Projective varieties are compactifications of affine varieties

750 Views Asked by At

This question has been asked here and here, though no answers have been given. It is exercise 3.2.1 in Smith's Invitation.

Showing $\mathbb{P}^n$ is compact boils down to showing it is the continuous image (under the projection map) of the (compact Euclidean) unit sphere in $\mathbb{C}^{n+1}\setminus \{0\}$, and then (as in a comment in the second post) projective varieties as a closed subspace of $\mathbb{P}^n$ (closed in Zariski topology, hence closed in Euclidean topology) is also compact.

I am struggling in showing that a projective variety is the compactification of an affine variety in the Euclidean (or Zariski) topology. Intuitively it makes sense - by adding on 'points at infinity' to some affine variety, it becomes compact - but I cannot seem to make it rigorous. I have the following ideas below.

If $X$ is an affine variety, the only associated projective variety $V$ I can think of is the one for which $X$ is the affine cone over $V$, where $V$ is constructed by identifying points in $X$ which lie on the same line. However, such a map $X \to V$ is clearly not injective, and so cannot be an extension (let alone compact). Should I be looking in a different direction here?