The Grothendieck group of a quantum projective space bundle (Q864995)

A basic result in the \(K\)-theory of algebraic varieties is the computation of the \(K\)-groups for projective space bundles. In this article the authors compute \(K_0\) for non-commutative analogues of projective space bundles. The motivation comes from the problem of classifying non-commutative surfaces. This is in its early stages, but the non-commutative analogues of ruled surfaces play a central role. An intersection theory is essential in this study, this can be developed by defining an intersection multiplicity as a bilinear \(\mathbb{Z}\)-valued form, the Euler form, on the Grothendieck group of the surface. The authors give a formula for the Grothendieck group for quantum ruled surfaces, and this is used to show that the associated intersection theory gives natural analogues of the commutative results: If \(X\) is a smooth commutative projective curve and \(f:\mathbb{P}(\mathcal{E})\rightarrow X\) a quantum ruled surface over \(X\), then the fibers do not meet and a section meets a fiber exactly once. The last two sentences represent the main results of the article, but on the way getting there, a lot of other interesting results and theories appears. First of all, the terminology and notation generalizes the commutative scheme theory: For abelian categories \(L\) and \(K\), the category \(\text{BIMOD}(K,L)\) of weak \(K\)-\(L\)-bimodules is the opposite of the category of left exact functors \(L\to K.\) If \(\mathcal F\) is is a weak \(K\)-\(L\)-bimodule, the notation is \(\text{Hom}_L(\mathcal F,-)\) for the corresponding left exact functor. \(\mathcal F\) is called a bimodule if \(\text{Hom}_L(\mathcal F,-)\) has a left adjoint, denoted \(-\otimes_K\mathcal F.\) This terminology leads to generalizations of algebras \(A\) as objects in \(\text{BIMOD}(K,K)\), and an \(A\)-module as an element \(M\in K\) together with a morphism \(M\otimes_K A\rightarrow M\) in \(\text{BIMOD}(\text{Ab},K)\) making the obvious diagrams commute. A quasi-scheme \(X\) is then a Grothendieck category, called \(X\) when thought of as a geometric object, \(\text{Mod}X\) when thought about as a category. The authors define graded \(X\)-modules and graded \(X\)-algebras, leading directly to the quasi-scheme \(\text{Proj}_{\text{nc}}A=\text{GrMod}A/ \text{Tors}A\). Then definitions of \(\mathcal{O}_X\)-bimodule algebras follows, and finally the definition of the analogue of a \(\mathbb{P}^n\)-bundle over a scheme \(X\), the quantum \(\mathbb{P}^n\)-bundles. This last definition makes the study of quantum ruled surfaces possible, leading up to the main results mentioned earlier. The sections of the rest of the article are ``Connected graded algebras over a quasi-scheme'' including a nice generalization of Nakayama's lemma, ``Flat connected graded \(X\)-algebras'', ``Grothendieck groups'' including the explicit expressions of the \(K_0\)-groups, and finally ``Applications to quantum projective space bundles over a commutative scheme'' defining the intersection theory and proving its main results.