Part III [03W0]
Original official author HTML, exact retained edition. Historical TeX conversion verdicts remain unchanged. Cited-edition alignment and mathematical self-containment are not assessed.
Complete original source context · Original author HTML
Part III
We fix field satisfying Zero Characteristic Assumption.
Let be a compact oriented surface, a finite set, and a sheaf defining a -affine structure on . We assume that all singularities of the underlying -affine structure are standard (see Section 6.4), and local monodromy around each acts on with a fixed point (see the Fixed Point property at the end of Section 7.1). Main result of Part 3 of the paper can be formulated such as follows.
Theorem 5
There exist a compact -analytic surface , a top degree analytic form and a continuous proper Stein map such that the set of -smooth points coincides with and the induced -affine structure coincides with the one given by .
In other words, the triple is a solution of the Lifting Problem.
By Stein property it suffices to construct the sheaf of -algebras on . We will see that outside of the finite singular set the sheaf is locally isomorphic to . In the next section we will describe the local model for the sheaf near each singular point. It will be glued together with a modification of the canonical sheaf . This modification depends on the data called lines. Appearance of lines is motivated by Homological Mirror Symmetry (see [Ko], [KoSo]) 66 6 The main idea is that is a component of the moduli space of certain objects (skyscrapper sheaves) in the derived category . These objects correspond to -local systems on Lagrangian tori in the Fukaya category of the mirror dual symplectic manifold.. Roughly speaking, lines correspond (for mirror dual K3 surface) to collapsing holomorphic discs with boundaries belonging to fibers of the dual torus fibration (see Section 5.1 and [KoSo]). Such “bad” fibers are Lagrangian tori, but they do not correspond to objects of the Fukaya category (A-branes in terminology of physicists). There are infinitely many such fibers and hence infinitely many lines. We will axiomatize this piece of data in Section 9. Subsequently, with each line we will associate an automorphism of the restriction of to . This will give us the above-mentioned modified canonical sheaf.
8 Model near a singular point
Here we will construct an analytic torus fibration corresponding to standard singularity (see Sections 3.2.4 and 6.4).
Let be the algebraic surface given by equation in coordinates , and be the corresponding analytic space. We define a continuous map by the formula where . Here denotes the multiplicative seminorm corresponding to the point (see Appendix A).
Proposition 4
The map is proper. Moreover
a) Image of is homeomorphic to .
b) All points of the image except of are -smooth.
Proof. Here is the plan of the proof.
- 1.
We define three open domains in three copies of the standard two-dimensional analytic torus , and continuous maps such that all points of the image are -smooth (i.e. each is an analytic torus fibration). Domains cover .
- 2.
For each we construct an open embedding .
- 3.
We construct an embedding such that each open set is homeomorphically identified with and . Moreover, -smooth points are mapped into -smooth points.
The Proposition will follow from 1)-3).
Let us describe the constructions and formulas. We start with open sets . Let us fix a number and define
Clearly . We define also a slightly modified domain as .
We define and . Then the projections are given by the formulas
In these formulas are coordinates on .
We define inclusion by the following formulas:
Let us decompose according to the sign of where is a point. It is easy to see that
From this explicit description we see that is proper and the image of is homeomorphic to .
Let us consider the embedding given by formula
One can easily check that the image of coincides with the image of , and for all . This concludes the proof of Proposition.
We can derive more from explicit formulas given in the proof.
Let us denote by the map . It is an analytic torus fibration outside of point . The induced -affine structure on is in fact the standard singular -affine structure described in Sections 3.2.4 and 6.4, as follows immediately from formulas for projections .
Let us introduce another sheaf on . It is defined as in each domain , with identifications
Let us consider the direct image sheaf . It is easy to see that on the sets and this sheaf is canonically isomorphic to . The isomorphism is given by the identification of coordinates on , and of coordinates and on . Therefore on the intersection we identify two copies of the canonical sheaf by certain automorphism of which preserves one coordinate (namely, the coordinate ). We will develop the theory of such transformations and their analytic continuations in Section 11. The explicit formulas for is
We would like to say now few words about analytic volume forms. Notice that each carries a nowhere vanishing top degree analytic form given by the formula . Then a straightforward calculation shows that is the pullback under of nowhere vanishing on analytic top degree form
Form satisfies Constant Norm Assumption, hence it gives a -affine structure on . On the other hand, the sheaf of algebras is also endowed with top-degree form , equal to in local coordinates.
Lemma 3
The -affine structure on associated with coincides with the one associated with .
Proof: Using definitions from Section 7.2 one sees immediately that the statement of the Lemma follows from the equality , which is straightoforward: .
In all the definitions and formulas in this section on can shift domains by vector for arbitrary , thus giving a map with singularity at the point .
Finally, we denote by . This will be our model for the sheaf near each point of the singular set .
9 Lines on surfaces
In this section we are going to describe axiomatically the notion of collection of lines on a surface.
9.1 Data
- a)
-
A compact oriented surface , a finite subset .
- b)
-
A -affine structure on with the standard singularities near each .
- c)
-
A set of lines. With each line there is an associated continuous map . We assume that is decomposed into a disjoint union of two subsets . Lines belonging to are called initial, while those in are called composite. We assume that for any there exists a continuous extension such that if and if .
- d)
-
A collection of covariantly constant nowhere vanishing integer-valued -forms . We assume that for in the standard coordinates near singular point we have: or for all sufficiently small , and .
- e)
-
A map (the letter stands for “parent”: one can think about these lines as “generating in a collision”).
Notice that since the form is invariant with respect to the monodromy, the condition in d) is coordinate-independent. The covector will be called a direction covector of at time . It gives rise to a half-plane
9.2 Axioms
To every we assign a pair , where is a choice of sign in (see data d) in the previous subsection). In this way we obtain a map .
- Axiom 1.
-
Map is one-to-one.
Let be a simply-connected domain, and line intersects . Let be an interval such that . Then there exists a covariantly constant closed non-zero -form in (with constant integer coefficients), such that , when both sides are restricted to .
- Axiom 2.
-
For any one has
Let , satisfy the condition . In this case we say that lines and have a collision at at the times and respectively.
- Axiom 3.
-
Under the above assumptions there are only two possibilities:
3a) either and , or
3b) covector is not proportional to . Then we may assume that . Under these conditions we require that for any coprime positive integers there exists a unique line such that , and .
In other words, and are “parents of ”, and the direction covector of at the intersection point is a primitive integral linear combination of those for and (see Figure 4).

- Axiom 4.
-
For every line there exist and such that they satisfy the condition 3b).
- Axiom 5.
-
For any there are no more than two pairs such that . In other words, there are no more than two lines intersecting at a point in .
Let mean the same as in the Axiom 3, and assume that . Let us consider the set of germs of all starting at (i.e. such that ).
- Axiom 6.
-
For any finite subset there is an orientation preserving homeomorphism of a neighborhood of onto a neighborhood of such that:
6a) Germs of oriented curves which are images of and get transformed into the germs at of coordinate axes and respectively.
6b) Germ of the image of gets transformed into the germ of the ray where .
Figure 5 illustrates this axiom.

- Axiom 7.
-
Let denotes either or . Then for any there exists such that if the line is well-defined then it belongs to .
This axiom says that any composed line appears as a result of finitely many collisions. The tree of ancestors of a given line form a tree embedded in , see Figure 6.

9.3 Example: gradient lines
Here we offer a construction of the set of lines satisfying the above axioms.
Let us use the standard as a model around each in order to fix a structure of smooth manifold on the whole surface . Let denotes the covering of such that the fiber over is .
Let us fix a generic smooth metric on . By the pull-back it gives a metric on . Notice that there is a canonical closed -form on such that , where . Using the metric we obtain dual to gradient vector field on .
For any and a choice of 1-form in local coordinates, we take the unique integral line of starting at . Set will be the set of all lines obtained in this way. Each line carries a covariantly constant closed -form . Using Axiom 2 as a definition, we obtain a canonical parametrization of each line by the time parameter . Since the metric is generic, a line cannot return to a point in .
Then we proceed inductively. If two already constructed lines meet at we produce a new integral line of with the direction covector satisfying the condition 3b) for any pair of coprime positive integers . In this way we construct a set of lines satisfying all the axioms. The only non-trivial thing to check is that for each line values of the parameter are in one-to-one correspondence with the interval . In order to see this we observe that the length of each line is infinite. Indeed, an integral curve of cannot have a limiting point in (since the flow generated by is smooth, and the lengths of tangent vectors are bounded from below because of the integrality of -forms).
We conclude that there exists a set of lines satisfying Axioms 1-7.
10 Groups and symplectomorphisms
In this section we are going to discuss the sheaf of groups of symplectomorphisms of the sheaf . Let be an open convex subset. By definition, a symplectomorphism of is an automorphism of -algebra preserving projection to and the canonical symplectic form (the latter is understood as an element of the algebra of Kähler differential forms). To each line we will assign a symplectomorphism of the restriction of to , so that the assignment will be compatible with the collision of lines. Then we are going to modify the sheaf using symplectomorphisms, associated with lines and obtain the sheaf . This sheaf will be glued with the sheaf near each point of .
10.1 Pro-nilpotent Lie algebra
Here it will be convenient to work in local coordinates on .
Let be a point, be -covectors such that . Denote by the closed angle
Let be a -algebra consisting of series , such that and for all we have:
- 1.
if then , where we identified with a covector in ;
- 2.
as long as .
Algebra is a Poisson algebra with respect to the bracket .
For an integer covector we denote by the monomial .
Let us consider a pro-nilpotent Lie algebra consisting of series
satisfying the condition
The latter condition is equivalent to .
Lie algebra admits a filtration by Lie subalgebras , , such that consists of the above series which satisfy the condition . Clearly , and .
Thus, is a topological complete pro-nilpotent Lie algebra over . We denote by the corresponding pro-nilpotent Lie group . It inherits the filtration by normal subgroups obtained from the corresponding Lie algebras.
10.2 Lie groups
For each we define a Lie subalgebra
Each is an abelian Lie algebra. It carries the induced filtration by Lie algebras . Denote by the corresponding pro-nilpotent group.
Lemma 4
For any given there exist finitely many such that for .
Proof. Indeed, for the monomial which maps non-trivially to the quotient we have: , where are non-negative integers. There are finitely many such non-negative integers and .
It follows from the Lemma that we have a natural isomorphism of vector spaces , hence the map
is well-defined and gives rise (after taking the projective limit as ) to the isomorphism .
In a similar way we define the map , the product is taken with respect to the natural order on . Namely, for any we define
and then set .
Theorem 6
Map is a bijection of sets.
Proof. Let be an integer. We claim that is a bijection of sets (this implies the proposition by taking the projective limit as ). We will prove the bijection by induction in . Case is obvious because all the groups under considerations are trivial.
We would like to prove that is a bijection assuming that is a bijection. Let be an element of and its image in . By the induction assumption there exist unique such that . Let be any liftings of to . Then , hence belongs to . The last inclusion holds because .
Next we observe that the isomorphism of abelian Lie algebras
implies an isomorphism of the corresponding abelian groups
Hence we can write uniquely , where . It follows that . Also it is now clear that this decomposition of is unique. This concludes the proof.
10.3 Function
For we will define an order function
(its meaning will become clear later) by the following inductive procedure:
- 1.
Let and be sufficiently small. Then in the standard affine coordinates near one has . We define . Then , and we can extend uniquely for all .
- 2.
Let and be parents of . In the notation of Axiom 3 we have and . Then we define . Again, using the condition and the knowledge of we can extend for .
Notice that can be thought of as affine function on the tangent space (in the induced integral affine structure). In particular, we have a half-plane defined by the inequality . The family of half-planes is covariantly constant with respect to .
Each half-plane contains strictly in its interior. Recall that at the end of Section 9.1 we defined another half-plane . It is easy to see that is the half-plane parallel to such that is on the boundary of .
10.4 Symplectomorphisms assigned to lines
In this section we are going to assign to each line a symplectomorphism
giving for each a transformation . This symplectomorphism in local coordinates will belong to the subgroup where is the slope of . More precisely, we demand that is of the form
where , operation is the Poisson bracket on and is an analytic function of one variable satisfying the following condition. Let us consider the pullback (by the exponential map) of the function to a section of the sheaf on vector space considered as a manifold with -affine structure. Then this pullback should admit an analytic continuation from to the half-plane , and obey there the bound
Let us explain the construction of , leaving the justification for the next sections.
Symplectomorphisms are constructed by an inductive procedure. Let be (in standard affine coordinates) a line in the half-plane emerging from (there is another such line in the half-plane ). Assume that is sufficiently small. Then we define on topological generators by the formula (as in Section 8)
Notice that , where is convergent for .
In order to extend to the interval , where is not small, we cover the corresponding segment of by open charts. Notice that change of affine coordinates transforms into a monomial multiplied by a constant from . Therefore extends analytically in a unique way to a global section over of the sheaf . Moreover, the norm strictly decreases as increases, and remains strictly smaller than . Hence can be canonically extended for all .
Each symplectomorphism is defined by a series which converges in the half-plane . Using the exponential map associated with the affine structure as well as estimates of , we can extend analytically into a neighborhood of .
Let us now assume that and collide at , generating the line . Then is defined with the help of factorization theorem in the group . More precisely, we set and the angle to be the intersection of half-planes . By construction elements and belong respectively to and . Then we can use the factorization Theorem 6 and write down the formula
where and the product on the right is in the increasing order. There is no clash of notations because it is easy to see that the boundary factors in the decomposition from above are indeed equal to and . Each term with corresponds to the newborn line with the direction covector . Then we set . This transformation is defined by a series which is convergent in a neighborhood of , and using the analytic continuation as above, we obtain for . The decomposition identity can be rewritten as
where each factor corresponds to half-lines at the collision point (see Figure 5), and the meaning of the identity is that the infinite composition of symplectomorphisms in the natural cyclic order on half-lines, is trivial.
11 Modification of the sheaf
11.1 Pieces of lines and convergence regions
Definition 13
A neighborhood of a point is convex if there exists an open convex which is isomorphic to by means of the exponential map associated with the affine structure on .
For let be convex neighborhoods of such that is relatively compact in . Let . Then there is a natural embedding .
Definition 14
A piece of defined by the pair is an element of the image of the set of connected components into under the above embedding.
In plain words a piece of is an equivalence class of a connected interval of . Two connected intervals are equivalent if they are contained in a larger connected interval of . The sole purpose of the introduction of the notion of a piece is to avoid some pathology. Namely, for any pair as above, any and any , there is only a finite number of pieces of in which have points with time parameter .
Let be a piece of defined by a pair . Then one can define an affine function in the following way. Let be such that belongs to . Since is convex, there is a unique continuation of to . This is an affine function which does not depend on the choice of . We will denote it by .
For any germ of a symplectomorphism at a point we define its convergence region as the maximal convex subset such that the pullback extends to . Since the definition of (and hence its convergence region) is covariant with respect to the affine connection we have the following result:
Proposition 5
Let belongs to a line . Then the convergence region of at contains an open half-plane .
It is clear that one can define convergence regions for symplectomorphisms associated with pieces of lines, and a similar property holds for them.
11.2 Main assumptions, and an apology
Let us suppose that our collection of lines satisfies the following assumptions:
- Assumption A1
-
There is a smooth metric and a collection of balls with centers at such that each ball contains exactly two lines outcoming of .
- Assumption A2
-
There exists such that for any the distance in between and the boundary of is greater or equal to .
We are going to show that such a collection does exist in Section 11.
Assumptions A1 and A2 are very artificial, they do not hold in physical picture which is the main motivation for the construction. It is quite possible that they can be weakened or even omitted. The main purpose of introducing them here is the possibility to define the sheaf of analytic functions by simple gluing. In complex geometry it is similar to the gluing of closed Riemann surfaces with boundaries by the mean of real-analytic identifications of the boundaries. It is well-known that one can replace real-analytic maps by smooth ones (or even by quasi-symmetric continuous maps). Maybe the rest of this section is unnecessary, and unpleasant technical arguments in Section 11.5 can be avoided.
11.3 Infinite product and its convergence
Denote by the set of all points of all lines. It has measure zero. Let be a point of . We consider two convex neighborhoods of such that is relatively compact in .
For any two points belonging to , and a path joining and in , we would like to define an infinite ordered product of transformations , where factors correspond to the intersection points of with all possible pieces relative to . Factors in the infinite product are ordered according to the time parameter of , the sign corresponds to the mutual position of orientations of and a piece at the intersection point.
In order to give a precise meaning to the infinite product the neighborhood of should be sufficiently small. Then we will have an analytic continuation of symplectomorphisms to , and the convergence of the infinite product. We are also going to prove that the product is independent of the choice of path . In order to achieve these goals it suffices to assume:
- C1
-
for any such the set is contained in ;
- C2
-
for any there is only a finite number of pieces of lines in such that .
Theorem 7
Assume two above conditions. Then the product defining converges at every point of and in fact gives an element of . Moreover, the product does not depend on the choice of path , and for any satisfies the relation .
Proof. Condition C1 implies that all transformations admit an analytic continuation to . Let us introduce a decreasing filtration by positive real numbers on group by the formula
This is a complete filtration, and condition C2 implies that in any quotient only a finite number of elements are non-trivial. Therefore we can define the product in the quotient group.
In order to prove independence of , we consider the quotient group , and the finite -dimensional CW-complex (graph) consisting of finitely many pieces , such that in the quotient. For each vertex of the graph there is a natural cyclic order on the edges incident to . The product taken in the cyclic order over the set of edges incident to is equal to (this follows from the construction of via factorizations). Since is simply-connected, we conclude that the image of in does not depend on . Using completeness of the filtration we see that does not depend on . Proof of the identity is similar.
Theorem 8
Assumptions A1 and A2 imply that for any there exist neighborhood (and also ) satisfying conditions C1 and C2.
Proof. Assumption A1 implies that the result near any singular point , as there are only two lines near . If we are far from then obviously A2 implies C1.
In order to check C2 we prove the following lemma
Lemma 5
Under Assumptions A1 and A2, for any the set
consists of a finite number of intervals.
Proof: We proceed by induction in “complexity of the line”. Let be the infimum of where has a collision at time . This number is strictly positive because the number of initial lines is finite, and by A1 there is no collisions at small times. Observe that the value of at the beginning of any composite line is greater or equal to the sum . Therefore the inequality in the lemma implies that the number of collisions is bounded from above by . Also we have an upper bound on integer coefficients in each collision (see Axiom in Section 9.2). Let us observe that the length of each edge of the ansector tree of is also bounded from above by , for some absolute constant . Hence we have only finitely many possibilities for intersections.
For point which is far from we chose as a neighborhood of radius where is constant from Assumption A2. Then for any point of a line we will have the inclusion
This implies that in for the corresponding piece is bounded below by
Since (by the last lemma) there exists only a finite number of pieces intersecting such , we obtain convergence condition C2.
11.4 Construction of the modified sheaf
For any point and a neighborhood satisfying conditions C1 and C2 we define the sheaf as the result of the identification of copies of the sheaf labeled by points , by isomorphisms . It follows from formulas in Section 8 that near singular points one can identify canonically this sheaf with the restriction of the sheaf to a punctured neighborhood of .
Proposition 6
For the modified sheaf one has a canonical nowhere vanishing section of the associated sheaf of -analytic -forms.
The -affine structure on associated with coincides with the initial one .
Proof. Existence of follows from the fact that all modifications associated with lines are symplectomorphisms. In order to finish the proof it suffices to check that the modification associated with a line does not change the -affine structure on . In local coordinates we may assume that and the modification is of the form , where is convergent in an appropriate domain. We need to check that the automorphism acts trivially on the quotient sheaf (see Section 7.2 for the notation). This check reduces to the calculation of
The latter is equal to because belongs to and therefore has no constant term.
Thus, we have a solution of the Lifting Problem under Assumptions A1 and A2.
11.5 Construction of the collection of lines
We would like to show that there exists a smooth metric and a collection of lines satisfying the Assumptions A1 and A2.
Let be an arbitrary smooth metric, flat near singular points. We define germs of lines in such a way that for each in local coordinates these lines are given by and . The metric will coincide with in a sufficiently small neighborhood of the singular set. Hence Assumption A1 will be satisfied.
In order to construct the whole family of lines we introduce a -dimensional manifold consisting of pairs where and is a half-plane in whose boundary contains zero. Here is a larger neighborhood of .
We would like to construct a smooth section of the pull-back to of the tangent bundle satisfying the following conditions:
- 1.
for any one has ;
- 2.
for any the map is an orientation-preserving diffeomorphism
- 3.
for every there exists a smooth extension of the piece of in to a larger piece intersecting such that
for such that ;
Let us associate with the section a nowhere vanishing vector field on in the following way:
- •
For each the vector is tangent to the horizontal distribution associated with the flat connection (the one which defines the affine structure on ).
- •
Projection of to coincides with , where .
Clearly these conditions determines uniquely. Now we formulate last condition:
- 4.
there exits such that for almost all (in the sense of Baire category) initial values the integral curve of starting at reaches the pullback of in finite time.
Using the vector field we will construct (under certain genericity assumptions) a set of lines satisfying Assumption A1. Namely, the data consisting of a line and an integer-valued -form (see Section 9) will be an integral line of .
We are going to construct lines by induction by the number of collisions. Lines will be constructed using condition 3. The genericity assumption mentioned after the condition 4 is the assumption that no more than two lines collide and that initial values for newborn lines will be sufficiently generic. Conditions 1 and 4 plus genericity imply that one can parametrize any line by the new “time” such that the Axiom 2 is satisfied. Axiom 6 follows from the condition 2. Other axioms and the Assumption A1 will be satisfied automatically.
Now we would like to discuss Assumption A2.
Proposition 7
Suppose that the metric and field described above are such that for any there exists such that
where is the normal unit vector to directed inside and is the covariant derivative of the metric considered as a symmetric tensor on the cotangent bundle.
Then the Assumption A2 is satisfied.
Proof. In order to satisfy Assumption A2 it suffices to find such that for any and any half-plane with the distance , and another half-plane parallel to such that , one has the following property: if is the half-plane obtained from by a small covariant (with respect to the affine connection ) shift in the direction of , then
Here etc. denotes the induced flat metric on the tangent space . This property guarantees that the condition will propagate along the line. For a new line obtained as a result of collision of and at the times and respectively one has
since contains the intersection point , see Figure 7.

One can easily see that the infinitesimal inequality from above is equivalent to
(the change of the distance consists of two summands: one corresponds to the shift along with the fixed metric, and the other one corresponds to the change of the metric). Taking the limit we arrive to the inequality for the covariant derivative of the metric with .
Now our goal is to construct the field of directions and the metric satisfying the conditions 1–4 and the inequality from the last Propostion. This will conclude the construction of the set of lines satisfying the Assumptions A1 and A2.
Since is a boundary of the convex set, we can locally model it by the graph of function such that , . We may assume that is the upper half-plane. Then we take
We extend this local model of near to in such a way that conditions 1 and 2 are satisfied. It is clear that we can satisfy conditions 3,4 as well by taking a small perturbation of . On Figure 8 there is a picture of the field .

For an arbitrary choice of the metric we have for all . The problem with inequality
arises only as the point approaches . Indeed, in this case the vector can be very close to the tangent vector to .
Lemma 6
With the above choice of assume that the metric satisfies for any the condition
where is the unit tangent vector to and is the normal vector to (all scalar products and lengths are taken with respect to the metric ).
Then there exists such that
for all .
Proof. We need to check that the ratio
is bounded for .
It suffices to prove the Lemma assuming that is the parabolic domain and is the upper half-plane. The vector field is given for by the formulas
The denominator is equal to near .
The numerator is equal to
where and are two -functions.
By assumption of the Lemma we have . Therefore where is a convenient local coordinate near the point . Notice also that .
Now we can estimate first summand of the numerator assuming that and are sufficiently small. As we have seen, it is bounded by
There are three cases which we need to consider.
a) If then .
b) if then .
c) If the .
We see that the numerator is bounded. This concludes the proof of Lemma.
Finally, we have the following result.
Lemma 7
There exists metric satisfying the conditions of Lemma 6.
Proof: First of all, the condition on from Lemma 6 is the condition on a loop of scalar products on 2-dimensional spaces, here . We can write where and is a smooth function. Then we have
The equation of Lemma 6 gives . The RHS of this expression is known as long as we know . Hence we can say that , where is a 1-form depending on the restriction . We see that it suffices to find such that (then and hence does exist).
Let us consider the functional . We can interpret a metric as a point in the Lobachevsky plane . More precisely, let us consider the space of pairs where is a positive quadratic form on such that and is a half-plane in (the meaning of is the inward oriented tangent half-plane to at point ). This space is naturally diffeomorphic to . The latter manifold can be identified in -equivariant way with the manifold consisting of pairs , where and belongs to the absolute. Hence is (locally) a non-parametrized path in (it would be a global path, if the bundle over given by the all metrics on with the determinant was trivial).
Next we observe that the variation , where is a -dimensional surface bounded by the paths defined by and , and is a canonical -invariant -form on . One can show that even by a small variation of the path defined by we can make an arbitrary real number. In particular, we can find such that . This concludes the proof of Lemma 6.
Summarizing, we have constructed a set of lines satisfying the Assumptions A1 and A2. This concludes the proof of Theorem 5. Thus we have obtained a solution of the Lifting Problem, which is a -analytic K3 surface.
11.6 Independence and uniqueness
It is natural to ask how the above construction of the -analytic K3 surface depends on the choice of the set of lines. We know that the “periods” of (they are encoded in the initial -affine structure) do not depend on (see Sections 7.3, 10.4). In the light of Torelli theorem (see Appendix B) it is natural to formulate the following conjecture.
Conjecture 11
The isomorphism class of the pair does not depend on the choice of the set of lines.
More precisely, the change of corresponds to the change of the projection (see Section 7.3).
Remark 4
For and with the standard singular -affine structure we have constructed a -analytic K3 surface depending on parameters in . More precisely, we have a -dimensional -analytic space of conjugacy classes of representations
such that the monodromy around each singular point is conjugate to the pair where is equal to
(compare with Section 3.3).
11.7 Remark on the case of positive and mixed characteristic
Our construction of works even without the assumption where is the residue field of . This can be explained from the point of view of factorization theorem (see Section 10.4). It turns out that symplectomorphisms which appear in the infinite product in the RHS of the factorization theorem are infinite series whose coefficients are integer polynomials in the coefficients of the “parent” symplectomorphisms.
For example, let and be two power series convergent when . Let us consider two symplectomorphisms: and and decompose into the infinite ordered product . Here
where . Then one can check that for any coprime and any one has
This implies that our construction works when one replaces by arbitrary commutative ring endowed with a complete non-trivial valuation .
11.8 Further generalizations
First of all, one can introduce a small parameter of noncommutativity in the picture, coordinates will not commute but instead satisfy the relation
For such a noncommutative analytic torus one can still define sheaf on by the “same” formula as in the commutative case:
where is connected. Also one can construct a non-commutative deformation of the model sheaf near the singular point. All arguments with the groups work as well. In this way we will obtain a kind of quantized K3 surface over a non-archimeden field.
Secondly, we believe that one can generalize our construction to higher dimensions. Instead of lines there will be codimension one walls which should be flat hypersurfaces with respect to -affine structure and carry foliations by parallel lines. Generically on the intersection of two such foliated hypersurfaces one can “separate” variables into the product of a purely 2-dimensional situation studied in the present paper, and dummy variables. Presumably everywhere except a countable union of codimension 2 subsets one can use 2-dimensional factorization and define gluing volume preserving maps. One can hope that by a kind of Hartogs principle the sheaf will have a canonical extension to the whole space .
Appendix A Analytic geometry
In this section we collect several facts and definitions about rigid analytic spaces and Clemens polytopes. Some of them are well-known, the rest is borrowed from [KoT].
We always work over a complete non-archimedean local field . The field carries a valuation map such that .
We will assume that the valuation is non-trivial. The ring
is called the ring of integers of . The residue field is defined as , where is the maximal ideal in .
Our main example is the field of Laurent series in one variable. In this case , as long as .
A.1 Berkovich spectrum
We refer the reader to [Be1] for the general definition of an analytic space and more details. In this Appendix we restrict ourselves to analytic spaces associated with algebraic varieties (although we use the general definition in the paper as well).
Let be a commutative unital finitely generated -algebra. The underlying set of the Berkovich spectrum can be defined in two ways. First one uses valuations (or, equivalently, multiplicative seminorms).
Definition 15
(Valuations) A point of is an additive valuation
extending , i.e. it is a map satisfying the conditions
- •
;
- •
;
- •
for all and all .
Having a valuation and a real number one can define the multiplicative seminorm . In particular, in the previous definition one can take seminorms instead of valuations . The reader has noticed that in the main body of the paper, for we often took . It is easy to translate the definition of Berkovich spectrum to the language of multiplicative seminorms. We use it freely in the paper.
The second way to define uses evaluations (characters).
Definition 16
(Evaluation maps) A point of is an equivalence class of homomorphisms of -algebras
where is a complete field equipped with a non-archimedean valuation, which extends the valuation , and such that is generated by the closure of the image of .
The field is determined by in a canonical way. We define for and the “value” as the image .
In order to pass from the first description of to the second, starting with a valuation one defines the field as the completion of the field of fractions of , where .
Definition 17
The topology on is the weakest topology such that for all the map
is continuous.
An element defines a function , where is the non-archimedean valuation field, which is the completion of the field of fractions of the domain . Since each carries a seminorm, we obtain a function .
A fundamental system of neighborhoods of a point is parametrized by the following data: a finite collections of functions
and numbers
such that , The corresponding neighborhood consists of points such that for all and .
Let us assume that elements generate , i.e.
where is an ideal. Let us consider the algebra of series
with constants , absolutely convergent when variables satisfy the above inequalities. The quotient of this algebra by the topological closure of ideal is the algebra .
As in the case of schemes we can glue into ringed spaces called analytic spaces (or rigid analytic spaces). Moreover we get a functor
Proposition 8
The space
a) is a locally compact Hausdorff space as long as is separated;
b) has the homotopy type of a finite -complex;
c) is contractible if has good reduction with irreducible special fiber.
Example 1
Let be the affine line. The analytic space contains, among others, points of the following types:
- •
;
- •
for define
This gives an embedding .
We see that contains, in a sense, both -adic and real points.
Define the cone over as
We interpret a point of as a -point of , where is a complete field with the -valued valuation
whose restriction to is proportional to . The set of points such that the valuation is -valued is denoted by .
A.2 Algebraic torus and the logarithmic map
Here we will describe explicitly the main example for our paper. Let be an algebraic torus. and the corresponding analytic space.
Firstly, we define an embedding . For real vector the corresponding point will be described in terms of valuations.
For every Laurent polynomial we set
Secondly, we define a projection by formula
The fiber over a point can be identified with the set of such seminorms that . We see is a kind of torus fibration77 7 This is the origin of the term “analytic torus fibration” introduced in Section 4.1.. Moreover, .
For any open connected the -algebra of analytic functions on consists of series with coefficients such that for any we have when . It is easy to see that where is the convex hull of .
The sheaf (canonical sheaf) plays an important role in the paper (see Sections 4.1, 7.3, 8).
A.3 Clemens polytopes
Let be a smooth proper scheme over the non-archimedean field . We assume that carries a discrete valuation such that .
Definition 18
A model of is a scheme of finite type flat and proper over , together with an isomorphism . Denote the special fiber of by
A model has no nontrivial automorphisms. Thus, the stack of equivalence classes of models is in fact a set, which we denote by . It carries a natural partial order. Namely, we say that if there exists a map over . Such a map is automatically unique.
Definition 19
A model has normal crossings if the scheme is regular and the reduced subscheme is a divisor with normal crossings.
By the resolution of singularities, in the case we know that every model is dominated by a model with normal crossings.
Definition 20
A model has simple normal crossings (snc model for short) if
- •
it has normal crossings;
- •
all irreducible components of are smooth and
- •
all intersections of irreducible components of are either empty or irreducible.
The set of equivalence classes of snc models will be denoted by . It is a filtered partially ordered set. The order is given by dominating maps of models which give the identity automorphism on the generic fiber.
It is easy to show that starting with any model with normal crossings and applying blow-ups centered at certain self-intersection loci of the special fiber we can get a snc model. In what follows we use snc models only. This choice is dictated by convenience and not by necessity. Working with snc models has the advantage that all definitions and calculations can be made very transparent. The reader can consult [Be2] for the approach in the general case, without the use of the resolution of singularities.
Let be an snc model and the set of irreducible components of . Denote by the divisor corresponding to . For any finite non-empty subset put
By the snc property the set is either empty or is a smooth connected proper variety over of dimension . For a divisor we denote by the order of vanishing of at , where is an uniformizing element, . Equivalently, is the multiplicity of in .
Definition 21
The Clemens polytope is the finite simplicial subcomplex of the simplex such that is a face of iff .
Clearly, is a nonempty connected CW-complex. We will also consider the cone over :
Analogously, we can define .
We identify with the following subset of :
Obviously, we can also describe as a quotient of :
A.4 Simple blow-ups
Let be an snc model, a non-empty subset and a smooth irreducible variety of dimension less or equal than . Let us assume that intersects transversally (in ) all subvarieties of (for ), and that all intersections are either empty or irreducible. It is obvious that the blow-up of with the center at is again a snc model.
Definition 22
For a pair of snc models as above we say that is obtained from by a simple blow-up. If we say that we have a simple blow-up of the first type. Otherwise (when ), we have a simple blow-up of the second type.
Let us describe the behavior of under simple blow-ups. To the set of vertices we add a new vertex corresponding to the divisor obtained from :
The degree of the new divisor is (for both the first and the second type)
For blow-ups of the first type we have automatically . Here is the list of faces of :
1) for ;
2) for ;
3) the vertex .
For blow-ups of the second type the list of faces
of is
1) for ;
2) for ;
3) the vertex .
On can deduce from results [AKMW] the following
Theorem 9
(Weak factorization) Assume that . Then for any two snc models there exists a finite alternating sequence of simple blow-ups
Corollary 3
Simple homotopy type of does not depend on the choice of a snc model .
A.5 Clemens cones and valuations
Let be a snc model of . We define a map
such as follows. For such that let us consider a point
and an affine Zariski open subset containing the generic point of . One can embed into the algebra of formal series where is the field of rational functions on and are equations of divisors . We define a valuation of by the formula
We define to be the image of the point in . It is easy to check that the element does not depend on the choice of the open subset .
The following proposition is obvious:
Proposition 9
The map is an embedding.
We will denote also by the induced embedding .
A.6 Clemens cones and paths
For a model we can interpret elements of as paths in , i.e. equivalence classes of maps
where is the ring of integers in a field with discrete valuation in , such that the image of does not lie in . We define the map
as
where is the multiplicity of the intersection of the path with the divisor .
The following proposition can be derived from [Be1].
Proposition 10
The map extends uniquely to a continuous -equivariant map . The map is a surjection.
We denote by the map induced by .
Let be a dominating map of models. Let us denote by the multiplicity of a divisor in the proper pull-back of . We define by the formulas . Let be the corresponding by map of Clemens polytopes.
Then we have the following result, which is easy to prove.
Lemma 8
For any dominating map of models we have
Corollary 4
For dominating maps we have
Theorem 10
For any algebraic the analytic space is a projective limit over the partially ordered set of snc models of Clemens polytopes . The connecting maps are .
With any meromorphic at family of smooth complex projective varieties one can associate a variety over the field . It is easy to see that for any snc model one can canonically complete the family by adding as the fiber over . The total space is not a complex manifold by just a Hausdorff locally compact space which maps properly to the dick . Passing to the projective limit we see that one can compactify the family at by .
Appendix B Torelli theorem for K3 surfaces
Here we recall the classification theory of complex K3 sufaces (see [PSS] and its extension to non-algebraic case in [LP]). Let be a complex K3 surface, i.e. smooth connected complex manifold with which admits a nowhere vanishing holomorphic -form , and such that .
It is known that the group endowed with the Poincare pairing is isomorphic to the lattice
of signature .
Complex -dimensional vector space satisfies the condition for any non-zero vector . Finally, it is known that admits a Kähler metric, and Kähler cone of all Kähler metrics on is an open subset of . In fact is a connected component of the set , where is the hyperplane orthogonal to .
Axiomatizing these data we arrive to the following definition.
Definition 23
K3 period data is a quadruple consisting of a free abelian group , a symmetric pairing , a -dimensional complex vector subspace and a set satisfying the following conditions:
- 1.
;
- 2.
is isomorphic to ;
- 3.
for any one has and ;
- 4.
the set is a connected component of , where and is the hyperplane orthogonal to .
The K3 period data form a groupoid. On the other hand, K3 surfaces also form a groupoid (morphisms are isomorphisms of K3 surfaces). Then classical global Torelli theorem can be formulated in the following way.
Theorem 11
Groupoid of K3 surfaces is equivalent to the groupoid of K3 period data.
In particular the automorphism group of a K3 surface is isomorphic to the automorphism group of its period data.
More generally one can speak about holomorphic families of K3 surfaces over complex analytic spaces. For a K3 surface over an analytic space the period data consist of a local system of integral lattices pointwise isomorphic to , a holomorphic line subbundle of which is isotropic with respect to the symmetric pairing , and satisfies pointwise the condition , and an open subset of the total space of the bundle over with the fibers ( is the orthogonal complement) satisfying pointwise the condition 4) from the definition of K3 period data. Then Torelli theorem holds for families as well.
References
[AKMW], D. Abramovich, K. Karu, K. Matsuki, J. Wlodarczyk, Torification and factorization of birational maps, J. Amer. Math. Soc. 15 (2002), no. 3, 531–572, and preprint math.AG/9904135.
[Ar] V. Arnold, Mathematical methods of classical mechanics, Springer, 1997.
[Au] M. Audin, Spinning tops. A course on integrable systems, Cambridge Studies Adv. Math., vol. 51, Cambridge University Press, 1996.
[Be1] V. Berkovich, Spectral theory and analytic geometry over non-archimedean fields, AMS Mathematical Surveys and Monographs, n. 33, 1990.
[Be2] V. Berkovich, Smooth p-adic analytic spaces are locally contractible, Inv.Math., 137, 1–84 (1999).
[Be3] V. Berkovich, Smooth p-adic analytic spaces are locally contractible, II, to appear.
[GS] M. Gross, B. Siebert, Mirror Symmetry via Logarithmic degeneration data, I, preprint math.AG/0309070.
[GW] M. Gross, P. M. H. Wilson, Large complex structure limits of K3 surfaces, J. Differential Geom. 55 (2000), no. 3, 475–546, and preprint math.DG/0008018.
[HZh] C. Haase, I. Zharkov, Integral affine structures on spheres and torus fibrations of Calabi-Yau toric hypersurfaces I, preprint math.AG/0205321.
[KN] S. Kobayashi, K. Nomizu, Foundations of differential geometry, vol. 1, John Wiley and Sons, 1963.
[Ko] M. Kontsevich, Homological algebra of mirror symmetry, Proc. ICM Zürich, vol.1, 1994, and preprint math.AG/9411018.
[KoSo] M. Kontsevich, Y. Soibelman, Homological mirror symmetry and torus fibrations, in Symplectic geometry and mirror symmetry (Seoul, 2000), World Sci. Publishing, River Edge, NJ, 2001, 203–263, and math.SG/0011041.
[KoT] M. Kontsevich, Yu. Tschinkel, Non-archimedean Kähler geometry, in preparation.
[LeS] N. C. Leung, M. Symington, Almost toric symplectic four-manifolds, preprint math.SG/0312165.
[LYZ] J. Loftin, S.-T. Yau, R. Zaslow, Affine manifolds, SYZ geometry and the “Y” vertex, preprint math.DG/0405061.
[LP] E. Looijenga, C. Peters, Torelli theorems for Kähler K3 surfaces, Comp. Math. 42 (1981), 145–186.
[Mi] G. Mikhalkin, Amoebas of algebraic varieties and tropical geometry, preprint math.AG/0403015.
[Mor] D. R. Morrison, Mathematical aspects of mirror symmetry, in Complex algebraic geometry (Park City, UT, 1993), IAS/Park City Math. Ser., 3, Amer. Math. Soc., Providence, RI, 1997, 265–327, and preprint alg-geom/9609021.
[PSS] I. I. Pjateckiĭ-S̆apiro, I. R. S̆afarevic̆, A Torelli theorem for algebraic surfaces of type K3, Math. USSR Izvestija 5 (1971), No. 3, 547–588.
[SYZ] A. Strominger, S.-T. Yau, E. Zaslow, Mirror symmetry is T-duality, Nucl. Phys. B479 (1996), 243-259.
[Tyu] A. Tyurin, On Bohr-Sommerfeld bases, Izv. Math. 64 (2000), no. 5, 1033–1064, and preprint math.AG/9909084.
[Zu] N. Zung, Symplectic topology of integrable Hamiltonian systems, II: Topological classification , Compositio Math., 138:2 (2003), 125-156.
Addresses:
M.K.: IHES, 35 route de Chartres, F-91440, France
maxim@ihes.fr
Y.S.: Department of Mathematics, KSU, Manhattan, KS 66506, USA
soibel@math.ksu.edu