Suppose we have a degree $d$ map between two closed compact oriented surfaces $f : \Sigma \to \Sigma'$. What does knowing $d$ tell us about the map on $f_\star : H_1(\Sigma, \mathbb{Z}) \to H_1(\Sigma', \mathbb{Z})$?

  • $\begingroup$ @QiaochuYuan: I think genus $1$ is a special case since there are lots of self-maps from the torus to itself which are higher degree. I doubt there exists a map from a genus 2 surface to itself of degree $\neq0,\pm1$. $\endgroup$ Jun 8, 2017 at 20:27
  • 3
    $\begingroup$ @GrumpyParsnip: There isn't; see this question. $\endgroup$ Jun 8, 2017 at 21:25

2 Answers 2


Here's an example to think about. You can get a map of any degree from a surface of any genus $\Sigma_g$ to a sphere. Take $d$ disjoint disks in $\Sigma_g$ and crush the complement of these disks to a point. Map the disks onto the sphere in an orientation preserving way, with the boundary going to a fixed point on the sphere. This is a degree $d$ map, but the map on $H_1$ is trivial no matter what $d$. So in this case, the map's degree is not related to $H_1$ in any way.


denote by $I$ and $I'$ the intersection matrices on $\Sigma$ and $\Sigma'$, in standard basis having form $$ \begin{pmatrix} 0 & 1 & 0 & 0 & \ldots \\ -1 & 0 & 0 & 0 &\\ 0 & 0 & 0 & 1 \\ 0 & 0 & -1 & 0 \\ \vdots & & & & \ddots \end{pmatrix}. $$

then for matrix $A$ of the homomorphism $f_*$ the following condition holds: $$ A\cdot I\cdot A^t=d\cdot I'. $$ here $A^t$ corresponds to the operator $f^*:H^1(\Sigma')\to H^1(\Sigma)$, and the equation follows from the fact that $f^*$ is a ring homomorphism.

  • $\begingroup$ I tried this formula with two different examples and it didn't work in either one! The simplest one I tried was the map $f$ from a genus $3$ surface to a genus $2$ surface that crushes a subsurface homeomorphic to a torus with open disk removed to a point. Then the matrix $A$ will be a 4 x 6 matrix which is essentially a 4x4 identity matrix with two extra columns of $0$s. The degree of this map is $1$, but $A^t I' A$ is not any multiple of $I$. I also tried looking at genus $3$ as a double cover of genus $2$ but it didn't work in that case either. $\endgroup$ Jun 12, 2017 at 16:43
  • $\begingroup$ @GrumpyParsnip are you shure that $A^t I' A$ is not equal to $I$? $\endgroup$ Jun 13, 2017 at 14:50
  • $\begingroup$ Yes. It has zeroes in the lower corner. $\endgroup$ Jun 13, 2017 at 17:45
  • $\begingroup$ @GrumpyParsnip surely, it is false! the correct statement tells about dual matrix, corresponding the map $H^1\to H^1$ $\endgroup$ Jun 14, 2017 at 8:46
  • $\begingroup$ I see, that is much more believable and consistent with my genus 0 example. Do you have a reference or sketch of proof? $\endgroup$ Jun 14, 2017 at 16:06

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.