We know that if an $n$-dimensional Fano manifold admits a Kahler-Einstein metric, it satisfies the following Chern number inequality $$nc_1^n\leq 2(n+1)c_2c_1^{n-2}.$$

My question is whether there exists a Fano manifold which does not satisfy this inequality (of course in this case this Fano manifold can not have any K-E metric). I calculated some examples which have been famously known that they don't support K-E metrics and find out that all of them still satisfy this inequality.

Thanks in advance.


I think a counterexample is given by $\mathbb{P}_{P}(\mathcal{O}_{P}\oplus \mathcal{O}_{P}(n-1))$ for $n\geq 4$, with $P:=\mathbb{P}^{n-1}$. The computation is a bit long but here are the main steps (I hope I didn't make mistakes). Let $h\in H^2(X,\mathbb{Z})$ be the class of the tautological bundle, and $f$ the pull back of the class of $\mathcal{O}_{P}(1)$. Standard computations give $$c_1(X)=2h+f\qquad c_2(X)= 2n\,hf-\binom{n}{2}f^2\, \quad\mbox{ hence }$$ $$\Delta :=2(n+1)c_2-nc_1^2=4n\,hf - n^3f^2\ .$$ I want to prove $\Delta .c_1^{n-2}<0$. This is a sum of terms $\Delta .h^pf^{n-2-p}$ with positive coefficients. Note that $h^2=(n-1)hf$ (Chern class relation), so $h^pf^{n-2-p}=(n-1)^{p-1}hf^{n-3}$ for $p\geq 1$. Then $$\Delta .hf^{n-3}=4n(n-1)-n^3\qquad \Delta .f^{n-2}=4n\ .$$ For $n\geq 4$ all the terms are highly negative except $\Delta .f^{n-2}$, one sees immediately that the sum is negative.

  • $\begingroup$ to abx:Thank you very much! Your calculation is correct! I also found out this example and some calculations in page 137 of Olivier Debarre's book "Higher-dimensional algebraic geometry". $\endgroup$ – Kevin Feb 22 '14 at 4:07

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.