Let $A\longrightarrow M$ and $B\longrightarrow N$ be Lie algebroids with anchors $\#_A$ and $\#_B$, respectively.

A morphism of Lie algebroids is a morphism of vector bundles $\Phi:A\longrightarrow B$ covering let's say $\phi:M\longrightarrow N$ such that the induced map $\Phi^*:\Omega^k(B)\longrightarrow \Omega^k(A)$ commutes with the exterior derivative. How can I show the following relation holds $$\#_B\circ \Phi=d\phi\circ \#_A?$$


$(i)$ $\Omega^k(A):=\Gamma(\Lambda^k A^*)$ for $k>0$ and $\Omega^0(A):=C^\infty(M)$ (that is, $\Omega^k(A)$ are the smooth sections of the vector bundle $\Lambda^k A^*$);

$(ii)$ The induced map $\Phi^*$ is given by $$(\Phi^* \omega)_p(v_1, \ldots, v_k)=\omega_{\phi(p)}(\Phi(v_1), \ldots, \Phi(v_k)).$$

$(iii)$ The exterior derivative $d_A$ is the map $d_A:\Omega^k(A)\longrightarrow \Omega^{k+1}(A)$ given by:

\begin{align*} \displaystyle (d_A\omega)(\alpha_0, \ldots, \alpha_k)&:=\sum_{j=0}^k (-1)^j \mathcal{L}_{\#_A\alpha_j}(\omega(\alpha_0, \ldots, \widehat{\alpha_j}, \ldots, \alpha_k))\\ &+\sum_{i<j} (-1)^{i+j} \omega([\alpha_i, \alpha_j], \alpha_0, \ldots, \widehat{\alpha_i}, \ldots, \widehat{\alpha_j}, \ldots, \alpha_k). \end{align*} Here we're using the isomorphism of $C^\infty(M)$-modules $$\Omega^k(A)\simeq \textrm{Alt}^k_{C^\infty(M)}(\Gamma(A), C^\infty(M)),$$ to define $d_A$. For degree zero: $$d_A(f)(\alpha):=\mathcal{L}_{\#_A \alpha}(f).$$

$(iv)$ The isomorphism $\Omega^k(A)\simeq \textrm{Alt}^k_{C^\infty(M)}(\Gamma(A), C^\infty(M))$ is given by: $$\omega\longmapsto ((\alpha_1, \ldots, \alpha_k)\longmapsto \omega(\alpha_1, \ldots, \alpha_k)),$$ where $\omega(\alpha_1, \ldots, \alpha_k):M\longrightarrow \mathbb R$ is the map: $$\omega(\alpha_1, \ldots, \alpha_k)(p):=\omega_p(\alpha_1(p), \ldots, \alpha_k(p)).$$ The inverse is given by: $$\theta\longmapsto (p\longmapsto \theta_p),$$ where $\theta_p:A_p\times \ldots\times A_p\longrightarrow \mathbb R$ is given by $$\theta_p(v_1, \ldots, v_k):=\theta(\alpha_1, \ldots, \alpha_k)(p),$$ where $\alpha_j\in \Gamma(A)$ are sections such that $\alpha_j(p)=v_j$ (which vary with $p$).

$(v)$ On the level of $\textrm{Alt}^k$ we could see $\Phi^*$ as given by:

\begin{align*} \displaystyle (\Phi^* \theta)(\alpha_1, \ldots, \alpha_k)(p)&=\theta(\beta_1, \ldots, \beta_k)(\phi(p)), \end{align*}

where $\beta_j\in \Gamma(B)$ are such that $\beta_j(\phi(p))=\Phi(\alpha_j(p))$.



A vector is determined by its action on functions as a directional derivative. This is just the Lie derivative in degree $0$ mentioned at the end of $(iii)$. Together with the naturality $(ii)$ of $d$ and the definition of $d\phi$, you get $$(\#_b\circ\Phi_p)(\alpha)(f)=d_B(f)_{\phi(p)}(\Phi\alpha)=(\Phi^*d_Bf)_p(\alpha)=(d_A(f\circ\phi))_p(\alpha)=(d\phi_p\circ\#_A)(\alpha)(f)$$ for all $f\in C^\infty(N)$ and all $\alpha\in A_p$. Your claim follows.

| cite | improve this answer | |

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.