Quasilinear equations and Aleksandrov-Bakelman-Pucci estimates: Difference between pages

From nonlocal pde
(Difference between pages)
Jump to navigation Jump to search
imported>Luis
No edit summary
 
imported>Russell
No edit summary
 
Line 1: Line 1:
Quasilinear equations are those which are linear in all terms except for the highest order derivatives  (whether they are of fractional order or not).
The celebrated "Aleksandrov-Bakelman-Pucci Maximum Principle" (often abbreviated as "ABP Estimate") is a pointwise estimate for weak solutions of elliptic equations.  The strength of the ABP estimate is that it is the main tool in the theory of non-divergence elliptic equations which gives a pointwise bound on solutions in terms of a measure theoretic quantity of the equation.  It is not only the backbone of the regularity theory of fully nonlinear second order elliptic equations <ref name="CC"/> and more recently for [[Fully nonlinear integro-differential equations]] <ref name="CS"/>, but it also is essential in stochastic homogenization of uniformly elliptic equations in both the second order and integro-differential setting.  


For instance,  the following equations are all quasilinear (and the first two are NOT semilinear)
== The classical Aleksandrov-Bakelman-Pucci Theorem ==


=== Mean curvature flow ===
Let $u$ be a viscosity supersolution of the linear equation:
\[u_t-\mbox{div} \left ( \frac{\nabla u}{\sqrt{1+|\nabla u|^2}}\right ) = 0 \]


\[ Lu \leq f(x) \;\; x \in B_1\]
\[ u \geq 0 \;\; x \in \partial B_1\]
\[ Lu:=a_{ij}(x) u_{ij}(x)\]


=== [[Nonlocal porous medium equation]] ===
The coefficients $a_{ij}(x)$ are only assumed to be measurable functions such that for positive constants $\lambda<\Lambda$ we have
\[ u_t = \mbox{div} \left ( u \nabla \mathcal{K_\alpha} u\right ),\;\;\; \mathcal{K_\alpha} u = u * |x|^{-n+\alpha} \]


\[ \lambda |\xi|^2 \leq a_{ij}(x) \xi_i\xi_j \leq \Lambda |\xi|^2 \;\;\forall \; \xi \in \mathbb{R}^n \]


=== [[Semilinear equations|Hamilton-Jacobi with fractional diffusion]] ===
Moreover, the function $f$ is assumed to be continuous. Then, the ABP Theorem says that
\[ u_t  + (-\Delta)^s u + H(x,t,u,\nabla u)= 0.\]


\[ \sup \limits_{B_1}\; |u_-|^n \leq C_{n,\lambda,\Lambda} \int_{\{ u=\Gamma_u \} } f_+^n dx \]


Equations which are NOT quasilinear, and thus involve no linearity assumption of any sort, are called [[Fully nonlinear equations]], they include for instance the [[Monge Ampére Equation]] and [[Fully nonlinear integro-differential equations]]. Note that all [[Semilinear equations]] are automatically quasilinear.
Where $\Gamma_u$ is the "convex envelope" of $u$: it is the largest non-positive convex function in $B_2$ that lies above $u$ in $B_1$. The fact that the integration on the right hand side takes place only on the set where $u$ agrees with its convex envelope is an important feature of the estimate and it is not to be overlooked <ref name="CC"/>.


== ABP-type estimates for integro-differential equations ==


{{stub}}
The setting for integro-differential equations is similar, what changes are the operators: let $u$ be a viscosity supersolution of the equation
 
\[ Lu \leq f(x) \;\; x \in B_1\]
\[ u \geq 0 \;\; x \in  B_1^c\]
\[ Lu:= (2-\sigma)\int_{\R^n} \delta u(x,y)\frac{a(x,y)}{|y|^{n+\sigma}} dy\]
 
Here $\delta u(x,y):= u(x+y)+u(x-y)-2u(x)$ and $\sigma \in (0,2)$. The function $a(x,y)$ is only assumed to be measurable and such that for some $\Lambda\geq\lambda>0$ we have
 
\[ \lambda \leq  a(x,y) \leq  \Lambda \;\;\forall\;x,y\in \R^n\]
 
As in the second order ABP, the function $f$ is assumed to be continuous. Then, Caffarelli and Silvestre proved <ref name="CS"/> there is an estimate
 
\[ \inf \limits_{B_1} \; |u_-|^n \leq  C_{n,\lambda,\Lambda,\sigma}\sum \limits_{i=1}^m  ( \sup \limits_{Q_i^*} |f|^n) |Q^*_i|  \]
 
For some finite collection of non-overlapping cubes $\{Q_i \}_{i=1}^m$ that cover the set $\{ u=\Gamma_u\}$, each cube having non-zero intersection with this set.  Moreover, all the cubes have diameters $d_i \lesssim 2^{-\frac{1}{2-\sigma}}$. As before, $\Gamma_u$ denotes the "convex envelope" of $u$ in $B_2$.
 
Furthermore,  although the sharp constant may depend on $\sigma$, it is uniformly bounded for all $\sigma$ bounded away from zero. In particular, as $\sigma \to 2^-$ the above constant does not blow up, and since the diameter of the cubes goes to zero as $\sigma$ approaches $2$, one can check that this last estimate implies the second order ABP in the limit.
 
So far, only very little is known about an ABP result which can capture the more refined measure theoretic information of the right hand side of the equation for such integro-differential $L$ as above.  The only known result applies to a very restricted family of $L$ which are indexed by degenerate elliptic matrices:
 
\[
L(u,x):=  (2-\sigma)\int_{\R^n} \delta u(x,y)\frac{y^TA(x)y}{|y|^{n+\sigma+2}} dy,
\]
where $\text{Tr}(A(x))\geq \lambda$.  For these operators, Guillen and Schwab <ref name="GS"/> proved that
\[
\sup_{B_1}|u_-|\leq \frac{C(n,\sigma)}{\lambda}\lVert f^+\rVert^{(2-\sigma)/2}_{L^\infty(\mathcal C)}\Vert f^+\rVert^{\sigma/2}_{L^n(\mathcal C)},
\]
where $\mathcal C$ is the contact set between $u$ and a $\sigma$-order replacement for the convex envelope.
 
== References ==
 
{{reflist|refs=
<ref name="CC">{{Citation | last1=Caffarelli | first1=Luis A. | last2=Cabré | first2=Xavier | title=Fully nonlinear elliptic equations | publisher=[[American Mathematical Society]] | location=Providence, R.I. | series=American Mathematical Society Colloquium Publications | isbn=978-0-8218-0437-7 | year=1995 | volume=43}}</ref>
<ref name="CS">{{Citation | last1=Caffarelli | first1=Luis | last2=Silvestre | first2=Luis | title=Regularity theory for fully nonlinear integro-differential equations | url=http://dx.doi.org/10.1002/cpa.20274 | doi=10.1002/cpa.20274 | year=2009 | journal=[[Communications on Pure and Applied Mathematics]] | issn=0010-3640 | volume=62 | issue=5 | pages=597–638}}</ref>
<ref name="GS">{{Citation | last1=Guillen | first1=N. | last2=Schwab | first2=R. | title=Aleksandrov-Bakelman-Pucci Type Estimates For Integro-Differential Equations | year=2010 | journal=Arxiv preprint arXiv:1101.0279}}</ref>
}}

Latest revision as of 20:33, 11 March 2012

The celebrated "Aleksandrov-Bakelman-Pucci Maximum Principle" (often abbreviated as "ABP Estimate") is a pointwise estimate for weak solutions of elliptic equations. The strength of the ABP estimate is that it is the main tool in the theory of non-divergence elliptic equations which gives a pointwise bound on solutions in terms of a measure theoretic quantity of the equation. It is not only the backbone of the regularity theory of fully nonlinear second order elliptic equations [1] and more recently for Fully nonlinear integro-differential equations [2], but it also is essential in stochastic homogenization of uniformly elliptic equations in both the second order and integro-differential setting.

The classical Aleksandrov-Bakelman-Pucci Theorem

Let $u$ be a viscosity supersolution of the linear equation:

\[ Lu \leq f(x) \;\; x \in B_1\] \[ u \geq 0 \;\; x \in \partial B_1\] \[ Lu:=a_{ij}(x) u_{ij}(x)\]

The coefficients $a_{ij}(x)$ are only assumed to be measurable functions such that for positive constants $\lambda<\Lambda$ we have

\[ \lambda |\xi|^2 \leq a_{ij}(x) \xi_i\xi_j \leq \Lambda |\xi|^2 \;\;\forall \; \xi \in \mathbb{R}^n \]

Moreover, the function $f$ is assumed to be continuous. Then, the ABP Theorem says that

\[ \sup \limits_{B_1}\; |u_-|^n \leq C_{n,\lambda,\Lambda} \int_{\{ u=\Gamma_u \} } f_+^n dx \]

Where $\Gamma_u$ is the "convex envelope" of $u$: it is the largest non-positive convex function in $B_2$ that lies above $u$ in $B_1$. The fact that the integration on the right hand side takes place only on the set where $u$ agrees with its convex envelope is an important feature of the estimate and it is not to be overlooked [1].

ABP-type estimates for integro-differential equations

The setting for integro-differential equations is similar, what changes are the operators: let $u$ be a viscosity supersolution of the equation

\[ Lu \leq f(x) \;\; x \in B_1\] \[ u \geq 0 \;\; x \in B_1^c\] \[ Lu:= (2-\sigma)\int_{\R^n} \delta u(x,y)\frac{a(x,y)}{|y|^{n+\sigma}} dy\]

Here $\delta u(x,y):= u(x+y)+u(x-y)-2u(x)$ and $\sigma \in (0,2)$. The function $a(x,y)$ is only assumed to be measurable and such that for some $\Lambda\geq\lambda>0$ we have

\[ \lambda \leq a(x,y) \leq \Lambda \;\;\forall\;x,y\in \R^n\]

As in the second order ABP, the function $f$ is assumed to be continuous. Then, Caffarelli and Silvestre proved [2] there is an estimate

\[ \inf \limits_{B_1} \; |u_-|^n \leq C_{n,\lambda,\Lambda,\sigma}\sum \limits_{i=1}^m ( \sup \limits_{Q_i^*} |f|^n) |Q^*_i| \]

For some finite collection of non-overlapping cubes $\{Q_i \}_{i=1}^m$ that cover the set $\{ u=\Gamma_u\}$, each cube having non-zero intersection with this set. Moreover, all the cubes have diameters $d_i \lesssim 2^{-\frac{1}{2-\sigma}}$. As before, $\Gamma_u$ denotes the "convex envelope" of $u$ in $B_2$.

Furthermore, although the sharp constant may depend on $\sigma$, it is uniformly bounded for all $\sigma$ bounded away from zero. In particular, as $\sigma \to 2^-$ the above constant does not blow up, and since the diameter of the cubes goes to zero as $\sigma$ approaches $2$, one can check that this last estimate implies the second order ABP in the limit.

So far, only very little is known about an ABP result which can capture the more refined measure theoretic information of the right hand side of the equation for such integro-differential $L$ as above. The only known result applies to a very restricted family of $L$ which are indexed by degenerate elliptic matrices:

\[ L(u,x):= (2-\sigma)\int_{\R^n} \delta u(x,y)\frac{y^TA(x)y}{|y|^{n+\sigma+2}} dy, \] where $\text{Tr}(A(x))\geq \lambda$. For these operators, Guillen and Schwab [3] proved that \[ \sup_{B_1}|u_-|\leq \frac{C(n,\sigma)}{\lambda}\lVert f^+\rVert^{(2-\sigma)/2}_{L^\infty(\mathcal C)}\Vert f^+\rVert^{\sigma/2}_{L^n(\mathcal C)}, \] where $\mathcal C$ is the contact set between $u$ and a $\sigma$-order replacement for the convex envelope.

References

  1. 1.0 1.1 Caffarelli, Luis A.; Cabré, Xavier (1995), Fully nonlinear elliptic equations, American Mathematical Society Colloquium Publications, 43, Providence, R.I.: American Mathematical Society, ISBN 978-0-8218-0437-7 
  2. 2.0 2.1 Caffarelli, Luis; Silvestre, Luis (2009), "Regularity theory for fully nonlinear integro-differential equations", Communications on Pure and Applied Mathematics 62 (5): 597–638, doi:10.1002/cpa.20274, ISSN 0010-3640, http://dx.doi.org/10.1002/cpa.20274 
  3. Guillen, N.; Schwab, R. (2010), "Aleksandrov-Bakelman-Pucci Type Estimates For Integro-Differential Equations", Arxiv preprint arXiv:1101.0279