0

Prove that any smooth, orientable, and connected surface can be equipped with a Riemann surface structure (in the sense of a complex 1-dimensional manifold).

My idea is to modify the chart so that the transition map satisfies the Cauchy-Riemann equation and is thus holomorphic. Since the manifold is orientable, the jacobian of the transition maps everywhere is invertible. And say let $f = \phi_\alpha \circ \phi_\beta^{-1}$ for some transition maps on $U \subset R^2$. I want to compose each $\phi_\alpha$ with some $\psi_\alpha$ where $$J_{\psi_\alpha \circ \phi_\alpha \circ \phi_\beta^{-1}\circ \psi_\beta^{-1}}(\psi_\beta^{-1}(p))= J_{\psi_\alpha}(f(p))J_f(p)J_{\psi_\beta^{-1}}(\psi_\beta^{-1}(p))$$

such that $$J_{\psi_\alpha}(f(p)) = \begin{bmatrix} a &b \\ b & -a \end{bmatrix}^{\frac{1}{2}}\begin{bmatrix} \partial_xf_1(p) &\partial_yf_1(p) \\ \partial_xf_2(p) & \partial_yf_2(p) \end{bmatrix}^{-\frac{1}{2}}$$

and $$J_{\psi_\beta^{-1}}(\psi_\beta^{-1}(p)) = \begin{bmatrix} \partial_xf_1(p) &\partial_yf_1(p) \\ \partial_xf_2(p) & \partial_yf_2(p) \end{bmatrix}^{-\frac{1}{2}}\begin{bmatrix} a &b \\ b & -a \end{bmatrix}^{\frac{1}{2}}$$ for some $a,b$ properly chosen.

If such construction exists, then the jacobian for the new transition map is $\begin{bmatrix} a &b \\ b & -a \end{bmatrix}$, satisfying the Cuacy-Riemann equation and thus the transition holomorphic. Does such construction work and does there exist such $\psi_\alpha$? I think properly not because I didn't use the connectedness property. So is there any modification I can do to make this construction valid?

The One
  • 834
  • 1
    Not "is a Riemann surface" but "can be equipped with a Riemann surface structure." The proof though is more complicated, you should put a Riemannian metric on your surface and use the isothermal coordinates. – Moishe Kohan May 11 '21 at 22:54
  • Thanks for your comment! I wonder if there is any way to directly prove the result instead of utilizing the equivalent relationship between complex structure and conformal structure on riemann surface – The One May 11 '21 at 22:57
  • 1
    I do not think so. Keep in mind that at some point you should invoke paracompactness of your surface (every Riemann surface is paracompact). – Moishe Kohan May 11 '21 at 23:06
  • To conclude: This issue was discussed at MSE many times, for instance here, just keep in mind that that accepted answer is incomplete (in the opposite direction), I gave a detailed proof instead. I find it very unlikely that you will find a more direct proof which would avoid constructing an auxiliary Riemannian metric and isothermal coordinates. If you ever find such a proof, let me know. For now, I am voting to close this as a duplicate. – Moishe Kohan May 13 '21 at 23:21

0 Answers0