Quotients of toric varieties by the action of a subtorus (Q1292763): Difference between revisions

If an algebraic group \(H\) acts on an algebraic variety \(X\) then a categorical quotient is a morphism \(p:X\to Y\) which is constant on \(H\)-orbits, such that all \(H\)-invariant morphisms \(f:X\to Z\) factor uniquely through~\(p\). In general such a \(p\) need not exist. In the special case where \(X\) is a toric variety and \(H\) is a subtorus of the open torus \(T\subset X\), a categorical quotient in the category of toric varieties does always exist. The authors prove this by giving an algorithm that constructs the fan of~\(Y\) starting from the fan of~\(X\). If \(X\) has a good quotient (in the sense of Seshadri), \(p:X\to Y\), under the action of~\(H\), then \(p\) coincides with the categorical quotient. However the toric categorical quotient will usually not be good. In the last part of this paper it is shown that \(X\) has a good model \(\overline X\), which is an associated toric variety with an action of~\(H\) that does have a good quotient. Thus the assignment \(X\mapsto \overline X\) is adjoint to the forgetful functor on toric varieties with \(H\)-action and a good quotient that forgets the existence of the good quotient. Again the proof is a direct construction of the appropriate fan.