I am doing algebraic topology course and I'm trying to compute the singular homology for some spaces. However I am stuck doing that for $\mathbb{R}P^{2}$.

Compute the singular homology of $\mathbb{R}P^{2}$ (the real projective plane).

  • $\begingroup$ Have you studied the Mayer Vietoris sequence yet? $\endgroup$ – tharris Apr 19 '13 at 13:22
  • $\begingroup$ yes i studied it $\endgroup$ – Koam Apr 19 '13 at 13:28
  • $\begingroup$ Ok I'll add an answer using it below. $\endgroup$ – tharris Apr 19 '13 at 13:30

View the real projective plane as the union of a disk $D^{2}$ and a Möbius strip $M$intersecting in an annulus, so that the edge of the disk goes all the way around the boundary of the Möbius strip. So when we consider $D^{2} \cap M \simeq S^{1}$, the circle "wraps around" the disk twice. Note that $\mathbb{R}P^{2}$ is a connected 2-dimensional manifold so $H_{0}(\mathbb{R}P^{2}) = \mathbb{Z}$ and $H_{n}((\mathbb{R}P^{2}) = 0$ for $n \geq 3$.

Now $M \simeq S^{1}$ and $M \cap D^{2} \simeq S^{1}$, the Mayer Vietrois sequence is

$$H_{2}(D) \oplus H_{2}(S^{1}) \longrightarrow H_{2}(\mathbb{R}P^{2})\longrightarrow H_{1}(S^{1}) \longrightarrow H_{1}(D) \oplus H_{1}(S^{1}) \longrightarrow H_{1}(\mathbb{R}P^{2}) \longrightarrow H_{0}(S^{1}) \longrightarrow H_{0}(D) \oplus H_{0}(S^{1}) \longrightarrow H_{0}(\mathbb{R}P^{2})\longrightarrow 0,$$ giving

$$0 \longrightarrow H_{2}(\mathbb{R}P^{2})\longrightarrow \mathbb{Z} \longrightarrow \mathbb{Z} \longrightarrow H_{1}(\mathbb{R}P^{2}) \longrightarrow \mathbb{Z} \longrightarrow \mathbb{Z} \oplus \mathbb{Z} \longrightarrow \mathbb{Z} \longrightarrow 0.$$

The map including the annulus into the Möbius strip wraps twice around the circle that is a retract of the Möbius strip, so the first map $\mathbb{Z} \rightarrow \mathbb{Z}$ is given by multiplication by $2$. Hence the map is injective and $H_{2}(\mathbb{R}P^{2}) = 0$. Now the last 3 groups in the sequence form a short exact sequence, so the map $H_{1}(\mathbb{R}P^{2}) \rightarrow \mathbb{Z}$ must be the zero map as $\mathbb{Z} \rightarrow \mathbb{Z} \oplus \mathbb{Z}$ is injective. Hence we have a short exact sequence

$$ 0 \longrightarrow \mathbb{Z} \stackrel{2}{\longrightarrow} \mathbb{Z} \longrightarrow H_{1}(\mathbb{R}P^{2}) \longrightarrow 0,$$

so $H_{1}(\mathbb{R}P^{2}) = \mathbb{Z} / 2 \mathbb{Z}$.


A perhaps more aesthetically pleasing way is to use that singular and simplicial homology coincides, and then find a triangulation of $\mathbb{R}P^2$.

One such looks like this: enter image description here

  • $\begingroup$ Nice! I didn't know a triangulation :) $\endgroup$ – tharris Apr 19 '13 at 14:01
  • 3
    $\begingroup$ I have never seen anyone call finding a triangulatio and actually computing the homology from the corresponding simplical complex «aesthetically pleasing»! :-) $\endgroup$ – Mariano Suárez-Álvarez Apr 19 '13 at 18:12
  • 2
    $\begingroup$ @MarianoSuárez-Alvarez : I think it depends on how you find the triangulation. If you just draw a bunch of triangles and chop them into pieces until it works, that's far from aesthetically pleasing. But if you notice that you can take a regular icosahedron and identify antipodes to get a triangulation of the projective plane, then I like it. $\endgroup$ – Andreas Blass Apr 19 '13 at 18:39
  • 1
    $\begingroup$ Everybody loves an icosahedron. But then you have to actually compute! $\endgroup$ – Mariano Suárez-Álvarez Apr 19 '13 at 18:39

$\mathbb{R}P^2$ is a CW complex with one 0-cell $v$, one 1-cell $a$, one 2-cell $e$ which is attached via the word $a^2$. Cellular homology comes from the chain complex $$0\rightarrow H_2(e,a)\rightarrow H_1(a,v){\rightarrow }H_0(v,\emptyset)\rightarrow 0$$ where $H_i$ are the $i$-th singular homology groups of consecutive pairs of skeletons for the CW structure. Each of these groups in this complex is just a direct sum of copies of $\mathbb{Z}$ with a generator for each $i$-cell. In this case, $$0\rightarrow\mathbb{Z}\rightarrow\mathbb{Z}\rightarrow\mathbb{Z}\rightarrow 0.$$ The boundary maps are computed via the degree of the attaching maps for each cell that generates a $\mathbb{Z}$. In this case the only intersting boundary map is $d_2:[e]\mapsto 2[a]$ where 2 is the degree of the attaching map of $e$ to $a$. All other boundary maps are $0$. So the cellular homology groups (which are the same a singular homology groups) are $H_n(\mathbb{R}P^2)=0$ for $n>1$, $H_1(\mathbb{R}P^2)=\mathbb{Z}_2$, $H_0(\mathbb{R}P^2)=\mathbb{Z}.$

Cellular homology works very efficiently for computing homology groups of most familiar CW spaces. i.e., $\mathbb{C}P^n, \mathbb{R}P^n$, genus $g$ surfaces, etc. Hatcher has a lot of examples in the middle of his Homology chapter, and it's very readable.


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.