Grothendieck's existence theorem

Grothendieck's existence theorem
Name	Grothendieck's existence theorem
Field	Algebraic geometry
Discoverer	Alexander Grothendieck
Year	1960s
Related	Formal GAGA, Artin approximation, Serre duality, EGA, SGA

Grothendieck's existence theorem Grothendieck's existence theorem establishes a fundamental bridge between formal schemes and algebraic schemes by giving conditions under which coherent sheaves and morphisms that exist formally along a closed subscheme algebraize to genuine coherent sheaves and morphisms on a neighborhood. The theorem sits at the nexus of work by Alexander Grothendieck, Jean-Pierre Serre, Oscar Zariski, David Mumford, Michael Artin, Jean-Louis Verdier, and others who shaped modern EGA and SGA foundations.

Statement of the theorem

Grothendieck formulated a precise criterion: given a proper morphism of noetherian schemes X → S and a closed subscheme S0 ⊂ S defined by a nilpotent ideal or an ideal of definition, the formal completion X̂ along the fiber over S0 yields an equivalence between the category of coherent O_X-modules on a neighborhood of X_{S0} (with compatible morphisms) and the category of coherent modules on the formal scheme X̂ satisfying the appropriate topological completeness. The theorem asserts full faithfulness and essential surjectivity for coherent sheaves and for finite type algebras under hypotheses closely related to properness, flatness, and cohomological finiteness such as the vanishing of higher direct images or coherence of R^i f_* for i ≥ 0. Influential companions to the statement include Serre's criterion, Cartier divisors, Hilbert scheme, Picard scheme, and statements about formal functions and completion functors found in EGA III, EGA IV, and SGA 1.

Historical context and motivation

Motivation arose from classical problems like algebraization of formal power series solutions in the work of David Hilbert, the development of formal geometry by Kurt Hensel, and questions about deformation and moduli addressed by Alexander Grothendieck and Michael Artin. Grothendieck framed the problem within the program of EGA to reconcile formal neighborhoods studied by Oscar Zariski and global algebraic structures used by Jean-Pierre Serre and André Weil. The theorem was driven by needs in the construction of the Picard scheme, representability problems treated by Grothendieck, and by the desire to control cohomology in families as exemplified in works by Jean-Louis Verdier, Michel Raynaud, Pierre Deligne, Phillip Griffiths, and David Mumford.

Sketch of proof and key ideas

The proof combines cohomological control via properness with formal patching and Artin approximation techniques. One begins with formal completion and uses the theorem on formal functions (derived from results in EGA III) to relate R^i f_* of coherent sheaves on X to the inverse limit of R^i f_* on thickenings, invoking coherence results due to Jean-Pierre Serre and duality frameworks associated with Serre duality and Grothendieck duality. Using finiteness theorems from EGA IV and descent methods elaborated in SGA 1 and SGA 4, one establishes full faithfulness by comparing Hom groups via inverse limits and Mittag-Leffler arguments similar to techniques used by Alexander Grothendieck in his treatment of prorepresentable functors. Essential surjectivity is achieved by constructing algebraizations locally using flattening stratifications of Raynaud–Gruson type and then gluing via descent as developed by Michel Raynaud and Laurent Lafforgue-era techniques. Complementary methods by Michael Artin on algebraic approximation contribute alternative proofs in special contexts, and later treatments incorporate derived category perspectives by Bernhard Keller and Jacob Lurie.

Applications and consequences

The theorem underpins representability results such as the existence of the Hilbert scheme and the Picard scheme, and it is crucial in deformation theory studied by Michael Artin, Barry Mazur, and Pierre Deligne. It enables algebraization in contexts encountered in the study of moduli spaces of curves as in work by Deligne–Mumford, and it influences formal geometry methods in the proofs of theorems by Faltings and Wiles where control of formal models matters. Consequences extend to the construction of Néron models studied by André Néron and Bosch–Lütkebohmert–Raynaud, to comparison theorems in arithmetic geometry used by Jean-Marc Fontaine and Alexander Beilinson, and to applications in complex analytic algebraization echoing results by Serre and H. Cartan.

Examples and counterexamples

Standard examples include the algebraization of formal coherent sheaves on projective schemes over complete local rings such as projective space over a p-adic integer ring, illustrating classical formal GAGA phenomena and examples drawn from Fano varieties and K3 surfaces where cohomological finiteness holds. Counterexamples occur when properness or coherence hypotheses fail: non-proper morphisms, schemes lacking noetherian hypotheses, or settings with infinite cohomology produce formal objects that do not algebraize, as seen in pathological constructions inspired by Zariski and later examples by Nagata and Serre demonstrating failure of finiteness and failure of the Mittag-Leffler property.

Variants and generalizations

Variants include formal GAGA for analytic spaces developed by Serre and refined by Grothendieck; Artin's approximation and algebraization theorems by Michael Artin provide broader algebraization criteria for algebraic stacks and formal moduli problems as in work by Jacob Lurie and Dmitry Kaledin. Derived and stacky generalizations appear in sources influenced by Toën–Vezzosi and Gabriele Vezzosi on derived algebraic geometry, with enhancements by Bertrand Toën and Jacob Lurie providing derived analogues for perfect complexes and for higher obstruction theories. Extensions to equivariant settings and formal patching interact with results by Serre, Grothendieck, Michel Raynaud, and more recent work on rigid analytic geometry by Vladimir Berkovich and Roland Huber.

Category:Algebraic geometry