3
$\begingroup$

Let $X$ be a smooth plane projective curve of degree $6$ and genus $10$ (over complex numbers). Then my question is the following :

Question : Is it possible that there exists a special divisor $D$ of degree $9$ on $X$ admitting exactly $4$ independent sections?

Observations : $(i)$ From Clifford's theorem: we have, $h^0(\mathcal O_X(D)) -1 =3\leq \frac{\text{deg}(D)}{2} =4.5$. Therefore, this theorem says that such divisor may exists on $X$.

$(ii)$ Note that, if such a divisor exists on $X$, then it belongs to $G^3_9(X)$. Since, $\rho(10,3,9)=-6<0$, we can't guarantee the non-emptiness of $G^3_9(X)$.

Any insight or remark from anyone is welcome.

$\endgroup$

1 Answer 1

3
$\begingroup$

No, this is not possible. Let $H$ be the divisor of a line. Use the base point free pencil trick to get an exact sequence $$0\rightarrow H^0(D-H)\rightarrow H^0(D)^2\rightarrow H^0(D+H)$$ Since $\deg(D+H)=15$, we have by Riemann-Roch $h^0(D+H)\leq 7$, hence $h^0(D-H)\geq 1$. Thus $D\equiv H+E$ with $E\geq 0$ of degree 3. But by Serre duality, $h^0(H+E)=h^0(2H-E)$. Now $E$ imposes independent conditions on conics, so $h^0(2H-E)=h^0(2H)-3=3$, contradicting $h^0(D)=4$.

$\endgroup$
19
  • 1
    $\begingroup$ The system $\lvert 2H\rvert$ is cut down by conics in $\mathbb{P}^2$, so $\lvert 2H-E\rvert$ is cut down by conics passing through $E$. Now any length 3 subscheme of $\mathbb{P}^2$ imposes 3 conditions on conics — see Lemma 0.1.1 in Beltrametti-Sommese Zero cycles and $k$-th order embeddings of smooth projective surfaces, INDAM Symposia Mathematica, 32,(1992), 33-48. $\endgroup$
    – abx
    Commented Mar 16, 2021 at 20:27
  • 1
    $\begingroup$ What is your question if $\deg(D)=8$? You can have $h^0=3$, and certainly not 4 — the exact sequence would give you a $g^1_2$. $\endgroup$
    – abx
    Commented Mar 17, 2021 at 16:53
  • 1
    $\begingroup$ $E$ cannot be a $g^1_2$, $X$ would be hyperelliptic. $\endgroup$
    – abx
    Commented Mar 17, 2021 at 17:37
  • 1
    $\begingroup$ If it is 4, $D+p$ for any $p\in X$ contradicts my previous answer. $\endgroup$
    – abx
    Commented Mar 17, 2021 at 18:26
  • 1
    $\begingroup$ Yes, this is what I mean. $\endgroup$
    – abx
    Commented Mar 17, 2021 at 19:27

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.