5
$\begingroup$

For a field $K\subset \mathbb{Q}(\zeta_p)$ $~$($\zeta_p$ a primitive pth root of unity, p a prime), it seems to be the case that the discriminant of $K$ is $p^{[K:\mathbb{Q}]-1}$ (according to Sage). How can I prove that/ is the proof implied by something written down somewhere?

I have a feeling this is somewhat deep so I ask it here; if it's insufficiently deep I'm happy to close it and ask elsewhere.

$\endgroup$
  • 5
    $\begingroup$ The Führerdiskriminantenproduktformel tells you that is it the product of conductors of characters, but all but the trivial character must have conductor p. $\endgroup$ – Chris Wuthrich Oct 18 at 12:31
9
$\begingroup$

The Führerdiskriminantenproduktformel tells you that is it the product of conductors of characters, but all but the trivial character must have conductor $p$.

$\endgroup$
10
$\begingroup$

This would also work for non-abelian, and even non-Galois extensions: if $L/\mathbb{Q}$ is totally ramified and tamely ramified at $p$ and $K$ is an intermediate field, then $K$ is also totally tamely ramified at $p$, so the exponent of $p$ in the discriminant of $K$ is $[K:\mathbb{Q}]-1$ (Serre, Corps Locaux, Proposition 13).

$\endgroup$
  • 3
    $\begingroup$ This is (IMO) a better answer. First, the different (and discriminant) of a totally tamely ramified extension is certainly a simpler notion than understanding the conductor of an abelian extension. Second, it is more general. Of course, both answers are correct. $\endgroup$ – Electric Penguin Oct 18 at 21:29
  • 1
    $\begingroup$ @ElectricPenguin theres a very elementary definition of conductor in this special case: the smallest n s.t. $\chi$ factors through (Z/pZ)*-->(Z/nZ)* $\endgroup$ – dessin d'enfant terrible Oct 19 at 22:52
  • 1
    $\begingroup$ @dessind'efantterrible To split hairs, I would say that this is a computation of the conductor in this case rather than a definition. The distinction is that this if you actually want to prove the Führerdiskriminantenproduktformel (in the abelian case) for a local extension $L/K$, you really need to grapple with $N_{L/K}(L^*)$ in some way (say via basic local class field theory), whereas one could compute (with a full proof) the different of a totally tamely ramified extension $L/K$ in a single MO comment. $\endgroup$ – Electric Penguin Oct 20 at 0:33
  • $\begingroup$ @GHfromMO The extension $\mathbf{Q}_p(p^{1/e})$ with $e > 1$ prime to $p$ (for example) is certainly totally tamely ramified and is never generated by a root of unity. Perhaps you are thinking about the fact that an abelian extension of $\mathbf{Q}_p$ is cyclotomic (local Kronecker-Weber)? $\endgroup$ – Electric Penguin Oct 20 at 0:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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