Take the 2-minute tour ×
MathOverflow is a question and answer site for professional mathematicians. It's 100% free, no registration required.

$M$ is a Riemannian manifold with metric $g$ and we have a map $F: M \to T^{\*}M$ with $F(p)=(p,f(p))$ with a 1-form $f$. On $T^{*}M$ we use the Sasaki-metric.

How can I prove or it is wrong?:

$F$ is harmonic iff $f$ is harmonic.

Thank you and best regards.

share|improve this question
2  
What is the difference between f and F? –  Michael Bächtold Jun 24 '10 at 12:47
    
Why don't you sit down and actually compute, relative to a local coordinate system, what the tension-field/energy-density is, and what the local coordinate expressions for harmonicity is for the map $F$ and for the one-form $f$? –  Willie Wong Jun 24 '10 at 12:52
    
When I compute this, I have to inverse the following metric $g^{\prime}_{ij}=g_{ij}+\left( \frac{\partial f^{k}}{\partial x^{i}}\delta_{kn}-f^{t}\delta_{tk}\Gamma_{in}^{k} \right)g^{nM} \left( \frac{\partial f^{a}}{\partial x^{j}}\delta_{aM}-f^{a}\delta_{ab}\Gamma_{jM}^{b} \right).$ Can you help me to get the inverse metric? –  Differentialgeometer Jun 24 '10 at 13:01
    
Oh, now I understood the question. –  Michael Bächtold Jun 24 '10 at 13:06
1  
What's the Sasaki metric on the cotangent bundle? I'm not familiar with this. (This looks like a straightforward formal computation, and it is unlikely that you need to find an explicit formula for the inverse of $g'_{ij}$ to do the calculation. Just call it $g'^{ij}$ and plunge ahead.) –  Deane Yang Jun 24 '10 at 13:17
show 4 more comments

1 Answer

It is perhaps easier to look at the the harmonicity as Euler Lagrange equations for some action functional. The energy density of the map $F$ can be computed to be $$e_F = N + |\nabla f|^2$$ directly using the definition of the Sasaki metric. $N$ is the dimension of the manifold. So the Euler-Lagrange equation gives that the equation satisfied by $f$ for $F$ to be a harmonic map is $$ \triangle_g f = 0$$ On the other hand, for $f$ to be a harmonic one-form, you need $$ (d\delta + \delta d) f = 0 $$ The Weitzenbock formula tells us that the two equations differ by a term coming from the Riemann curvature of $(M,g)$. So no, in general the two expressions are not equal.

share|improve this answer
add comment

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.