1
$\begingroup$

It is well-known that if one assumes algebraic closedness and characteristic 0 of the residue field then finite covers of complete DVRs are all of the form $A[x]/(x^m-a)$ for some $a \in A$ (direct sums of such, more precisely).

Is there a similar description for finite covers of complete regular local rings in higher dimensions?

$\endgroup$
2
  • $\begingroup$ In the DVR case your statement is true only if the residue field has characteristic zero. $\endgroup$ Feb 7, 2011 at 19:24
  • 1
    $\begingroup$ For covers étale out of a normal crossing divisor, the description is rather simple. See SGA1, XIII, §5. $\endgroup$
    – Qing Liu
    Feb 7, 2011 at 23:57

1 Answer 1

2
$\begingroup$

Depending on what you mean by cover your statement isn't true in the DVR case. To make it true in that case you can throw in the condition that the cover be normal and it may also be a good idea to assume the map is flat. However, under your assumptions the complete regular local ring is a power series ring over the residue field $k$ in $n$ say variables. Now, for $n$-dimensional complete local normal ring $R$ with residue field $k$ there is a finite map from $R$ to the power series ring in $n$ variables. If $R$ is also Cohen-Macaulay then the map is flat. Hence, any $n$-dimensional complete local normal Cohen-Macaulay ring with $k$ as residue field appears as covering. When $n>1$ there is an enormous number of possible such ring so those of the suggested type only cover a very small number of cases (for instance they have embedding dimension at most $n+1$ and there are normal Cohen-Macaulay local rings of fixed dimension and arbitrary high embedding dimension).

$\endgroup$
2
  • $\begingroup$ yes, I meant a normal cover. what if one assumes normality of the cover in the n-dimensional case? $\endgroup$ Feb 7, 2011 at 20:48
  • 1
    $\begingroup$ As I said, there are far too many normal Cohen-Macaulay rings. $\endgroup$ Feb 7, 2011 at 21:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.