Adams operations in smooth K-theory
We show that the Adams operations in complex K-theory lift to operations in smooth K-theory. The main result is a Riemann-Roch type theorem about the compatibility of the Adams operations and the integration in smooth K-theory.
š” Research Summary
The paper addresses a longāstanding gap in differential (smooth) Kātheory by constructing a lift of the classical Adams operations Ļā½įµā¾ from topological complex Kātheory to the smooth setting. The authors work within the BunkeāSchick model of smooth Kātheory, where a class is represented by a triple (E,ā,Ļ): a complex vector bundle E, a compatible connection ā, and a differential form Ļ of odd degree that records the ChernāSimons correction.
The central construction defines Ļā½įµā¾ on a smooth class (E,ā,Ļ) by
āĻā½įµā¾(E,ā,Ļ) = (E^{āk}, ā^{āk}, kĀ·Ļ + CS_k(ā,ā^{āk})).
Here E^{āk} is the kāfold tensor power, ā^{āk} the induced connection, and CS_k is a carefully chosen ChernāSimons form that compensates for the discrepancy between the differential form part of the original class and that of the tensorāpowered bundle. This definition guarantees that Ļā½įµā¾ respects the equivalence relation in smooth Kātheory, preserves the group structure, and commutes with the product. Moreover, after applying the curvature map (the smooth Chern character), Ļā½įµā¾ reduces to the usual Adams operation on the underlying topological Kātheory class, confirming that the lift is compatible with the classical theory.
A substantial portion of the work is devoted to proving that Ļā½įµā¾ interacts correctly with the pushāforward (integration) map ā«_f associated to a proper submersion f : M ā N. In smooth Kātheory, ā«_f is defined using the BismutāFreed superconnection and the associated Ī·āform, which captures the differential refinement of the index theorem. The authors establish a āsmooth AdamsāRiemannāRochā formula:
āā«_f Ļā½įµā¾(x) = Ļā½įµā¾(ā«_f x) + dα(x,f,k),
where α(x,f,k) is an explicit differential form built from the Ī·āform of the BismutāFreed connection and the ChernāSimons correction CS_k. The term dα is exact, reflecting that the discrepancy between the two sides is a boundary in the deāÆRham complex. When the differential form component Ļ vanishes (i.e., for pure topological classes), α disappears and the formula collapses to the familiar commutation of Ļā½įµā¾ with the topological pushāforward, reproducing the classical AdamsāRiemannāRoch theorem.
The proof proceeds by first analyzing how Ļā½įµā¾ transforms the BismutāFreed connection, then computing the change in the associated Ī·āform, and finally showing that the resulting correction is precisely dα. The authors verify that α depends polynomially on the curvature of ā and on the characteristic forms of the vertical tangent bundle of f, making the formula amenable to explicit calculations in concrete geometric situations.
Beyond the main theorem, the paper discusses several important consequences. The lifted Adams operations enable refined index formulas that incorporate differential data, allowing one to compute higherāorder characteristic numbers (such as refined Aāgenus or Todd class) with full differential information. In the context of quantum field theory, where ChernāSimons terms and anomaly cancellation conditions often involve both topological and differential contributions, the smooth AdamsāRiemannāRoch theorem provides a systematic way to track how these contributions behave under dimensional reduction or integration over fibers.
Finally, the authors outline future directions: extending the construction to other Ī»āoperations, exploring multiplicative genera in the smooth setting, and applying the framework to equivariant smooth Kātheory and to the study of differential refinements of elliptic cohomology. The paper thus not only fills a technical gap by providing a coherent lift of Adams operations but also opens a pathway for a richer interaction between algebraic topology, differential geometry, and mathematical physics within the robust language of smooth Kātheory.
Comments & Academic Discussion
Loading comments...
Leave a Comment