Arcara-Bertram Stability Conditions on Surfaces

• Arcara–Bertram stability conditions are geometric Bridgeland conditions that tilt Coh(X) via slope torsion pairs and complexified ample classes to bridge classical μ-stability with derived categories.
• They enable precise wall-crossing analysis and moduli space construction by using mini-walls, quadratic support properties, and clear phase equations.
• The framework extends to singular surfaces, deformed Hermitian–Yang–Mills settings, and higher dimensions, establishing a versatile method in modern stability theory.

Arcara–Bertram stability conditions are geometric Bridgeland stability conditions on derived categories of surfaces obtained by tilting $\mathrm{Coh}(X)$ at a slope-theoretic torsion pair and equipping the tilted heart with a central charge built from a complexified ample class. In the surface setting, later papers describe them as a standard “large volume” family, as the natural Bridgeland framework for wall-crossing on $D^b(X)$, and as a reference point for comparisons with polynomial stability, deformed Hermitian–Yang–Mills stability, birational contractions, and moduli-theoretic constructions on surfaces and related geometries (Lo et al., 2011).

1. Surface construction and basic formalism

For a smooth complex projective surface $X$, the construction uses numerical data $B,\omega\in \mathrm{NS}(X)_{\mathbb R}$ with $\omega$ ample. The central charge is written

$$Z_{B,\omega}(E)=-\int_X e^{-(B+i\omega)}\operatorname{ch}(E),$$

and equivalently, in the notation of another surface paper,

$$Z_{\beta,\omega}(E)=-\ch_2^\beta(E)+\frac{\omega^2}{2}\ch_0(E)+i\,\omega\cdot \ch_1^\beta(E),$$

where

$$\ch^\beta(E)=\ch(E)\exp(-\beta), \qquad \ch_1^\beta(E)=\ch_1(E)-\beta\,\ch_0(E), \qquad \ch_2^\beta(E)=\ch_2(E)-\beta\cdot\ch_1(E)+\frac{\beta^2}{2}\ch_0(E).$$

The heart $A_{B,\omega}$ is the tilt of $\mathrm{Coh}(X)$ with respect to slope stability for

$D^b(X)$0

It consists of complexes with cohomology only in degrees $D^b(X)$1 and $D^b(X)$2, with $D^b(X)$3 in the torsion part and $D^b(X)$4 in the free part of the associated torsion pair. In standard notation one writes

$D^b(X)$5

or, in the one-parameter large-volume slice,

$D^b(X)$6

The Arcara–Bertram theorem is recalled in later work in the form that if $D^b(X)$7, then $D^b(X)$8 defines a Bridgeland stability condition on $D^b(X)$9 (Vilches, 9 Aug 2025).

This construction is “geometric” in the sense emphasized in later surface work: skyscraper sheaves are stable of the same phase. It is also the standard surface prototype for the later tilt-based Bridgeland constructions in higher dimension. A plausible implication is that the Arcara–Bertram framework is best viewed not merely as a specific family of central charges, but as a canonical interface between classical $X$0-stability of sheaves and Bridgeland stability on $X$1.

2. Hearts, slopes, and wall structures

The tilted heart is defined from the torsion pair cut out by the slope threshold $X$2 or $X$3. In one standard form,

$X$4

$X$5

For line bundles $X$6, the heart membership test is especially explicit:

• $X$7 iff $X$8,
• $X$9 iff $B,\omega\in \mathrm{NS}(X)_{\mathbb R}$0,
• if $B,\omega\in \mathrm{NS}(X)_{\mathbb R}$1, then $B,\omega\in \mathrm{NS}(X)_{\mathbb R}$2 is automatically $B,\omega\in \mathrm{NS}(X)_{\mathbb R}$3-stable.

Variation of stability in the Arcara–Bertram family is organized by walls and chambers. For the one-parameter family $B,\omega\in \mathrm{NS}(X)_{\mathbb R}$4, later work defines mini-walls for a fixed numerical type $B,\omega\in \mathrm{NS}(X)_{\mathbb R}$5 inside $B,\omega\in \mathrm{NS}(X)_{\mathbb R}$6, proves that mini-walls are locally finite, and proves that there exists $B,\omega\in \mathrm{NS}(X)_{\mathbb R}$7, depending only on $B,\omega\in \mathrm{NS}(X)_{\mathbb R}$8, such that there are no mini-walls in $B,\omega\in \mathrm{NS}(X)_{\mathbb R}$9. The same paper identifies Bayer’s polynomial Bridgeland semistability with $\omega$0-semistability for all $\omega$1, uniformly for a fixed numerical type (Lo et al., 2011).

This wall structure is central to the utility of the theory. It permits one-dimensional “slices” of the stability manifold to be studied by explicit phase equations, while retaining compatibility with classical moduli problems in the large-$\omega$2 regime. This suggests that Arcara–Bertram stability conditions serve simultaneously as an analytic continuation of sheaf stability and as a wall-crossing formalism with effective boundedness properties.

3. Large-scaling limits and the deformed Hermitian–Yang–Mills correspondence

A recent development studies Arcara–Bertram-type stability in the “large scaling limit” and compares it to deformed Hermitian–Yang–Mills stability for line bundles on smooth complex projective surfaces. For a line bundle $\omega$3, the large scaling limit with respect to $\omega$4 is defined by the existence of $\omega$5 such that

$\omega$6

and in the twisted setting this becomes stability with respect to $\omega$7. The motivation given is that dHYM is invariant under scaling $\omega$8, $\omega$9, whereas ordinary Bridgeland stability is not scaling invariant in general.

On a smooth complex projective surface, the same work proves the comparison:

1. if $$Z_{B,\omega}(E)=-\int_X e^{-(B+i\omega)}\operatorname{ch}(E),$$0 is $$Z_{B,\omega}(E)=-\int_X e^{-(B+i\omega)}\operatorname{ch}(E),$$1-stable for all $$Z_{B,\omega}(E)=-\int_X e^{-(B+i\omega)}\operatorname{ch}(E),$$2, then $$Z_{B,\omega}(E)=-\int_X e^{-(B+i\omega)}\operatorname{ch}(E),$$3 is $$Z_{B,\omega}(E)=-\int_X e^{-(B+i\omega)}\operatorname{ch}(E),$$4-twisted dHYM-semistable;
2. assuming the Arcara–Miles conjecture for $$Z_{B,\omega}(E)=-\int_X e^{-(B+i\omega)}\operatorname{ch}(E),$$5, if $$Z_{B,\omega}(E)=-\int_X e^{-(B+i\omega)}\operatorname{ch}(E),$$6 is $$Z_{B,\omega}(E)=-\int_X e^{-(B+i\omega)}\operatorname{ch}(E),$$7-twisted dHYM-semistable, then $$Z_{B,\omega}(E)=-\int_X e^{-(B+i\omega)}\operatorname{ch}(E),$$8 is $$Z_{B,\omega}(E)=-\int_X e^{-(B+i\omega)}\operatorname{ch}(E),$$9-stable for all $$Z_{\beta,\omega}(E)=-\ch_2^\beta(E)+\frac{\omega^2}{2}\ch_0(E)+i\,\omega\cdot \ch_1^\beta(E),$$0.

In dimension $$Z_{\beta,\omega}(E)=-\ch_2^\beta(E)+\frac{\omega^2}{2}\ch_0(E)+i\,\omega\cdot \ch_1^\beta(E),$$1, the dHYM condition is expressed numerically by the Collins–Jacob–Yau criterion: $$Z_{\beta,\omega}(E)=-\ch_2^\beta(E)+\frac{\omega^2}{2}\ch_0(E)+i\,\omega\cdot \ch_1^\beta(E),$$2 for every curve $$Z_{\beta,\omega}(E)=-\ch_2^\beta(E)+\frac{\omega^2}{2}\ch_0(E)+i\,\omega\cdot \ch_1^\beta(E),$$3. The reverse implication depends on the Arcara–Miles conjecture, which states that failure of stability for a line bundle is controlled by negative self-intersection curves and phase inequalities for subobjects such as $$Z_{\beta,\omega}(E)=-\ch_2^\beta(E)+\frac{\omega^2}{2}\ch_0(E)+i\,\omega\cdot \ch_1^\beta(E),$$4 or analogous subobjects of $$Z_{\beta,\omega}(E)=-\ch_2^\beta(E)+\frac{\omega^2}{2}\ch_0(E)+i\,\omega\cdot \ch_1^\beta(E),$$5. The same note remarks that this conjecture has been proved in several special cases, including surfaces with no negative curves, rank $$Z_{\beta,\omega}(E)=-\ch_2^\beta(E)+\frac{\omega^2}{2}\ch_0(E)+i\,\omega\cdot \ch_1^\beta(E),$$6 Néron–Severi surfaces with a unique negative curve, and certain del Pezzo surfaces of Picard rank $$Z_{\beta,\omega}(E)=-\ch_2^\beta(E)+\frac{\omega^2}{2}\ch_0(E)+i\,\omega\cdot \ch_1^\beta(E),$$7 (Fan, 24 Apr 2026).

A common misconception is that fixed-scale Bridgeland stability for a line bundle should coincide with dHYM stability. The large-scaling analysis explicitly warns that this is false: a line bundle can be stable for one scale but fail at others. The genericity hypothesis on $$Z_{\beta,\omega}(E)=-\ch_2^\beta(E)+\frac{\omega^2}{2}\ch_0(E)+i\,\omega\cdot \ch_1^\beta(E),$$8 is therefore substantive rather than cosmetic, because semistability and stability can differ on walls.

4. Reider-type arguments, normal surfaces, and degenerations to nef limits

The Arcara–Bertram method was used in the smooth case to reinterpret Reider’s theorem via Bridgeland stability, and later work extends this strategy to normal surfaces using Langer’s construction. In the normal-surface setting, one works with the tilt heart

$$Z_{\beta,\omega}(E)=-\ch_2^\beta(E)+\frac{\omega^2}{2}\ch_0(E)+i\,\omega\cdot \ch_1^\beta(E),$$9

where

$$\ch^\beta(E)=\ch(E)\exp(-\beta), \qquad \ch_1^\beta(E)=\ch_1(E)-\beta\,\ch_0(E), \qquad \ch_2^\beta(E)=\ch_2(E)-\beta\cdot\ch_1(E)+\frac{\beta^2}{2}\ch_0(E).$$0

and with the Langer-modified central charge

$$\ch^\beta(E)=\ch(E)\exp(-\beta), \qquad \ch_1^\beta(E)=\ch_1(E)-\beta\,\ch_0(E), \qquad \ch_2^\beta(E)=\ch_2(E)-\beta\cdot\ch_1(E)+\frac{\beta^2}{2}\ch_0(E).$$1

The only formal difference from the smooth Arcara–Bertram setup is the extra $$\ch^\beta(E)=\ch(E)\exp(-\beta), \qquad \ch_1^\beta(E)=\ch_1(E)-\beta\,\ch_0(E), \qquad \ch_2^\beta(E)=\ch_2(E)-\beta\cdot\ch_1(E)+\frac{\beta^2}{2}\ch_0(E).$$2-term, inserted because Langer’s Bogomolov inequality on a normal surface is weakened. When $$\ch^\beta(E)=\ch(E)\exp(-\beta), \qquad \ch_1^\beta(E)=\ch_1(E)-\beta\,\ch_0(E), \qquad \ch_2^\beta(E)=\ch_2(E)-\beta\cdot\ch_1(E)+\frac{\beta^2}{2}\ch_0(E).$$3 is smooth in characteristic $$\ch^\beta(E)=\ch(E)\exp(-\beta), \qquad \ch_1^\beta(E)=\ch_1(E)-\beta\,\ch_0(E), \qquad \ch_2^\beta(E)=\ch_2(E)-\beta\cdot\ch_1(E)+\frac{\beta^2}{2}\ch_0(E).$$4, one may take $$\ch^\beta(E)=\ch(E)\exp(-\beta), \qquad \ch_1^\beta(E)=\ch_1(E)-\beta\,\ch_0(E), \qquad \ch_2^\beta(E)=\ch_2(E)-\beta\cdot\ch_1(E)+\frac{\beta^2}{2}\ch_0(E).$$5, recovering the original Arcara–Bertram setup exactly. This permits Reider-type vanishing and divisor-existence theorems on normal surfaces, in positive characteristic, and with the Du Bois variant $$\ch^\beta(E)=\ch(E)\exp(-\beta), \qquad \ch_1^\beta(E)=\ch_1(E)-\beta\,\ch_0(E), \qquad \ch_2^\beta(E)=\ch_2(E)-\beta\cdot\ch_1(E)+\frac{\beta^2}{2}\ch_0(E).$$6 (Larsen et al., 2024).

A complementary recent direction studies birational morphisms $$\ch^\beta(E)=\ch(E)\exp(-\beta), \qquad \ch_1^\beta(E)=\ch_1(E)-\beta\,\ch_0(E), \qquad \ch_2^\beta(E)=\ch_2(E)-\beta\cdot\ch_1(E)+\frac{\beta^2}{2}\ch_0(E).$$7 from a smooth surface to a singular surface and constructs pre-stability conditions whose central charges depend on a nef class $$\ch^\beta(E)=\ch(E)\exp(-\beta), \qquad \ch_1^\beta(E)=\ch_1(E)-\beta\,\ch_0(E), \qquad \ch_2^\beta(E)=\ch_2(E)-\beta\cdot\ch_1(E)+\frac{\beta^2}{2}\ch_0(E).$$8. The central charge is the Arcara–Bertram formula with $$\ch^\beta(E)=\ch(E)\exp(-\beta), \qquad \ch_1^\beta(E)=\ch_1(E)-\beta\,\ch_0(E), \qquad \ch_2^\beta(E)=\ch_2(E)-\beta\cdot\ch_1(E)+\frac{\beta^2}{2}\ch_0(E).$$9 replaced by $A_{B,\omega}$0: $A_{B,\omega}$1 The resulting pre-stability conditions are described as limits, in the topology of $A_{B,\omega}$2, of the Arcara–Bertram stability conditions $A_{B,\omega}$3 as $A_{B,\omega}$4. In the cyclic quotient case, including $A_{B,\omega}$5 singularities, the support property is proved using a quadratic form

$A_{B,\omega}$6

and the paper also proves a converse obstruction: if such a limit exists, then $A_{B,\omega}$7 cannot contract any smooth curve of genus $A_{B,\omega}$8 (Vilches, 9 Aug 2025).

These developments show that Arcara–Bertram stability conditions are robust under two distinct kinds of generalization: singular-surface corrections within the same tilt geometry, and degeneration from ample to nef classes along birational contractions. This suggests that the surface theory is flexible enough to encode both cohomological vanishing arguments and boundary phenomena in the stability manifold.

5. Moduli spaces, wall-crossing, and explicit geometries

Arcara–Bertram stability conditions play a central role in explicit wall-crossing descriptions of moduli spaces. On a principally polarized abelian surface $A_{B,\omega}$9 with $\mathrm{Coh}(X)$0, one studies the tilted hearts $\mathrm{Coh}(X)$1 and central charges

$\mathrm{Coh}(X)$2

for $\mathrm{Coh}(X)$3. For Mukai vector $\mathrm{Coh}(X)$4, the large-$\mathrm{Coh}(X)$5 chamber gives the fine moduli space

$\mathrm{Coh}(X)$6

parametrizing twisted ideal sheaves $\mathrm{Coh}(X)$7. Wall-crossing is analyzed by exact sequences in $\mathrm{Coh}(X)$8, and the paper expresses the corresponding birational transformations as Mukai flops. Fourier–Mukai transforms are used to relate small-$\mathrm{Coh}(X)$9 and large-$D^b(X)$00 chambers and to prove projectivity of the resulting moduli spaces (Maciocia et al., 2011).

On $D^b(X)$01, Arcara–Bertram/Bayer–Macrì stability conditions organize the interpolation problem for monomial zero-dimensional schemes. For a monomial scheme $D^b(X)$02, the destabilizing wall of $D^b(X)$03 is explicitly the semicircle centered at

$D^b(X)$04

where $D^b(X)$05 is extracted combinatorially from the block diagram of $D^b(X)$06. The paper proves that there exists a vector bundle $D^b(X)$07 with interpolation for $D^b(X)$08 if and only if $D^b(X)$09, and confirms for monomial schemes the Arcara–Bertram–Coskun–Huizenga correspondence between Bridgeland walls and Mori chamber walls on $D^b(X)$10 (Coskun et al., 2013).

A related wall-to-divisor correspondence is developed for $D^b(X)$11 and its Sklyanin deformations $D^b(X)$12. In that setting, the destabilizing semicircular wall with center $D^b(X)$13 corresponds to the stable base locus wall spanned by $D^b(X)$14, and the paper proves a one-to-one correspondence between semicircular actual walls and stable base locus walls in the relevant quadrant (Li et al., 2013).

These examples display the concrete strength of the theory: the abstract wall-and-chamber structure on $D^b(X)$15 becomes directly calculable in terms of subschemes, torsion sheaves, divisors, and birational transformations of moduli spaces.

6. Broader generalizations and higher-dimensional analogues

Arcara–Bertram-type ideas also appear in product constructions and in threefold double-tilt theories. For products $D^b(X)$16, one paper starts from a stability condition $D^b(X)$17 on a smooth projective variety $D^b(X)$18, constructs the Abramovich–Polishchuk global heart

$D^b(X)$19

and for a curve $D^b(X)$20 defines central charges

$D^b(X)$21

on a tilted heart $D^b(X)$22. The paper states explicitly that it is inspired by the “Arcara–Bertram-type” philosophy of constructing stability conditions on products by modifying heart and central charge to reflect relative geometry, though it generalizes this to arbitrary stability conditions on $D^b(X)$23 and to arbitrary smooth projective curves $D^b(X)$24 (Liu, 2019).

In dimension three, the direct surface construction is replaced by the Bayer–Bertram–Macrì–Toda double-tilt program, but the lineage is explicit in later papers. For smooth quintic threefolds, Calabi–Yau double/triple solids, and the Calabi–Yau threefold $D^b(X)$25, the common structure is:

1. start from slope stability;
2. tilt once to form $D^b(X)$26;
3. define a tilt slope $D^b(X)$27;
4. prove a strengthened Bogomolov–Gieseker-type inequality, often via a Clifford-type inequality on a curve;
5. use the resulting quadratic inequality to construct an open family of Bridgeland stability conditions on the double-tilt heart.

For the quintic, the resulting stability conditions are parameterized by

$D^b(X)$28

For $D^b(X)$29, the same parameter inequalities appear in the final family

$D^b(X)$30

These papers explicitly place themselves in the Arcara–Bertram / Bayer–Bertram–Macrì–Stellari–Toda lineage, even though the original surface heart is replaced by a second tilt and a $D^b(X)$31-inequality (Li, 2018).

In this higher-dimensional context, “Arcara–Bertram stability conditions” no longer refers to the original surface theorem literally. Instead, it names a methodological template: tilt-based construction of hearts, explicit central charges from twisted Chern characters, wall analysis, quadratic support inequalities, and moduli-theoretic wall-crossing. A plausible implication is that the enduring significance of the Arcara–Bertram framework lies less in a single formula than in having established the surface model from which a wide range of later Bridgeland constructions took their geometric form.