
give the equation
|
(45)
|
-
The terms of the form
give the equation
|
(46)
|
-
The terms of the form
give the equation
|
(47)
|
-
The terms of the form
give the equation
|
(48)
|
-
The terms of the form
give the equation
|
(49)
|
Solving the equations ( 1 )-( 49 ) gives the following coefficients (let
):
| |
| |
| |
| |
| |
| |
A Summary of the known relations used in the paper
We put here a summary of what we used in the calculations.
-
| |
| |
| |
| |
-
The combined effort of the first term vanishes.
-
The combined effort of the second term when
is at a marked point also vanishes.
-
Symmetries: For example,
and
.
-
WDVV
for every
,
-
TRR in
-
TRR in
-
Getzler's Relation in
:
-
Mumford's Relation in
:
| |
| |
| |
| |
-
Getzler's Relation in
:
| |
| |
| |
| |
| |
| |
| |
| |
| |
-
Belorousski-Pandharipande Relation in
:
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
B A relation in
in codimension
We use Getzler's and Mumford's relations to derive a new relation for
in codimension
. Start with Getzler's relation with
and
:
| |
| |
| |
| |
| |
We can then use Mumford's relation to simplify the left-hand side, and TRR's to simplify the right-hand side. If we set the two sides equal to each other and simplify, we obtain the equation
| |
| |
| |
| |
| |
C Relations in
in codimension
First of all, we use Getzler's relation in
to find relations in
in codimension
. Since we are only interested in terms with a total number of three markings at genus
vertices, we start with Getzler's relation written as follows:
After we disimmetrize, and add a fifth marked point
, we obtain the relation
| |
| |
| |
| |
If we identify the marked point
with the marked point
, we obtain that
| |
| |
| |
| |
and if we identify the marked point
with the marked point
(and rename the marked point
by
), we obtain that
| |
| |
Now we use Belorousski-Pandharipande's, Getzler's and Mumford's relations to derive other relations for
in codimension
. Start with the Belorousski-Pandharipande relation and add one descendence to each marked point
(note that, for doing this, we need to disimmetrize the relation):
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
Then use Getzler's and Mumford's relations and TRR's to simplify and obtain
| |
| |
| |
| |
| |
| |
| |
| |
| |
Using the two relations we found above from Getzler's relation in
, this simplifies into
| |
| |
| |
| |
References
-
D. Arcara, Y.-P. Lee, Tautological equations in genus
via invariance conjectures, arXiv:math.AG/0502488.
-
P. Belorousski, R. Pandharipande, A descendent relation in genus 2, Ann. Scuola Norm. Sup. Pisa Cl. Sci. (4) 29 (2000), no. 1, 171–191.
-
E. Getzler, Intersection theory on
and elliptic Gromov-Witten invariants, J. Amer. Math. Soc. 10 (1997), no. 4, 973–998.
-
E. Getzler, Topological recursion relations in genus
, Integrable systems and algebraic geometry (Kobe/Kyoto, 1997), 73–106, World Sci. Publishing, River Edge, NJ, 1998.
-
T. Graber, R. Vakil, Relative virtual localization, and vanishing of tautological classes on moduli spaces of curves, math.AG/0309227, Duke Math. J., to appear.
-
A. Givental, Y.-P. Lee, preliminary draft. Part of the core calculations can be found at http://www.math.utah.edu/~yplee/research/ite_talk.pdf.
-
T. Kimura, X. Liu, A genus-3 topological recursion relation, math.DG/0502457.
-
Y.-P. Lee, Witten's conjecture, Virasoro conjecture, and invariance of tautological relations, math.AG/0311100.
-
Y.-P. Lee, Witten's conjecture and Virasoro conjecture up to genus two, math.AG/0310442. To appear in the proceedings of the conference ”Gromov-Witten Theory of Spin Curves and Orbifolds”, Contemp. Math., AMS.
-
Y.-P. Lee, Conjectures on the tautological equations, http://www.math.utah.edu/~yplee/research/cite.pdf.
Department of Mathematics, University of Utah, 155 S. 1400 E.,Room 233, Salt Lake City, UT 84112-0090, USA E-mail address : arcara@math.utah.edu E-mail address : yplee@math.utah.edu