---
title: 'Dolgachev Surfaces: Fibrations & Categories'
url: https://www.emergentmind.com/topics/dolgachev-surfaces
type: topic
---

# Dolgachev Surfaces: Fibrations & Categories

Dolgachev surfaces are minimal elliptic surfaces over $\PP^1$ with vanishing irregularity and geometric genus, equipped with exactly two multiple fibers of relatively prime multiplicities. In the formulations used in the recent literature, a Dolgachev surface of type $(p,q)$ or $D_{p,q}$ is a smooth projective surface $S$ or $Y$ with $p_g=q=0$, Kodaira dimension $\kappa=1$, and an elliptic fibration whose only multiple fibers have multiplicities $p$ and $q$; when $\gcd(p,q)=1$, the surface is simply connected [1506.05213, 2605.06668, 2509.07467]. The subject now spans several tightly linked themes: classical logarithmic transformations of rational elliptic surfaces, $\QQ$-Gorenstein smoothings of singular rational surfaces, classifications of rational homology disk degenerations, compactified moduli of elliptic surfaces with multiple fibers, and derived-categorical phenomena such as exceptional collections, quiver embeddings, and phantom subcategories [1506.05213, 2204.13225, 2605.06668, 2509.07467, 2310.13319].

## 1. Definition and core invariants

A Dolgachev surface of type $(p,q)$ is described as a simply-connected, minimal, nonsingular projective surface with
\[
p_g(S)=h^2(\mathcal O_S)=0,\qquad q(S)=h^1(\mathcal O_S)=0,
\]
Kodaira dimension $\kappa(S)=1$, and an elliptic fibration $f\colon S\to\PP^1$ whose only multiple fibers have multiplicities $p$ and $q$; the multiplicities are assumed coprime, with $p>q>0$ in one formulation and $p,q\ge 2$ in another [1506.05213, 2204.13225]. A parallel notation writes the two multiple fibers as
\[
F_p=p\cdot C_p,\qquad F_q=q\cdot C_q,
\]
for simple smooth curves $C_p$ and $C_q$ [2204.13225].

Several numerical formulas recur across the literature. For a general fiber $F$ of the elliptic fibration, the canonical bundle formula is written
\[
K_S \equiv F-\frac1pF-\frac1qF
\]
in the sense of $\QQ$-divisors, equivalently
\[
K_S \equiv \left(1-\frac1p-\frac1q\right)F
\]
[1506.05213]. In a different notation, the canonical divisor is recorded as
\[
K_Y \equiv (p+q-2)F-C_p-C_q
\]
[2204.13225], and also as
\[
K_S \sim_{\QQ} f^*\mathcal O_{\PP^1}(-1)+(p-1)F_p^{\rm red}+(q-1)F_q^{\rm red}
\]
[2509.07467]. The same sources also record
\[
K_Y^2=0,\qquad \chi(\mathcal O_Y)=1,
\]
and, by the Noether formula,
\[
12\,\chi(\mathcal O_S)=K_S^2+e(S),
\]
so that $e(S)=12$ when $p_g=q=0$ [2605.06668, 2509.07467].

The topological behavior is controlled by the multiplicities. One source states that $\pi_1(Y)=1$ if $\gcd(a,b)=1$, and more generally that $\pi_1$ is cyclic of order $\gcd(a,b)$ [2605.06668]. The same paper notes that $\kappa(Y)=1$ if $ab>4$, while the case $(a,b)=(2,2)$ yields Enriques surfaces [2605.06668]. A section exists exactly when one multiplicity is $1$, in which case the surface is a Halphen surface of index $\max(a,b)$ [2605.06668]. For the generic Dolgachev surface of type $(p,q)$, the Picard number is recorded as $\rho(S)=10$ [2509.07467].

## 2. Constructions from rational elliptic surfaces and singular models

The classical construction starts from a rational elliptic surface obtained by blowing up the nine base points of a pencil of plane cubics. If
\[
\langle C_1,C_2\rangle\subset |\mathcal O_{\PP^2}(3)|
\]
is such a pencil, then the blow-up $Y\to\PP^2$ at its nine base points carries an elliptic fibration whose general fiber $C_0$ is a smooth genus-one curve. Performing logarithmic transforms of orders $p$ and $q$ at two distinct smooth fibers produces a simply-connected minimal elliptic surface with two multiple fibers of multiplicities $p$ and $q$, recovering the canonical bundle formula and the invariants $p_g=q=0$ and $\kappa=1$ [1506.05213]. A related formulation chooses two points $x,y\in\PP^1$ and torsion line bundles
\[
\eta_x\in\Pic^0(F_x),\qquad \eta_y\in\Pic^0(F_y)
\]
of exact orders $p$ and $q$; the resulting elliptic surface $S(p,q)\to\PP^1$ is isomorphic to the original rational Jacobian surface away from $x$ and $y$, but acquires multiple fibers of orders $p$ and $q$ there [2509.07467].

A second construction replaces logarithmic transforms by $\QQ$-Gorenstein smoothing of a singular rational surface. Starting again from a rational elliptic surface $Y$, one contracts two chains of curves to obtain a normal projective surface
\[
\pi\colon Y\to X
\]
with two Wahl $(T_1)$ singularities, of types
\[
P_1\sim \frac14(1,1),\qquad P_2\sim \frac1{n^2}(1,na-1),
\]
with $\gcd(n,a)=1$ [1506.05213]. In the Picard–Noether group one has
\[
\pi^*K_X \equiv \left(\frac12-\frac1n\right)C_0,
\]
so $K_X$ is nef, $K_X^2=0$, and $H^i(\mathcal O_X)=0$ for $i>0$ [1506.05213]. Wahl’s theory and Manetti’s vanishing $H^2(T_X)=0$ then yield a one-parameter $\QQ$-Gorenstein smoothing
\[
\varphi\colon \mathcal X\to\Delta,\qquad \mathcal X_0=X,\qquad \mathcal X_t=X^{\rm gen}\ (t\ne 0),
\]
whose general fiber is a minimal simply-connected elliptic surface with $p_g=q=0$, $\kappa=1$, and two multiple fibers of orders $2$ and $n$ [1506.05213].

The deformation-theoretic analysis also gives a detailed model of the central fiber after base change and weighted blow-up. After a base change of order $a$, one resolves the family near the singular points and obtains a central fiber
\[
\widetilde{\mathcal X}_0=\widetilde X_0\cup W_1\cup W_2,
\]
where
\[
W_1\simeq \PP^2,\qquad W_2\simeq \PP(1,n-1,1),
\]
and the components meet $\widetilde X_0$ along smooth rational curves $Z_1,Z_2$ [1506.05213]. This reducible model is the starting point for the later construction of exceptional bundles by gluing.

| Construction | Starting data | Output |
|---|---|---|
| Logarithmic transforms | Rational elliptic surface from the blow-up of nine base points of a cubic pencil | Minimal elliptic surface with multiple fibers of orders $p,q$ |
| $\QQ$-Gorenstein smoothing | Singular rational surface with two Wahl singularities | General fiber $X^{\rm gen}$ of Dolgachev type |
| Weighted blow-up central fiber | Base change and local resolution of the smoothing | $\widetilde X_0\cup W_1\cup W_2$ with $W_1\simeq\PP^2$ and $W_2\simeq\PP(1,n-1,1)$ |

Taken together, these constructions show that Dolgachev surfaces can be accessed either through explicit elliptic-surface surgery or through the deformation theory of singular surfaces. The latter viewpoint is the one used most heavily in the derived-categorical papers [1506.05213, 2204.13225].

## 3. Singular fibers, rational homology disk degenerations, and compactified moduli

Beyond the two multiple fibers, the remaining fibers are constrained by the elliptic fibration. One source states that all non-multiple fibers are irreducible of Kodaira type $I_n$ $(n\ge 0)$, $II$, $III$, or $IV$ [2605.06668]. In the compactification framework for elliptic surfaces with a multiple fiber, the remaining singular fibers may be any reduced Kodaira fibers of types
\[
I_n,\ II,\ III,\ IV,\ I_n^*,
\]
subject to the Euler–Kimura constraint that the total contribution to $e(S)$ is $12$ [2509.07467].

A recent classification concerns $\QQ$HD degenerations of Dolgachev surfaces. For a $\QQ$ smoothing whose general fiber belongs to $D_{a,b}$ with $ab>4$, the central fiber is a normal surface with one non–Du Val weighted-homogeneous $\QQ$ singularity and one Wahl singularity of type
\[
\frac1{b^2}(1,b-1)
\]
[2605.06668]. More precisely, the central fiber carries a minimal elliptic fibration with two multiple fibers of multiplicities $a$ and $b$, one supporting the non-log-canonical $\QQ$ singularity and the other the Wahl singularity. Up to “sliding,” the non-log-canonical singularities arise from $M$-modifications of an elliptic fiber of type $II$, $III$, or $IV$ [2605.06668].

The weighted-homogeneous singularity is encoded by a star-shaped resolution graph with central curve $C_0$ of self-intersection $-e_0$ and Hirzebruch–Jung chains on the legs. Its Seifert invariants are written
\[
(0;e_0;q_1/m_1,q_2/m_2,\dots,q_t/m_t),
\]
with $m_i/q_i=[e_{i,1},\dots,e_{i,r_i}]$, and the discrepancy of the central curve is
\[
d(C_0)=-1-\chi/e,\qquad \chi=-2+t-\sum 1/m_i,\qquad e=e_0-\sum q_i/m_i.
\]
The condition $0<\chi/e<1$ forces one non-log-canonical leg of length at least $3$ [2605.06668]. Every one-point $\QQ$HD degeneration of $D_{a,b}$ appears in this classification [2605.06668].

The same paper constructs minimal semi-log-canonical birational models of these degenerations. The procedure consists of a Seifert partial resolution extracting only the central curve
\[
\Delta\simeq \PP^1,
\]
followed by a semistable flip in a log-terminal threefold. The result is a central fiber
\[
W^+\cup Z^+
\]
glued along $\Delta$, with one new Wahl singularity on each component and with $K+\Delta$ nef [2605.06668]. Unobstructedness is proved by vanishing of the local-global obstructions, ultimately using $H^2(W,T_W)=0$ and vanishing statements on the components of the slc limit [2605.06668].

From the moduli viewpoint, Dolgachev surfaces of fixed type $(p,q)$ with $\gcd(p,q)=1$ fit into an irreducible $10$-dimensional parameter space $T$ of quintuples
\[
(X,x,y,\eta_x,\eta_y),
\]
where $X\to\PP^1$ is a rational elliptic surface of index $1$, $x\ne y$ are points over which the fibers are smooth or of type $I_n$, and $\eta_x,\eta_y$ are torsion line bundles of orders $p$ and $q$ [2509.07467]. A stable-pair compactification is obtained by choosing a marking divisor $A$ and the boundary
\[
B:=c\bigl(F_p^{\rm red}+F_q^{\rm red}\bigr),\qquad 0<c\le 1,
\]
and then invoking Birkar’s theory to construct a projective coarse moduli space
\[
M_{2,c,v,\sigma}
\]
[2509.07467]. In this framework, when a multiple fiber degenerates to an additive Kodaira fiber, boundary components are described by semi-log-canonical limits.

Kawamata’s classification of moderate degenerations imposes further restrictions on permissible normal central fibers with $\kappa=1$. The only fiber types on such a central fiber are
\[
mI_d(r,a),\qquad II(r),\qquad III(r),\qquad IV(r),
\]
and no
\[
I_b^*,\ II^*,\ III^*,\ IV^*
\]
occur on the central fiber of a permissible normal degeneration [2509.07467]. The invariants remain constant under degeneration:
\[
K_{X_0}^2=K_{X_t}^2,\qquad \chi(\mathcal O_{X_0})=\chi(\mathcal O_{X_t}),\qquad e(X_0)=e(X_t),
\]
and the log-canonical thresholds governing wall crossing are computed for the additive types [2509.07467].

## 4. Picard groups and explicit models of type $(2,3)$

The case of the $(2,3)$-Dolgachev surface is the setting in which the Picard group and derived category have been worked out most explicitly. For the $\QQ$-Gorenstein smoothing model with singularities of types
\[
\frac14(1,1)\quad\text{and}\quad \frac19(1,2),
\]
certain divisors on the resolution $Y$ satisfying parity and divisibility conditions can be lifted to Cartier divisors on the smooth fiber $X^{\rm gen}$ by Hacking’s gluing method [1506.05213]. This produces ten divisors
\[
G_i^{\rm gen}\in \Pic X^{\rm gen}\qquad (i=1,\dots,10),
\]
defined from lifts of $p^*(2H)$ and of classes $F_i-F_j$. Under the explicit definitions, the intersection matrix
\[
\bigl((G_i^{\rm gen}\cdot G_j^{\rm gen})\bigr)_{1\le i,j\le 10}
\]
is
\[
\operatorname{diag}(-1,\dots,-1,+1),
\]
and hence
\[
\Pic X^{\rm gen}\simeq \ZZ^{10}
\]
with basis $G_1^{\rm gen},\dots,G_{10}^{\rm gen}$ [1506.05213].

A complementary explicit model of a special $(2,3)$-surface begins with $\PP^2$ with coordinates $[x:y:z]$, two lines $L_1=\{z=0\}$ and $L_2=\{y=0\}$, and a singular quartic chosen so that after blowing up nine suitably infinitely near points one obtains a Halphen pencil of index $1$ [2310.13319]. The resulting elliptic surface $X\to\PP^1$ has one fiber of type $I_9$, three further nodal $I_1$ fibers, and one multiple fiber $2C$ [2310.13319]. On $X$ one finds a divisor
\[
D=F+(2L_1+R_3)+(2R_1+L_2)+(2R_6+R_7)
\]
that is divisible by $3$ in $\Pic X$; taking the $3{:}1$ cover branched along $D$, then contracting and resolving, gives a smooth surface $Y$ with elliptic fibration $p_Y\colon Y\to\PP^1$ whose singular fibers are exactly four fibers of type $I_3$, one multiple fiber $2C$, and one multiple fiber $3F$ [2310.13319].

For this explicit surface one has
\[
p_g(Y)=q(Y)=0,\qquad K_Y=2F-C,\qquad K_Y^2=0,\qquad c_2(Y)=12
\]
[2310.13319]. The same paper exhibits ten divisor classes, denoted $R_2,R_5,R_8,H_1,H_2,F_1,F_2,G_1,G_2$ together with $K_Y$, computes their full intersection matrix, and obtains a unimodular lattice of signature $(1,9)$ [2310.13319]. This explicit lattice-theoretic control is what makes a twelve-term line-bundle exceptional collection possible in that example.

## 5. Exceptional bundles, exceptional collections, and quiver realizations

The smoothing picture provides a natural source of exceptional bundles. On the weighted-projective component
\[
W_2\simeq \PP(1,n-1,1)
\]
one constructs an exceptional rank-$n$ bundle $G_2$, and on
\[
W_1\simeq \PP^2
\]
one takes the rank-$2$ tangent twist
\[
G_1=\mathcal T_{\PP^2}(-1),
\]
with restrictions
\[
G_2|_{Z_2}\simeq \mathcal O_{Z_2}(1)^{\oplus n},\qquad G_1|_{Z_1}\simeq \mathcal O_{Z_1}(1)^{\oplus 2}
\]
[1506.05213]. Gluing these to copies of $\mathcal O_{\widetilde X_0}(D')$ yields exceptional bundles on the reducible central fiber, and these deform uniquely to exceptional bundles on the smooth Dolgachev surface [1506.05213].

In the $(2,3)$ case, setting $G_0^{\rm gen}=0$ and $G_{11}^{\rm gen}:=2G_{10}^{\rm gen}$, one obtains an exceptional sequence
\[
\langle \mathcal O(G_0^{\rm gen}),\mathcal O(G_1^{\rm gen}),\dots,\mathcal O(G_{11}^{\rm gen})\rangle
\]
of length $12$ in $D^b(X^{\rm gen})$ [1506.05213]. Semiorthogonality and exceptionality are proved by explicit cohomology vanishings, using upper semicontinuity, Riemann–Roch on $X^{\rm gen}$, and degeneration to $Y$ and the components $W_i$ [1506.05213]. In the special $(2,3)$ example constructed via the triple cover, the paper [2310.13319] gives an explicit line-bundle realization
\[
\langle A_1,\dots,A_{12}\rangle\subset D^b(Y),
\]
where the $A_i$ are written as explicit $\QQ$-linear combinations of the lattice generators; Theorem 2.6 there states that these twelve line bundles form an exceptional collection of maximal possible length $12$.

A different but related construction applies to all Dolgachev surfaces $D_{p,q}$. The paper [2204.13225] proves the existence of a strong exceptional collection of length $10$,
\[
E_9,E_8,\dots,E_0\in D^b(Y),
\]
where each $E_i$ is a Hacking–Pak bundle obtained by extending the Kawamata bundle on the $N$-resolution $W^-$ to the smoothing. The Chern characters are computed as
\[
\operatorname{ch}(E_i)=\bigl(n_i,\,-n_i(A+\Gamma_1+\cdots+\Gamma_i),\,c_2(E_i)\bigr),
\]
with
\[
c_1(E_i)=-n_i(A+\Gamma_1+\cdots+\Gamma_i)\in H_2(Y,\ZZ),
\]
and
\[
c_2(E_i)=\frac12\left(\frac{n_i-1}{n_i}\right)\bigl(c_1(E_i)^2+n_i+1\bigr)
\]
[2204.13225]. The same paper establishes
\[
\operatorname{Ext}^k(E_i,E_j)=0\quad\text{for all }k>0\text{ whenever }i<j,
\]
and for the unique nonzero $\delta$-block,
\[
\operatorname{Hom}(E_1,E_0)=0,\qquad \operatorname{Ext}^1(E_1,E_0)\simeq \CC^{pq-p-q},
\]
with all other $\operatorname{Ext}^k(E_1,E_0)$ vanishing [2204.13225].

The endomorphism algebra of $\bigoplus_{i=0}^9 E_i$ is the path algebra of a star-shaped quiver $Q_{p,q}$ with ten vertices and
\[
\delta_1=pq-p-q
\]
arrows in the distinguished block [2204.13225]. This yields an admissible embedding
\[
\Phi\colon D^b(\CC Q_{p,q}\text{-}\mathrm{mod})\longinto D^b(Y)
\]
sending the simple modules to the exceptional bundles $E_0,\dots,E_9$ [2204.13225]. The same work shows that mutations of exceptional collections correspond exactly to antiflips in the passage from $M$-resolutions to $N$-resolutions.

## 6. Non-fullness, phantom categories, and categorical limits

The most striking categorical feature is that maximal or numerically maximal exceptional collections on Dolgachev surfaces need not be full. For the twelve-term line-bundle collection on the $(2,3)$ surface arising from $\QQ$-Gorenstein smoothing, if
\[
\mathcal A=\langle \mathcal O(G_0),\dots,\mathcal O(G_{11})\rangle^\perp,
\]
then
\[
K_0(\mathcal A)=0,\qquad HH_*(\mathcal A)=0,\qquad \mathcal A\ne 0,
\]
and the pseudoheight of the collection is at least $2$; by Kuznetsov’s height formalism, $HH^0(\mathcal A)=HH^0(X^{\rm gen})=\CC$, so $\mathcal A$ is a nontrivial phantom category [1506.05213].

The explicit construction in [2310.13319] gives a different proof of non-fullness for a special $(2,3)$-surface. Let $S=\mathrm{Jac}\,Y$ be the Jacobian elliptic surface; the paper identifies $S$ with the classical Hesse pencil rational elliptic surface, with Mordell–Weil group $(\ZZ/3)^2$ and four $I_3$ fibers [2310.13319]. The Galois group $\mu_3$ of the $3{:}1$ cover acts on $Y$ by translation by a $6$-torsion section $R_8$, and restriction of the exceptional objects to the generic fiber $F_\zeta$ shows that they generate $D^b(F_\zeta)$ [2310.13319]. The $\mu_3$-action exchanges a putative semiorthogonal decomposition
\[
\langle \mathcal O_{F_\zeta},\mathcal O_{F_\zeta}(R_8)\rangle,
\]
contradicting the theorem of Kawatani–Okawa that no nontrivial semiorthogonal decomposition exists for the derived category of a smooth elliptic curve. The conclusion is again that the right orthogonal is a nonzero phantom with trivial $K_0$ and Hochschild homology [2310.13319].

This phenomenon extends beyond a single special surface. The same paper states that any other $(2,3)$-Dolgachev surface obtained by analytic logarithmic transforms from the same Jacobian surface inherits an exceptional collection of the same length, and the same $\mu_3$ argument produces a phantom in its derived category [2310.13319]. In parallel, [2204.13225] identifies for general $D_{p,q}$ a rank-$2$ Mukai lattice as the orthogonal complement to the length-$10$ strong exceptional collection and states that a full exceptional collection on $Y$ is impossible unless $pq-p-q=3$, equivalently $(p,q)=(3,2)$.

Within the literature summarized here, Dolgachev surfaces therefore serve as a meeting point of elliptic surface theory, singularity smoothing, and categorical birational geometry. The explicit exceptional collections on $D_{2,3}$, the strong length-$10$ collections on all $D_{p,q}$, the quiver realizations, and the appearance of phantoms together show that the derived categories of Kodaira-dimension-$1$ surfaces with multiple fibers are substantially subtler than their numerical invariants alone would suggest [1506.05213, 2204.13225, 2310.13319].

Source: https://www.emergentmind.com/topics/dolgachev-surfaces