3
$\begingroup$

Is there an example of an integral projective curve $X$ over an algebraically closed field $k$ with nodal singularities whose normalization is a (smooth) genus 1 curve?

More generally, given any $g,n$, can we always find an integral projective curve $X$ over $k$ with $n$ nodes whose normalization is genus $g$?

$\endgroup$
  • $\begingroup$ A quartic with 2 double points has genus $1$. Is this what you are looking for ? $\endgroup$ – Rene Schipperus May 1 '17 at 20:52
  • $\begingroup$ @ReneSchipperus How would you construct such a quartic? Is there a general procedure for constructing curves with $n$ nodes and whose normalization is genus $g$? $\endgroup$ – oxeimon May 1 '17 at 21:28
  • $\begingroup$ Doesnt such a curve automatically have genus 1 by the genus formula ? $\endgroup$ – Rene Schipperus May 1 '17 at 21:40
2
$\begingroup$

The answer to your question is yes. Let $X$ be a smooth genus $g$ curve and let $L$ be a very large degree line bundle ($\deg L>2g+2n+2$ should do). Let $P_i, Q_i\in X, 1\leq i\leq n$ be $2n$ distinct points. Then you have a surjection $H^0(L)\to H^0(L_{|{\cup P_i\cup Q_i}})=\prod k(P_i)\times\prod k(Q_i)$. Take $\prod_{i=1}^n k\to \prod k(P_i)\prod k(Q_i)$ given by each $k$ mapping to $k(P_i)\times k(Q_i)$ as the diagonal embedding. Let $V\subset H^0(L)$ the sections mapping to this $\prod k\subset \prod k(P_i)\times \prod k(Q_i)$. Then, you can consider the the map given by this linear system $V$ and check that the image is precisely a curve where the $P_i,Q_i$ for all $i$ are identified into nodal points. The checking is tiresome, but follows the same principle as you would find in Hartshorne's book.

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ Aren't you just saying that any curve is birational to a plane curve with ordinary singularities ? $\endgroup$ – Rene Schipperus May 2 '17 at 3:04
  • $\begingroup$ @ReneSchipperus Not really, since if genus is fixed, the number of double points for a plane curve birational to it is fixed too, so not all $n$ would work. $\endgroup$ – Mohan May 2 '17 at 13:39
  • $\begingroup$ @ReneSchipperus You are right. The only minor extra in the above construction is that you can choose the points you want to identify, a bit more difficult when you are dealing with generic projections to plane curves, at least in my mind. $\endgroup$ – Mohan May 2 '17 at 14:09

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.