Source author record

Kevin Tucker

Explore this author’s publications, collaborators and research topics. This catalog record has not been claimed by its author.

ResearcherUnclaimed source record

Catalog footprint

Research overview

14works
2topics
4close collaborators

Published work

14 publications

preprint2022arXiv

An analog of adjoint ideals and PLT singularities in mixed characteristic

We use the framework of perfectoid big Cohen-Macaulay algebras to define a class of singularities for pairs in mixed characteristic, which we call purely BCM-regular singularities, and a corresponding adjoint ideal. We prove that these satisfy adjunction and inversion of adjunction with respect to the notion of BCM-regularity and the BCM test ideal defined by the first two authors. We compare them with the existing equal characteristic PLT and purely FF-regular singularities and adjoint ideals. As an application, we obtain a uniform version of the Briançon-Skoda theorem in mixed characteristic. We also use our theory to prove that two-dimensional KLT singularities are BCM-regular if the residue characteristic p>5p>5, which implies an inversion of adjunction for three-dimensional PLT pairs of residue characteristic p>5p>5. In particular, divisorial centers of PLT pairs in dimension three are normal when p>5p > 5. Furthermore, in the appendix we provide a streamlined construction of perfectoid big Cohen-Macaulay algebras and show new functoriality properties for them using the perfectoidization functor of Bhatt and Scholze.

preprint2022arXiv

Covers of rational double points in mixed characteristic

We further the classification of rational surface singularities. Suppose (S,n,k)(S, \mathfrak{n}, \mathcal{k}) is a strictly Henselian regular local ring of mixed characteristic (0,p>5)(0, p > 5). We classify functions ff for which S/(f)S/(f) has an isolated rational singularity at the maximal ideal n\mathfrak{n}. The classification of such functions are used to show that if (R,m,k)(R, \mathfrak{m}, \mathcal{k}) is an excellent, strictly Henselian, Gorenstein rational singularity of dimension 22 and mixed characteristic (0,p>5)(0, p > 5), then there exists a split finite cover of $\mbox{Spec}(R)$ by a regular scheme. We give an application of our result to the study of 22-dimensional BCM-regular singularities in mixed characteristic.

preprint2021arXiv

Bertini Theorems for FF-signature and Hilbert-Kunz multiplicity

We show that Bertini theorems hold for FF-signature and Hilbert--Kunz multiplicity. In particular, if X⊆PnX \subseteq \mathbb{P}^n is normal and quasi-projective with FF-signature greater than λλ (respectively the Hilbert--Kunz multiplicity is less than λλ) at all points x∈Xx \in X, then for a general hyperplane H⊆PnH \subseteq \mathbb{P}^n the FF-signature (respectively Hilbert--Kunz multiplicity) of X∩HX \cap H is greater than λλ (respectively less than λλ) at all points x∈X∩Hx \in X \cap H.

preprint2014arXiv

F-singularities via alterations

For a normal F-finite variety XX and a boundary divisor ΔΔ we give a uniform description of an ideal which in characteristic zero yields the multiplier ideal, and in positive characteristic the test ideal of the pair (X,Δ)(X,Δ). Our description is in terms of regular alterations over XX, and one consequence of it is a common characterization of rational singularities (in characteristic zero) and F-rational singularities (in characteristic pp) by the surjectivity of the trace map π∗ωY→ωXπ_* ω_Y \to ω_X for every such alteration π Y→Xπ\: Y \to X. Furthermore, building on work of B. Bhatt, we establish up-to-finite-map versions of Grauert-Riemenscheneider and Nadel/Kawamata-Viehweg vanishing theorems in the characteristic pp setting without assuming W2W2 lifting, and show that these are strong enough in some applications to extend sections.

preprint2014arXiv

Test ideals of non-principal ideals: Computations, Jumping Numbers, Alterations and Division Theorems

Given an ideal a⊆Ra \subseteq R in a (log) QQ-Gorenstein FF-finite ring of characteristic p>0p > 0, we study and provide a new perspective on the test ideal τ(R,at)τ(R, a^t) for a real number t>0t > 0. Generalizing a number of known results from the principal case, we show how to effectively compute the test ideal and also describe τ(R,at)τ(R, a^t) using (regular) alterations with a formula analogous to that of multiplier ideals in characteristic zero. We further prove that the FF-jumping numbers of τ(R,at)τ(R, a^t) as tt varies are rational and have no limit points, including the important case where RR is a formal power series ring. Additionally, we obtain a global division theorem for test ideals related to results of Ein and Lazarsfeld from characteristic zero, and also recover a new proof of Skoda's theorem for test ideals which directly mimics the proof for multiplier ideals.

preprint2012arXiv

Explicitly Extending Frobenius Splittings over Finite Maps

Suppose that π Y→Xπ\: Y \to X is a finite map of normal varieties over a perfect field of characteristic p>0p > 0. Previous work of the authors gave a criterion for when Frobenius splittings on XX (or more generally any p−ep^{-e}-linear map) extend to YY. In this paper we give an alternate and highly explicit proof of this criterion (checking term by term) when ππ is tamely ramified in codimension 1. Some additional examples are also explored.

preprint2012arXiv

F-signature of pairs and the asymptotic behavior of Frobenius splittings

We generalize FF-signature to pairs (R,D)(R,D) where DD is a Cartier subalgebra on RR as defined by the first two authors. In particular, we show the existence and positivity of the FF-signature for any strongly FF-regular pair. In one application, we answer an open question of I. Aberbach and F. Enescu by showing that the FF-splitting ratio of an arbitrary FF-pure local ring is strictly positive. Furthermore, we derive effective methods for computing the FF-signature and the FF-splitting ratio in the spirit of the work of R. Fedder.

preprint2012arXiv

F-signature of pairs: Continuity, p-fractals and minimal log discrepancies

This paper contains a number of observations on the {FF-signature} of triples $(R,Δ,\ba^t)$ introduced in our previous joint work. We first show that the FF-signature $s(R,Δ,\ba^t)$ is continuous as a function of tt, and for principal ideals $\ba$ even convex. We then further deduce, for fixed tt, that the FF-signature is lower semi-continuous as a function on $\Spec R$ when RR is regular and $\ba$ is principal. We also point out the close relationship of the signature function in this setting to the works of Monsky and Teixeira on Hilbert-Kunz multiplicity and pp-fractals. Finally, we conclude by showing that the minimal log discrepancy of an arbitrary triple $(R,Δ,\ba^t)$ is an upper bound for the FF-signature.

preprint2012arXiv

On the behavior of test ideals under finite morphisms

We derive transformation rules for test ideals and FF-singularities under an arbitrary finite surjective morphism π:Y→Xπ: Y \to X of normal varieties in prime characteristic p>0p > 0. The main technique is to relate homomorphisms F∗OX→OXF_{*} O_{X} \to O_{X}, such as Frobenius splittings, to homomorphisms F∗OY→OYF_{*} O_{Y} \to O_{Y}. In the simplest cases, these rules mirror transformation rules for multiplier ideals in characteristic zero. As a corollary, we deduce sufficient conditions which imply that trace is surjective, i.e. TrY/X(π∗OY)=OXTr_{Y/X}(π_{*}O_{Y}) = O_{X}.

preprint2011arXiv

F-signature exists

Suppose R is a Noetherian local ring with prime characteristic p>0. In this article, we show the existence of a local numerical invariant, called the F-signature, which roughly characterizes the asymptotic growth of the number of splittings of the iterates of the Frobenius endomorphism of R. This invariant was first formally defined by C. Huneke and G. Leuschke and has previously been shown to exist only in special cases. The proof of our main result is based on the development of certain uniform Hilbert-Kunz estimates of independent interest. Additionally, we analyze the behavior of the F-signature under finite ring extensions and recover explicit formulae for the F-signatures of finite quotient singularities.

preprint2011arXiv

Test ideals via a single alteration and discreteness and rationality of FF-jumping numbers

Suppose (X,Δ)(X, Δ) is a log-$\bQ$-Gorenstein pair. Recent work of M. Blickle and the first two authors gives a uniform description of the multiplier ideal $\mJ(X;Δ)$ (in characteristic zero) and the test ideal τ(X;Δ)τ(X;Δ) (in characteristic p>0p > 0) via regular alterations. While in general the alteration required depends heavily on ΔΔ, for a fixed Cartier divisor DD on XX it is straightforward to find a single alteration (e.g. a log resolution) computing $\mJ(X; Δ+ λD)$ for all λ≥0λ\geq 0. In this paper, we show the analogous statement in positive characteristic: there exists a single regular alteration computing τ(X;Δ+λD)τ(X; Δ+ λD) for all λ≥0λ\geq 0. Along the way, we also prove the discreteness and rationality for the FF-jumping numbers of τ(X;Δ+λD)τ(X; Δ+ λD) for λ≥0λ\geq 0 where the index of KX+ΔK_X + Δ is arbitrary (and may be divisible by the characteristic).

preprint2009arXiv

On the number of compatibly Frobenius split subvarieties, prime FF-ideals, and log canonical centers

Let XX be a projective Frobenius split variety over an algebraically closed field with splitting θ:F∗ØX→ØXθ: F_* Ø_X \to Ø_X. In this paper we give a sharp bound on the number of subvarieties of XX compatibly split by θθ. In particular, suppose $\sL$ is a sufficiently ample line bundle on XX (for example, if $\sL$ induces a projectively normal embedding) with $n = \dim H^0(X, \sL)$. We show that the number of dd-dimensional irreducible subvarieties of XX that are compatibly split by θθ is less than or equal to (nd+1){n \choose d+1}. This generalizes a well known result on the number of closed points compatibly split by a fixed splitting θθ. Similarly, we give a bound on the number of prime FF-ideals of an FF-finite FF-pure local ring. Compatibly split subvarieties are closely related to log canonical centers. Our methods apply in any characteristic, and so we are also able to bound the possible number of log canonical centers of a log canonical pair (X,Δ)(X, Δ) passing through a closed point x∈Xx \in X. Specifically, if nn is the embedding dimension of XX at xx, then the number of dd-dimensional log canonical centers of (X,Δ)(X, Δ) through xx is less than or equal to (nd){n \choose d}.

Research graph

See the researcher in context

Open full explorer

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.