Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Let $\Omega\subset\mathbb{R}^d$ be open,and $P(D)=\operatorname{\sum_{|\alpha|\le N}}f_{\alpha}D^{\alpha}$ be an elliptic differential operator. Rudin proves in Functional Analysis Part II the Regularity Theorem for Elliptic Operators that if $f_{\alpha}$ are smooth and $f_{\alpha}$ are constants for $|\alpha|=N$, then $P(D)u$ is locally in $H_s(\Omega)$ if and only if $u$ is locally in $H_{s+N}(\Omega)$, where \begin{equation} H_s=\{u\in D'(\mathbb{R}^d):(1+|y|^2)^{s/2}\hat{u}\in\mathcal{L}^2\}. \end{equation}

I know almost nothing about partial differential equations, and am ignorant of special examples especially. I need some examples showing the conclusion is false for other types of differential operators. That is, something like $Lu=v$ where $v$ is good but $u$ is bad when $L$ is not elliptic.

Thanks!

share|improve this question

1 Answer 1

up vote 2 down vote accepted

Some trivial examples are always useful. We work on $\mathbb{R}^2$, with coordinates $(x,y)$. One can apply partial differential operators to distributions, and in particular to functions in $L^1_{\text{loc}}(\mathbb{R}^2)$.

  1. Take $D = \frac{\partial}{\partial x}$. Take any function as bad as you want, that depends only on $y$, and still belongs to $L^1_{\text{loc}}(\mathbb{R}^2)$. Then, $Du = 0$, the nicest, smoothest function you could possibly imagine, but still, $u$ is ugly.
  2. Take $D = \frac{\partial^2}{\partial x \partial y}$. Then any distribution $u$ has $Du = 0$, but this says nothing about $u$.
  3. Take $D = \frac{\partial^2}{\partial x^2} - \frac{\partial^2}{\partial y^2}$. The equation $Du = 0$ is the one dimensional wave equation. Any function $u$ of the form $u(x,y) = F(x - y) + G(x + y)$, where $F$ and $G$ are arbitrary, ugly as you want, functions solves $Du = 0$ formally. Not all such functions can be understood as distributions, but those that can, and for which you can perform integration by parts, yield examples of non-smooth distribution solutions. $D$ is an example of a hyperbolic operator.
share|improve this answer

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.