While reading Peter Topping's “Lectures on the Ricci Flow”, I came across the term “linearization” and was unsure of its precise meaning in this context. The book considers a nonlinear differential operator acting between spaces of smooth sections of a vector bundle over a compact manifold. I understand that linearization refers to a form of linear approximation. However, the space of smooth sections is a Fréchet space, not a Banach space, while the Fréchet derivative is typically defined for operators between Banach spaces. This leads me to the following questions:
- In this context, does “linearization” refer to the Fréchet derivative of the operator once it has been extended to act between suitable Sobolev spaces? Or is it something defined more directly on the Fréchet space, perhaps related to the Gâteaux derivative? Are these approaches equivalent?
- Is it possible to define a Fréchet derivative directly on Fréchet spaces? For instance, could one simply replace the norm in the definition with a seminorm? Why does this approach seem to be avoided? I suspect that for operators with good regularity, such a definition might coincide with the practical linearization used in the field, but I'm not sure why this isn't standard.
If anyone have any ideas, or have good references please let me know.
Thank you.