Let $K$ be a number field of class number $1$(in other words, it's ring of integers $O_K$ is PID.
Let $p$ be a nonzero prime ideal of $O_K$.
$L/K$ is an extension with $n=[L:K]$, we say that $L$ is totally ramified at $p$ if $p\mathcal{O}_L=q^n$ for some prime $q$ of $L$.
On the other hand, $L_p/K_p$ is tottally ramified if $L$ and $K$ have the same residue field.
Then, what is the relation between '$L/K$ is totally ramified at $p$' and '$L_p/K_p$ is totally ramified as local field' ?
I think latter does not imply former in general, but if I add some condition (like $[L:K]=[L_p:K_p]$), can't we say something like the latter implies the former ?
Thank you for your help.
Best Answer
Nothing you're describing depends on the class number. The general situation is as follows (see, for example, Milne's Algebraic Number Theory, Thm. 3.34; Neukirch's Algebraic Number Theory, Prop. I.8.2; or Serre's Local Fields, Prop. I.10).
Let $A$ be a Dedekind domain with field of fractions $K$, let $L$ be a finite separable extension of $K$ with $n = [L : K]$, and let $B$ be the integral closure of $A$ in $L$. Then $B$ is a Dedekind domain. For each prime $P$ of $A$ and each prime $Q$ of $B$ dividing $PB$, let $e_Q$ be the ramification index, i.e., the largest integer such that $Q^{e_Q}$ divides $PB$, and let $f_Q = [B/Q : A/P]$ be the residue degree (also called the inertia degree). Then $$n = \sum_{Q \mid P} e_Q f_Q.$$ In particular, the following two notions of "totally ramified at $P$" are equivalent:
In the case of an extension of local fields, there's only one prime in total, so the uniqueness part of the second condition is automatic.