Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I am currently trying to understand the local expression of a (pseudo)differential operator $$ \int_{R^n} e^{(x - y)\cdot \xi} \sigma(x,\xi) \, d \xi $$ on a manifold $M$ (compact and boundaryless, say), where $\sigma(x,\xi)$, as usual, is the symbol of the operator. (alternatively one may also write this as $$ \int_{R^n} e^{(x - y)\cdot \xi} a(x,y,\xi) \, d \xi $$ where $a(x,y,\xi)$ belongs to the same symbol class as $\sigma$.)

now the picture I have in my head is that the variable $(x,\xi)$ stands for a local coordinate in the cotangent bundle $T^*M$. but then, at the same time, the variables $(x,y)$ are local coordinates for some subset $U \times V$ of $M \times M$, so I fear I am mixing something up here?

Also, I was wondering whether the dot product $$(x - y) \cdot \xi$$ in the exponential term has anything to do with the pairing of elements from $T^*M$ with elements of the tangent bundle $TM$? some sources for example write the dot product in terms of a braket, $$(x - y) \cdot \xi = \langle (x - y), \xi \rangle$$ which makes this even more suggestive. but I struggle to make sense of this, I don't really know how to relate $(x - y)$ to the tangent bundle $TM$, making it very likely that this picture is wrong anyways.

Many thanks for your comments!

share|cite|improve this question
up vote 3 down vote accepted

A $\Psi$DO $P$ on a compact manifold $M$ is (modulo smoothing operators) a finite sum $P = \sum_k P_k$, where each $P_k$ is in local coordinates a $\Psi$DO on $\mathbb{R}^n$ with compact support.

Each $P_k$ has a symbol $\sigma_k$ on $T^\ast \mathbb{R}^n$. Clutching them together with a fixed partition of unity gives a function on $T^\ast M$. In order to make this independent of the partition of unity, we have to quotient out the lower order symbols, and so we get the principal symbol $[\sigma_P]$ of $P$.

Now given a symbol $\sigma$ on $T^\ast M$ (not a principal symbol, but a concrete function on $T^\ast M$ satisfying the needed estimates in local charts), we can construct a $\Psi$DO $P$ on $M$ with the same principal symbol, i.e., $[\sigma_P] = [\sigma]$. This seems like what you want: Going from the symbol to the corresponding $\Psi$DO defined by it. But this construction is a little bit technical, since we first have to define the Fourier transform on a manifold. Details for this may be found in, e.g., Lawson, Michelsohn, Spin Geometry, Princeton University Press, 1989, at the end of Chapter III, §3.

share|cite|improve this answer

Your Answer


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.