Local equations for transverse fibre products #
For maps f : X → N and g : Y → N, coordinates φ on their common target give the
local equation φ(f x) - φ(g y) = 0. On the preimage of the coordinate domain this equation
cuts out exactly the fibre product. At a pair with f x = g y, its derivative is the
invertible coordinate derivative applied to df u - dg v.
Consequently the equation is regular exactly when the two derivative ranges span the target
tangent space. Its kernel is the space of pairs with equal images under the two derivatives,
with dimension dim X + dim Y - dim N in finite dimension at a transverse pair.
These are the local equations and tangent spaces used in the regular-level construction of a
transverse fibre product. The equal target tangent spaces are identified explicitly by
tangentSpaceCast; no identification of tangent spaces at distinct points is used.
No separation or countability hypotheses on the manifolds are needed.
References #
- J. M. Lee, Introduction to Smooth Manifolds, 2nd ed., Springer (2013),
Theorem 6.30 (transverse preimages), applied to the diagonal in
N × N.
The difference of target coordinates, whose zero set locally defines a fibre product.
Only on the preimage of φ.source in both factors does its vanishing characterize equality.
Instances For
The equation cuts out exactly the fibre product on the coordinate domain.
Smooth maps give a smooth defining equation throughout the coordinate domain.
At a pair in the fibre product, the vector-valued derivative of its local equation is the coordinate derivative applied to the difference of the map derivatives.
The tangent kernel of the defining equation consists of pairs of vectors with equal images under the two map derivatives. This description is independent of target coordinates.
Transversality is exactly regularity of the local fibre-product equation. The range of
dg is transported to the equal tangent space at f p.1 before taking the supremum.
At a transverse pair, the kernel of the defining equation has the expected dimension. Surjectivity forces the target dimension not to exceed the sum of the source dimensions.