Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Consider map from $S^{n}$ to $\mathbb{RP}^{n}$ $$\varphi:S^{n}\to\mathbb{RP}^{n}$$ which maps point $x\in S^{n}$ to corresponding direction in $\mathbb{R}^{n+1}$. This map induces map $${\varphi}^{*}:\Omega^{\bullet}(\mathbb{RP}^{n})\to\Omega^{\bullet}(S^{n})$$ and map of de Rham cogomology groups $${\psi}^{*}:H^{\bullet}(\mathbb{RP}^{n})\to H^{\bullet}(S^{n}).$$ as I see for n-th cohomology $$\dim H^{n}(\mathbb{RP}^n) = \dim\operatorname{Im} {\psi}^{*} + \dim\operatorname{Ker} {\psi}^{*}$$ Am I right that $$\dim\operatorname{Ker} {\psi}^{*} = 0$$ (why?) and $$\dim\operatorname{Ker} {\psi}^{*} = 0,1~~~\text{depending on the parity of}~n?$$

Thanks a lot!

share|improve this question
1  
Дорогой Aspirin, это интересный вопрос: +1 –  Georges Elencwajg Mar 10 '12 at 23:40
    
Спасибо! (Thanks a lot!) =) –  Aspirin Mar 11 '12 at 1:50

1 Answer 1

up vote 3 down vote accepted

The involution $i:S^n\to S^n:x\mapsto -x$ induces a decomposition $\Omega^n(S^{n})=\Omega^n_+(S^{n}) \oplus \Omega^n_-(S^{n})$ where $\Omega^n_\pm(S^{n}) $ consists of the differential forms satisfying $i^*\omega=\pm \omega$.
This yields a decomposition $H^n(S^{n})=H^n_+(S^{n}) \oplus H^n_-(S^{n})\cong \mathbb R$

On the other hand the quotient map $\phi:S^{n}\to\mathbb{RP}^{n}$ induces an isomorphism $\phi^*:\Omega^n(\mathbb P^{n})\stackrel {\cong}{\to }\Omega^n_+(S^{n})$ and then an isomorphism $\phi^*:H^n(\mathbb P^{n})\stackrel {\cong}{\to }H^n_+(S^{n})$.

Finally there remains to notice that the the canonical generator $[\omega_0]\in H^n(S^{n})=\mathbb R [\omega_0]$ is in $H^n_+(S^{n})$ or $H^n_-(S^{n})$ according as $n$ is odd or even, since $i^*(\omega_0)=(-1)^{n+1}\omega_0$.
We can thus conclude that $H^n(\mathbb P^{n})=\mathbb R$ for $n$ odd and $H^n(\mathbb P^{n})=0$ for $n$ even.

Reminder
The form $\omega_0\in \Omega^n(S^n)$ is defined by $$\omega_0(s)(v_1,\cdots,v_n)=det(s,v_1,\cdots,v_n)$$

Explanation of this formula: The point $s$ lies on the sphere which is itself embedded in $\mathbb R^{n+1}$, i.e. $s\in S^n\subset \mathbb R^{n+1}$.
The value $\omega_0(s) $ at $s$ of $\omega_0$ is an $n$-multilinear form on $T_s(S^n)\subset T_s(\mathbb R^{n+1})=\mathbb R^{n+1}$ and $T_s(S^n)$ consists of those $v\in \mathbb R^{n+1}$ orthogonal to the vector $s$, i.e. $T_s(S^n)=s^\perp$.
The formula says that the value of that multilinear function $\omega_0(s) $ on a $n$-tuple of vectors $v_1,\cdots,v_n\in T_s(S^n)$is the determinant of the matrix of the $n+1$ vectors $s,v_1,\cdots,v_n$ seen as column vectors in $\mathbb R^{n+1}$.

share|improve this answer
    
perhaps you wanted to say that ${\omega}_{0}(s, v_1,\ldots ,v_n) = det (v_{1}(s),\ldots ,v_{n}(s))$ –  Aspirin Mar 11 '12 at 1:40
    
Dear @Aspirin, no, no, I wrote what I meant. I have added a few words of explanation of the formula to clarify. –  Georges Elencwajg Mar 11 '12 at 10:26

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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